Bitcoin Forum

https://i.imgur.com/5OC15IT.jpg

With the bull market mining is again profitable, but I haven't seen much info about new GPUs in here...
Would be nice to share what we know and always there is some guy from factory or wholesale which has good info about the new GPUs so hopefully we can hear good news from someone like that also :D
As far as I know the RX 600 series is a rebrand of old Polaris again and recently there are news about RX 3000 (5000) Navi series? What a stupid idea to change the series to be alike Nvidia's... Hopefully it will be RX 5000 at least.


RX 5700 benchmarks

CN-GPU - XMR-stak

---

Update 13 June:
RX 5700 https://www.amd.com/en/products/graphics/amd-radeon-rx-5700 (https://www.amd.com/en/products/graphics/amd-radeon-rx-5700)
RX 5700 XT https://www.amd.com/en/products/graphics/amd-radeon-rx-5700-xt (https://www.amd.com/en/products/graphics/amd-radeon-rx-5700-xt)

Update 21 June
https://www.pcgamer.com/amd-radeon-rx-5700-xt-specs-release-date-price-and-everything-we-know/

https://i.imgur.com/IJaJPRZ.jpg
https://i.imgur.com/zWFarsP.png
https://i.imgur.com/CpTvyw0.jpg

New NAVI RX 3000 (5000) specs
https://i.imgur.com/JmOdaUK.png (https://i.imgur.com/JmOdaUK.png)
Source: https://www.tomshardware.com/news/amd-navi-everything-we-know,38242.html (https://www.tomshardware.com/news/amd-navi-everything-we-know,38242.html)

Polaris rebrand RX 600
https://i.imgur.com/tbnffnp.png
Source: https://www.tomshardware.co.uk/amd-radeon-rx-640-radeon-630-graphics-card,news-60635.html (https://www.tomshardware.co.uk/amd-radeon-rx-640-radeon-630-graphics-card,news-60635.html)


Update 3 June 2019: https://www.pcgamesn.com/amd/navi-rx-5000-gpu-release-date-performance (https://www.pcgamesn.com/amd/navi-rx-5000-gpu-release-date-performance)

i am afraid navi prices will be a bit higher.

Hey, matthew is this already launched? As far as i know, if this is still rumour or speculation because I was reading from some sites. NaVi series is created to be a competitor for RX series, and that's the only thing that I can catch from AMD's live presentation about that.  :D
but if that's cheap and more efficient and I will move to that series. I just heard this series would be launched in the next months.

Yeah, just rumours. RX should be just rebranded Polaris, according to the Tom's Hardware the Navi should be pretty powerful and cheap.

Navi prototype
https://www.tomshardware.com/news/navi-prototype-cards,39528.html (https://www.tomshardware.com/news/navi-prototype-cards,39528.html)

These are just rumors for now. More info will be release at E3

The series is really interesting and the main thing is very competitive when compared with what other companies offer.
But it still needs to be tested.
Intel should definitely think about it and offer its own option or level the prices of its products so as not to lose the market.

Good news that they will release new AMD cards around June or July.

What do you think will be the hashrate of these cards?

It seems it is more cheaper than buying Nvidia cards. I hope that we will get news here in my country from the distributor about this card I'm planning to start mining with AMD and let me test if the hashrate is higher than rx400 series.

If the price is not inflated and the above is indeed MSRP, then I'm sure it will sell great.
AMD is really giving Intel and Nvidia a run for the money.
With Ryzen 3700 and RX3080, it will be a bomb sub $1000 PC to build in Q4 2019.

Didnt they say that the price for the 2 navi cards that they talked about would be 399 & 499 usd?

But If prices are as in the slide u linked it will surely be pure awesomeness.

Funny those price rumors, reminds us about ryzen 3000 series price rumors. If the gpu can compete x a rtx 2070, it will not cost 50% less than that. Rumors saying the 3080 will cost $249 is ludicrous. RTX 2070 price point is around $499, so expect the 3080 if is indeed on par x the rtx 2070 to cost as much as 20% cheaper and that is a lot, 20% cheaper is $100, so $399 and that is being very generous. Remember AMD knows stupid miners will pay a lot for them. But even if rx 3080 is $100 cheaper, I myself see nvidia having a slightly edge in performance and a little higher power consumption. I myself think rx 3080 will have a brighter future than the rtx 2070 in the next 2 years or so.

Yeah, but imagine if it would be true :D

https://i.imgur.com/eFfmdb3.jpg

This says more realistic pricing, RX 5080 should be priced at 499 USD.
https://www.pcgamesn.com/amd/navi-rx-5000-gpu-release-date-performance

Latest rumours point to the prices being very close to their respective nVidia equivalents. We'll see.
What will be most interesting imo is whether we'll be able to tweak GDDR6 timings. If so, there might be a good advantage to getting some Navi cards vs Team Green... which afaik have not opened up turing memory settings (read-only as I've been told).

Intel?  You sound like a bot that is post farming with generic responses.

That is on par with my analysis, amd is pricing it high because there will be stupid miners that will buy it because they know coin prices are rising and will keep rising. You have to justify the purchased with performance x power consumption and miners will want the rx 5080 because power consumption matters for them, gamers would careless about that. Stupid gamers trade off 5 fps for 300 watts hehe, their psu burns their houses for only 5 fps more performance and they are happy with it, miners trade off 20% lower consumption for 10% lower performance and that is where rx 5080 is heading x rtx 2070.

I'm tired with the bullshit, I want the days of polaris to return rx 480 for $160, not sure if we will see it return in our generation, 30 - 50 years or so. I think that is over if cryptocoins keep rising. AMD and Nvidia will always have an excuse to price higher their products if cryptocoins keep rising, I think both have been pumping cryptocoins hehe.

160 USD for RX 480 on Newegg? I wish we have something like this in EU also :D

New Sapphire 570s were £199 a few weeks back on Amazon.

https://www.pcgamesn.com/amd/navi-rx-5000-gpu-release-date-performance

They did it but it's still speculation. But i thought that that makes sense for the top of NAVI card will be priced at that rate. If that was true and i will try to get it rather than spend a lot of money on rtx2080. I can get two for how much money that i must spend to buy one rtx 2080.

They were $160 back in July 2016 and if the mining crazy never happened then the 5070 would be the same value next month.

So we got the information now.

379 USD and 449 USD for the new navi cards...

Will be intresting to see how they perform but pricing seems 100 USD to high for my taste.

As for me - we have to wait and to see the real prices. And to compare with NVIDIA GPUs. If everything would be similar - i`d prefer NVIDIA surely. But if would be like as TS says - AMD can become better choice.

Not sure what to think about RX 5700 / 5700 XT for mining purposes.
Raw computing power doesn't seem higher than Vega 64, same for memory bandwith. So unless the new architecture can feed the compute units and the memory faster and with more consistency, I don't think I'll purchase one of these.

While they had a good performance increase in games, same can't be said for mining. I still think nvidia is overall better for most algorithms and when I say most, I say at moment 90%.

I thought that they will have the same performance as radeon VII because they have used 7nm fabrication that similar with radeon VII.
While radeon VII is more efficient compared with vega 64. FYI vega 64 still used 14nm fabrication.

It will be the same as Vega 64 in terms of performance. Vega 7 has 1TB of memory bandwidth while 5700 is ~ 480 GB....they will just be more efficient than the 64.

I nevertheless hope their efficiency upgrade is real, which would mean less wasted clock circles giving more consistent workloads. It would be excellent for all core heavy algos and may allow profitable dual mining.

Like the post above, vega will still be better only in cases where an algorithm needs that hmb high memory bandwidth.

It may be true, but i don`t think that it can be better then NVIDIA in core algos. The memory always was the main profit of AMD and in this series there is no this profit. The own way - the dual mining may be, but even in this way i dont sure. Only wait to see prices and miners.

And another shit from Metroid.
We all know that you are NVIDIA fanboy, so please don´t talk such shit anymore  ::)

90% of algos are better for NVIDIA - you made my day dude

It is the truth with Nvidia. Has nothing to do with fanboy or not. Nvidia brings less frustation and more algos.

But anyways. Dont forget the new Nvidia "Super" series from 2060 to 2080 (maybe also the 2080 Ti). Which will attack AMD's new series and increase hashrate (for the same price as before).

The only thing that I see that is good here is the price cuts that the consumer will be having i the current GPUs in the market. Specially that Nvidia will have a refresh of their current line-up with higher frequencies, and if its true that the prices will be the same with the current line up then price cuts in the current gen is inevitable. I am also quite curious about what the new Radeon GPUs can offer as we all should know that raw performance is not the same thing as programs that are optimized to that hardware. We can site examples with Adobe premiere that is running better in Intel CPUs and the new Windows update that are better optimized for Ryzen.

Prices in the table of course are very attractive, but with the release of these devices prices will immediately change in a big way.

Guys, what about Vega 56? If I reflash the bios from Vega 64, will it give an additional profit? So far, I do not see much point in waiting for a new series from AMD, as the prices for Vega 56 are very "tasty"

Use this https://bitcointalk.org/index.php?topic=5123724.0

No point in speculating, let's wait until these cards get put in rigs so we can get real numbers on hashrate and power consumption.

Yes cards of course cheaper and better in line with the manufacturer Nvidia, but I wonder what speed they will give in mining. Now I think this question torments many people reading this topic

It's 8gb gddr6, should be equal to a RTX2070 which does 41mh/s in eth. Efficiency doesn't look good with that tdp.

Ok i think you guys can not handle AMD cards.

New Super series from Nvidia? Ok and AMD do nothing? No new cards?

So please give me all algos where NVIDIA is faster (think about you say 90% of all algos, so i want all algos where is NVIDIA better than AMD)

Sure, posting hundreds of coins/algos for just one single mad guy. There are only 2 great miners outthere for AMD, which can squeeze out the highest possible hashrate. All other miners cant get even close to that results. The top algos for AMD you can count on one single hand. Take down your purple sunglasses mate.

Buying AMD can work, but its risky and not that easy to handle. Still have tons of Vega's in my farm, but I hate every freaking update...

maybe, just maybe the new RDNA arhicecture will hande more algos than GNC.

Well, the big boys are coming out when Navi 20 comes to market. That would be somewhere next year from what I've heard. Looking for some 80 compute units, hope they won't be too powerhungry and make too much noise.

After 3 weeks we will know the real hashrates of Navi, I am expecting ~50MH/~130-140W.

On ETH / ETC, hashrate won't match the Vega 64, because Vega still has a better bandwith and the algo is memory intensive.

Anyway, new GPU's are allways good. Will be hoping for a better prices for previous generations.

Navi 20 coming out next year will beat Vega 64 easily. Not sure what the naming will be, but maybe it will be called 5900 XT or thereabout. Should be a winner at mining.

And NVIDIA updates Turing series. The last i`ve heard - that they would start sales in July 2019, so we have to wait about a month. But they are making just little customize, as a understand - some more cores, new memory, but nothing new.

Generally worth buying AMD? if i'm going to mine ETH ? It seems to say that the  NV 1600 is a good option. Can anyone have a link to see the comparison in Youtube ?

That already stated if the top series of NAVI has a similar spec with radeon VII but it less efficient as per my understanding from a lot of rumors that already spread among the news site.
The top type of navy GPU can beat vega 64 easily but it looks not for another type that placed below the top type of NAVI series.

ETH might POS in january 2020.

To beat Vega 64 on Etash, you only need more bandwith, so 16 GBPS GDDR6.

---

A 70$ used RX570 will give you the same HR on ETH (same as GTX1660 / 1660Ti).

50Mh where did you pull that number from your ass

we already know gddr6 + 256 and RTX2070 use same memory and with max OC (where NV have better memory controller) max is ~41mh so do your math

also 5700XT is 225w TDP, those card will be HUGE FLOP already!

NV now own them 100% in mining as lead with faster memory and price is gone  ;D

The only thing that could potentially make them NAVIs more interesting is if AMD allow the GDDR6 timings to be tweaked... Polaris/Vega style. So far, Turing doesn't allow timings to be modded so it could play to AMD's advantage again... Fingers crossed.

Okay, we will see.

1st leaked gaming benchmark, RX5700 is faster than RTX2060. https://spectatorx.wordpress.com/2019/06/15/radeon-navi-rx5700-first-real-benchmark-on-level-of-geforce-rtx2060-in-far-cry-5/

I don't think it will fall behind in terms of compute power and bandwith.

We'll see in 3 weeks for 5700 and 5700XT full testing.

is this already confirmed and i thought that ETA for POS still become a rumour. Because ethereum still not yet implement another hardfork to support casper POS.

I thought that he was mentioning the wrong name because 5900XT was a nvidia FX series.

In two months later we will see the real comparison.

January 3rd 2020 will be POS phase 0 for ETH. So the network will still work with POS.

Where did you get that info from? I was under the impression they were considering a PoW+PoS solution at first, possibly implementing ProgPow... Have you got any sources to the claim?

Phase 0 means you can stake for future POS rewards. Doesn't mean at all ETH will move to POS at this date. I may have been confusing you, my apologies.



https://www.youtube.com/watch?v=izspfej05lE&feature=youtu.be&t=2646

AMD already said they will block any bios mod, so any tweak you do over direct access to vbios will be doable on NV too.... so 57xx= rtx2070 1:1 + NV have 2-5% better memory controler

only if they would offer 256bit memory card for 200$, but that will not happen

OP updated
https://www.pcgamer.com/amd-radeon-rx-5700-xt-specs-release-date-price-and-everything-we-know/

Guess you should check where you pull your info from ;)

50MH should be easily achievable on this architecture (people forget they completely revamped the GCN cores). Wave 32 instructions are executed in a single clock cycle now, so even with no miner side optimizations (ie running old Wave64 code) the RDNA architecture can run split a single Wave 64 and run it in parallel essentially executing it twice as fast.

TDP is also TDP, with the changes above you can probably down clock these to half their advertised speed for ethash, youll be looking at under 100 watts for 50+ MH.

Also didn't mention the doubling of L0 to ALU bandwidth, and addition of L1 cache to the architecture, which clever developers can take advantage off.

Should definitely be interesting to play with :D

Bios block is no longer a problem, check claymore latest miner.

prices for video cards are certainly acceptable, but still I think that even 8 pieces to buy RX570 or RX580 and will be cheaper for mining.

RX 5700 GPU-Z Validation
https://www.techpowerup.com/gpuz/details/urekc

Who wants to bet the 5700 will be the same exact hardware as the 5700 XT and will be flashable to 5700 XT bios to unlock all goodies accordingly? ;D

Good ol'AMD. They wouldn't dare disappointing us, would they?

AMD Radeon RX 5700 XT 50th Anniversary Edition also available in the EU
https://www.guru3d.com/news-story/amd-radeon-rx-5700-xt-50th-anniversary-edition-also-available-in-the-eu.html

---

3DMark Time Spy Result shows Radeon RX 5700 XT close to GeForce RTX 2070
https://www.guru3d.com/news-story/3dmark-time-spy-result-shows-radeon-rx-5700-xt-passing-geforce-rtx-2070.html

It be good for mining? It seems that it does not have much computing power. it seems oriented for gaming not for mining

Who's grabbing a RX 5700 /RX5700XT on launch day?

At this moment RX 5700 XT don't work with claymore ETH miner. Drivers or miner dunoo. ???

have any one got it and try to mine with claymore ?

yes https://a.imge.to/2019/07/07/IPp8A.jpg

Just for visibility.

Got any numbers yet?

Since this card is not hbm then I guess ethash hashrate will be around 2070 hashrate which is 37mhs unmodded, modded and with the right timings I believe close to 50mhs. This card will fly on mtp and progpow.

Tests and reviews are out, but I can't find any with mining results...anyone had better luck and willing to share here?

256bit GDDR6 will be 34mhs stock up to 37-38mhs OC

ppl that say 50mhs are insane

Well, no idea about tehnicalities and difference between DDR5 and DDR6, but if we only
compare memory throughput of non-bios modified, non overclocked, new cards have
exactly 1.75 times better numbers

1.75 X 30 =?

(30 should be replaced with nonmodified rx580 hashrate at least at start, before people
manage to find out how to OC and modify timings of new cards, though)

What about monero hashrates, any thoughts?

soon we will se all.I wonder how performs nvidia dominate algos like x16r,equhash..

I think 50-60 Ethhash will be possible.

Ram subsystem is fast enough.

just got my 5700XT up and running with 19.7.1 drivers.
The only thing working and not correctly are:
Phoenix 4.2c ethash @ 4Mhash
Lolminer 0.8.3 Equihash192  @0.671 hash
Lolminer 0.8.3 Equihash144 @0.202 hash

so... yeah nothing else works.  :( :(

i guess once everyone starts updating their software to work with navi, then we should start getting some real results.

it seems that RDNA is a new architecture, nothing to do with GCN that all previous AMD gpu are based . need a miner with new kernel for navi

Have you tried ethminer?  It should have a fallback opencl kernel for cards that it doesn't have a binary kernel for.  It won't be as fast as a properly tuned kernel, but it would be an interesting data point.

Any idea which kernel is PhoenixMiner using? Have you tried with flag


clKernel <n> Type of OpenCL kernel: 0 - generic, 1 - optimized, 2 - alternative, 3 - turbo (1 is the default)

clKernel 0 gives @4Mhash also.

---

ethminer isn't recognizing it under opencl.  I tried the --cl-only flag.

Well that's a bummer.
If it's not too much effort, could you post what ethminer outputs and also the output of running clinfo on the machine with the 5700xt?

Thank you for the informative topic. AMD group is in an advancement as I understand. Nvidia must be careful. Especially these days when the indication of bull market sprouting everywhere.

So few tests of new AMD graphics cards :-*

https://www.cowcotland.com/articles/2749-4/test-amd-radeon-rx-5700-xt.html

No mining tested there  :(

Or here

http://www.pcekspert.com/clanak/amd-radeon-rx-5700-xt-rx-5700-recenzija/

Can't test mining when dev don't have the cards to update their mining softwares. It's RDNA, not CGN anymore.

Understandable...but i'd expect at some level openCL hardware abstraction would be able to hide
hardware and do it in unoptimized way...maybe those 4MH/s someone mentioned are
measured that way, but sounds way too slow

;D test for mining

Someone just has to compile the open source ethminer under new RDNA compiler. Even with no updates Wave64 code should be able to run nearly twice as fast on RDNA...ill see if I have some time to post a compiled binary later. No reason why these wouldn't run 40+ MH with unoptimized code/timings.

Based on reviews though the OpenCL compiler seems to have major issues so that could be part of the problem. (they probably did it on purpose to keep miners at bay until gamers buy them  ;D )

or they realized "shit... they don't mine.  Better reduce the price so gamers buy them"    :o

 ;D ;D ;D

It seems,
5700 come swith Micron D9WCW MT61K256M32JE-14:A they are specified to run at 1750 MHz (14 Gbps GDDR6 effective).
5700xt comes with Samsung K4Z80325BC-HC14 : they are specified to run at 1750 MHz (14 Gbps GDDR6 effective) too.

waiting for results with optimised kernels...

The recap of my 2.5h stream can be found here: https://www.youtube.com/watch?v=1j5WM29ECPk (https://www.youtube.com/watch?v=1j5WM29ECPk)

Tonights stream I will share what I got back from AMD

Thanks for that info, interesting.


Going on that logic then they will change that everytime they release a new gpu hehe, dont think that is the case but who really knows.

Ordered one 5700XT for testing. Devs you have some days for update your miners :D

Got the asus rx5700 xt today so hope there will be a miner soon;)

Something while no one is going to release a miner for RX 5700

making software takes time...plus developers have to buy card, too

So it'll probably take few more days

That's a complete cut and paste, and tells us nothing new. I see Claymore has a card and is testing it, which is excellent news. Fingers crossed we get RDNA mining figures soon...

I say that it is pretty much guaranteed, knowning claymore's reputation

What I fear is that we small miners will get a cut version and large mining farms a much more efficient one.

When others start to build code for RDNA then competition will force everyone to provide good performance to the public too.

The miners you have listed leverage a generic CL Kernel, that is the only reason they run at all. There are no released kernels atm for the new architecture. Once those are released you will see better results. The change in the way AMD handles compute on this card should be interesting. Excited to see post Kernel releases

Don't forget, these are PCI-E 4.0 cards, thats twice as fast loading DAG into card memory

(need X570 chipset and Ryzen 3xxx processor, though)

What about big navi coming out later. Will it be twice as powerfull as the 5700 XT?

Hi guys,

well, i'm also waiting for a miner...

meanwhile, I have a Brute Force software, that uses both Memory and Core intensively, and just for you to have an idea of the XR5700XT performance relative to other cards:

- RX470 - 35.5 to 41.6 MKeys/s
- GTX1050ti - ~27 MKeys/s
- GTX1060 - ~48 MKeys/s
- GTX1070ti - ~78 MKeys/s

- RX5700XT - 158 to 161 MKeys/s (YES, THAT'S A LOT for a 400$ Card!)

I Don't know if it's the new Architecture, but i was SHOCKED when I compiled the code and got it Runnig...

so, and i know this things are not linear, we have a factor of x3.87 compared to the RX470...

extrapolating to ETH:

RX470 - 30 to 32MH/s (mine does 33.2 MH/s :P)
RX5700XT - 116.1 to 123.84 MH/s

take this calculations with 'a full gram' of salt!

dag is not loading to the gpu memory. It is calculated by gpu in it originally

Well, who knows, I usually see info like this:

https://2miners.com/blog/ethereum-mining-with-gtx-1060-3gb-cards-stopped-mining-eth-etc/

"GTX 1060 3GB stopped mining ETC because of the DAG file. This file is a data block of considerable size that is loaded into the GPU memory during miner launch.
Without DAG (Directed Acyclic Graph), miners can’t search for block solutions in Ethash-based cryptocurrency networks. "

I admit I didn't research how exactly algo works, so maybe I'm relaying wrong info  ;D

1070Ti makes ~35MH/s, so 5700XT is making x2 - ~70MH/s, in reality maybe not more than 50MH/s.

Talked to the developer of the miner Wild Rig. He says that AMD does not provide tools for completely new videocaeds and even under RVII there are no normal tools for writing miners. That's sad. Apparently, we will not mine on 5700/5700XT soon.
Peбятки... вce плoxo. Mнoгиe нe знaют, нo пиcaть мaйнepы пoд AMД cлoжнee, чeм пoд Hвидия. T.к. AMД пpocтo нe дaeт нopмaльныx инcтpyмeнтoв. Инcтpyмeнтoвв нeт дaжe нa RVII. He тo чтo для 5700/5700XT. Bидимo, пoмaйним мы нa этиx кapтax нe cкopo.

Well, e.g. sgminer will work, xmrig and xmr-stak should too ) Any miner that uses kernels written with pure OpenCL and not compiling them into binaries, will work fine

Supports Navi in latest version: https://bitcointalk.org/index.php?topic=3167363.msg51785677#new

Anyone can post results?

German tom's hardware guide just got the 5700 xt within an inch of 2080 performance by boosting it to 2.276 GHz @1.25V with an EK waterblock.
Using 250 watts.

Google translated article:
https://translate.google.com/translate?sl=auto&tl=en&u=https%3A%2F%2Fwww.tomshw.de%2F2019%2F07%2F11%2Fungefesselt-radeon-rx-5700-xt-auf-ueber-22-ghz-uebertaktet-break-the-limits-mit-den-neuen-softpowerplaytables-fuer-die-rx-5700-und-rx-5700-xt%2F


 igor'sLAB

 

Unrestrained: Radeon RX 5700 XT overclocked to over 2.2GHz. Break the limits with the new SoftPowerPlayTables for the RX 5700 and RX 5700 XT

By: Igor Wallossek on 07/11/2019 at 9:41 am posted in graphics cards

In contrast to Nvidia's restrictions, you can virtually override and change the default settings of the Radeon cards - be it clock caps, power limit or even the voltage. Always provided you have the right profile at hand. The whole thing is called SoftPowerPlayTable and, for the end user quite simple, can also be quickly implemented in the registry. As with the Radeon VII and earlier at Vega, our forum user hellm has once again done a great job here and left me the new tables to experiment. Yes, the RTG (Radeon Tweaker Group) is alive!

Important preliminary remark

But I have to preface this with an important preliminary remark. The here linked in the article SoftPowerPlayTables (short SPPT) are an experimental intervention in the system and certainly not suitable for 24/7 use in production systems! Especially the now possible, higher voltages are a big unknown in the usual overclocker-Milchmädchen bill. Because if you overdo it, you might save the money for the next faster card at the moment, but you risk losing your hardware and thus the entire investment for a long time. The article should show what we could do with it and where there are possible limits is nothing more than a short description of what we have done. No liability is assumed for your hardware and the consequences, because you are completely responsible for that!

Test the limits!

It was a tedious approach, because there was no experience. By the way, you do not have to use a chiller to get similar values, simple water cooling is fine if you do not overdo it, so you do not even have to use a chiller. The delta between the water temperature and the GPU diode was in extreme cases about 35 degrees, with only medium overclocking then much lower. As long as you remain significantly below 50 ° C GPU temperature, even with the maximum clock actually no major hurdles. Below about 45 ° C is ideal, but requires a water temperature of 30 ° C maximum. You can do it that way, but not with air coolers.

AMD's Power Tune on the Navi cards is a pretty interesting thing, because the clock control is usually something of Chill, the power-saving feature of the AMD driver. The maximum clock (including voltage) is therefore not always on, but only if the graphics load requires it. This style of play is quite smart, of course leaves no constant tact to specify, but a good timing, which adjusts the game load dynamically. If you like it consistently, Witcher 3 invites you to stop. Then the stove beats as if Hansel and Gretel had to burn 100 witches at once.

In the WattMan you can, depending on the implemented SPPT, adjust the settings of the card (chip quality) and cooling:

For my test with the 2.2 GHz + peak clock I have chosen a setting with 2276 MHz for the clock and the full 1.25 Volt. The curve can be left as it is and later flattened if necessary. For the power consumption that plays almost no role. The setting for the power I have set to 95%. we'll see on the next page that this does not end in extravagant stream orgies, but it may be important for a smoother frametimes.

That's exactly what I noticed when I logged the frequencies. In the maximum profile, we see a slightly softer course with fewer drops in fluctuating loads. I have again tested Shadow of the Tomb Raider in Full HD, but all settings on Ultra. And no, it does NOT limit the CPU, as many suspect, because even an RTX 2080 Ti OC scales well with shaders and clock, which I've tried in a plausibility test, of course, before times. Only the GPU is limited here.

These high peak clock rates are quite remarkable for this card because whenever they occur, the graphics load is especially high.

I'll show you on the next page, that the performance has risen again neatly and the power consumption is NOT extremely exploded. The increase in tension at this point had no dramatic consequences. But, I must emphasize once again, the high voltages cause even with perfect cooling in the long run a creeping death. And that's why it's not for everyday use. You have to pay attention!

Test system and measurement methods

I have already described the new test system and the methodology in the basic article " How we test graphics cards , as of February 2017 " (English: " How We Test Graphics Cards ") in great detail and therefore reference only this detailed description for the sake of simplicity , So if you want to read everything exactly again, is invited to do so.

Those who are interested in the summary in tabular form will quickly get a quick overview:
Test systems and measuring rooms
Hardware:
   Intel Core i7-8700K @ 5 GHz
MSI Z370 Gaming Pro Carbon AC
16GB KFA2 DDR4 4000 Hall Of Fame
1x 1 TB Toshiba OCZ RD400 (M.2, System SSD)
2x 960 GB Toshiba OCZ TR150 (Storage, Images)
Be Quiet Dark Power Pro 11, 850-watt power supply
Cooling:
   Alphacool Ice Block XPX + EKWB Vector Radeon RX 5700
8x Be Quiet! Silent Wings 3 PWM, 2x 480mm Alphacool Nexxos UT60 480mm
Thermal Grizzly Kryonaut (for radiator change)
Casing:
   Lian Li PC-T70 with expansion kit and modifications
Modes: Open Benchtable, Closed Case
Monitor:    Eizo EV3237-BK
power consumption:
   Non-contact DC measurement on the PCIe slot (riser card)
Non-contact DC measurement on the external PCIe power supply
Direct voltage measurement on the respective feeders and on the power supply
2x Rohde & Schwarz HMO 3054, 500 MHz multichannel oscilloscope with memory function
4x Rohde & Schwarz HZO50, current clamp adapter (1 mA to 30 A, 100 KHz, DC)
4x Rohde & Schwarz HZ355, probe divider (10: 1, 500 MHz)
1x Rohde & Schwarz HMC 8012, digital multimeter with memory function
thermography:
   Optris PI640, infrared camera
PI Connect evaluation software with profiles
acoustics:
   NTI Audio M2211 (with calibration file)
Steinberg UR12 (with phantom power for the microphones)
Creative X7, Smaart v.7
own anechoic measuring room, 3.5 x 1.8 x 2.2 m (LxTxH)
Axial measurements, perpendicular to the center of the sound source (s), measuring distance 50 cm
Noise development in dBA (slow) as RTA measurement
Frequency spectrum as graphic
operating system    Windows 10 Pro (1903, all updates)



contents

    Introduction and Settings
    Gaming performance and power consumption
    SoftPowerPlayTables - Application and Download


f

<@1.25V>
God damn.  Well its going to be a lot of fun.  Like vega right after release

Water Blocks are already available to preorder!
https://www.ekwb.com/news/ek-vector-blocks-engineered-for-amd-navi-gpus/

I'm going to go with not more than about 56MH/s ... based on memory bandwidth given the algorithm is memory bound.

nanominer v1.5.1
@Grumpy-Dwarf Grumpy-Dwarf released this 18 hours ago

Added Ethash binaries for AMD Radeon RX 5700 and AMD Radeon RX 5700 XT cards. Performance is awful but maybe it will boost with next driver update.

https://github.com/nanopool/nanominer/releases

tests?

what is the performance with this new miner nanominer v1.5.1

PCI-E 4 isn't compatible with current GPU risers.  Some people have been able to mess with BIOS and get them working by changing them to pcie 3.0 speeds, but not all mobos work like this.  So plug the GPUs directly into the mobo 16x Slot to utilize it and even then you wont see any mining improvements.

PCI-E are backward compatible so you can run cards even on 1.1. Which lot of miners did on older mobos to get 6 and more cards running back then. Every mining mobo and most of the newer mobos have option to set PCI-E version.

That is true for pcie 3.0 and below but im talking about the newest 4.0 as it has a different power requirement and there are reports of it not working at lower speeds with some bios. it's something that needs to be fleshed out over time.

I didn't know that 4.0 is not backward compatible... that's bad.

Subbing - really sad to see this new arch having trouble getting off the ground..

I hope when the non-reference models are out we will have working miners.

Its AMD's openCL compiler...its totally bugged for the new arch. Im sure the ASM guys already have working kernels that are probably super fast but obviously wont put anything public for a while.

It seems 5700XT OpenCL score falls between 2080 and 2080TI:

https://www.notebookcheck.net/Powerful-system-sporting-AMD-s-Radeon-RX-5700-XT-racks-up-colossal-OpenCL-score-on-Geekbench.426956.0.html

Gigabyte RX5700
Cryptonight-GPU (Ryo)
1290 h/s stock
1400 h/s OC 1786mhz core
150W (GPUz)

which miner?

I was thinking its for nanominer.. .but i'm getting "could not find binary for gfx1010 device" for cnr, so maybe not for nanominer.

Before everyone starts buying out your local Bestbuy: Brute force is EXACTLY what is sounds like. I'm no fancy bit city AMD low level engineer, but as mining goes we are also looking at quite a few other factors that extrapolate fairly badly with a straight brute force test. That said, awesome work, and I'm sure the new AMD cards are gonna be pretty killer. I seriously doubt that a 5700 XT will push EThash higher than a Radeon VII for many reasons, but I look forward to be proven wrong.

My semi-educated guess is that the 5700 XT will land around 1080ti performance based on memory bandwidth, and chipset (as a wild median). Outperform 1080/1080ti (2070) in general, but otherwise not break the laws of progressive performance except perhaps in some cases.

Remember (and I'm trying real hard not to bash AMD here), AMD do make great 'on paper' hardware. But driver and utilization wise, they have literally never over-performed.

I got "No GPU devices were detected" error with Nanominer. Claymore's Dual Miner and PhoenixMiner can detect the card, so I don't know what's the problem with Nanominer. Btw I think that "awful" means ~3-4MH/s for Ethash mining with RX 5700, so the same what I can get using the other softwares.

I hope that the above mentioned developers will update their miner soon - but it won't be easy because this is a totally new architecture.

xmr-stak

soooo many lies here... do AMD ppz can`t stand the truth

RX5700 is very bad card with ~200w power usage on par with rtx2060 super deal with it!

and monero 1400h/s speed is for 6x5700 cards NOT ONE CARD!!!! FFS

stop the FUD

5700XR is very good card almost on par with  2080 , but for half the price


https://www.tomshw.de/wp-content/uploads/2019/07/Shadow-of-the-Tomb-Raider-FPS-1920x1080-DX12-Ultra-4.png

Depends on how much power it’s using to get this FPS

About 250W which is not too much above default consumption:

https://www.tomshw.de/2019/07/11/ungefesselt-radeon-rx-5700-xt-auf-ueber-22-ghz-uebertaktet-break-the-limits-mit-den-neuen-softpowerplaytables-fuer-die-rx-5700-und-rx-5700-xt/2/

https://www.tomshw.de/wp-content/uploads/2019/07/01-Power-Consumption-5.png

Definition of FUD being competing emotional claims without any supporting evidence. Word from the devs is the tools are broke, so any hashrate claims (whether high or low) are currently nonsense, unless you've got your own dev team.

Thanks for adding a well researched and beautifully punctuated counterpoint.

Actually the 5700xt is pretty good compared to what could offer in the past vega or even rx 590 x gtx 1070, both failed attempt, amd has now succeeded with 5700xt but to say is close to rtx 2080 is spreading misinformation around. 5700xt a little bit better than the rtx 2070 and 20% slower than the rtx 2080.

https://tpucdn.com/review/amd-radeon-rx-5700-xt/images/relative-performance_2560-1440.png

I wonder how is the hashrate, gaming has nothing to do with mining and should not serve as a comparison in this case because amd changed how it works so much that this might be the worst gpu to date for mining x price, so better to wait to see how miner developers work their magic here.

I wonder if there has been any progress on these cards yet   ::)

Like vega in 2017, nobody told anything, they just bought all the cards they could find and made huge profits during those 3 months ehhe

As usual new cards are expensive but the specs are quite nice.  Drivers are shit too so dont expect miracles right away

if 5700s start to become hard to find, that would seem a likely outcome

I bought at first day of release, cause hbm2&4096sp, no need waiting ppl review, do the math, i do remember when mining 20$/day per vega its glorius days

These cards should be no better than the Rtx on Eth mining as both cards use the Gddr6 memory.Maybe they will perform better because they have lower nm architechture than Rtx cards have.We need some actual facts from people mining with them.

modified an existing miner for monero using vulkan, not opencl. reaching 700h/s on rx5700 using 6gb vram of the 8gb.

 :o

https://twitter.com/BenjaminWegener/status/1152134210452410368

hopefully we will get higher rates after further optimizations

Perhaps not "better" for raw numbers, but maybe for efficiency

That is strange, I always thought monero algo doesn't need much ram...6GB is even more than memory hungry ethash

Not bad numbers as for stock timings. RX's without modified timings shows not much too.

My VEGA 64 on the same SOFTWARE at 1200/800 received 991h/s
UPD:
 "cards" : [
  {
    "index"    : "0",
    "cu"       : "64",
    "factor"   : "48",
    "worksize" : "8",
    "mem_chunk": "4"
  },

1200/800 = 1324 h/s

5700 have totaly diferenet arhiceture, called RDNA, not GNC like before. So, devs need to update thier miners with new kernels for this cards, till this happen... low hashrate.

in theory, 5700 should mine preety well. we will see if it will happen.

Guys, you should try to run this miner, maybe he will form the kernel and the graphics card can work.

https://github.com/fancyIX/sgminer-phi2-branch/releases

1800 its very close to vega 56, power consumption?

It is comparable with Vega card results and the main question here is what power consumption it have ?  My second question is what about ethash algo , have you tried it ? I think results on ethash should be more impressive .

take a screen shot by GPU-Z

Why? HWinfo is more precise (if we talk about power consumption)

Hi!
What version of the miner and driver are you using?

Rx 5700/XT convo only please.

Take the shilling to the scam section of the website

---

My regular RX 5700 does 1300 H/s on XMR stak cryptonight GPU on stock settings.

https://m.youtube.com/watch?v=CCQ0QJ_6_kQ
rx 5700 NOT xt 1200 h/s @ 1200/850 @ 80W

overclockd maybe 1700

Can you try latest SRB 1.9.2 with GPU algo and share the results ?
It has better results on Vegas than Xmr Stak, i wonder is this the case on 5700 too.
Thanks

Enumerating OpenCL devices, please wait...

GPU0: AMD Radeon RX 5700 [gfx1010] [8176 MB][CU: 18][I: 100.0][W: 8][T: 2][F: 0][BUS: 3]

-----------------------------------------------------------------------------------------
SRBMiner Cryptonight AMD GPU miner 1.9.2
DevFee ~0.85%

Press 's' to display stats
Press 'h' to display hashrate
Press 'r' to reload pools
Press 'p' to switch to the next pool
Press 'o' to switch to the previous pool
Press 0-9 to disable/enable GPU 0-9, shift+0-9 for GPU 10-19
-----------------------------------------------------------------------------------------

Algorithm               : Cryptonight Gpu
Algo switching          : disabled
Coin forking            : disabled
Gpu watchdog            : enabled [normal mode]
Startup monitor         : enabled
Tweaking                : disabled


[2019-07-21 21:46:33] Pools defined -> 1
[2019-07-21 21:46:33] Tweaking disabled - no supported devices found
3 errors generated.
[2019-07-21 21:46:34] Error CL_COMPILE_PROGRAM_FAILURE when calling clCompileProgram [ctx->Program] for DeviceID 0 (Thread 0)
Error while creating GPU threads.. Stopping miner process.

Shutting down miner...


C:\Users\user\Desktop\srbminer192>

---

FIY xmrig amd on cn/gpu yields ~300h/s with intensity 1700

anyone test this?

have you tried the last version XMR-Stak 2.10.6 that was released a few hours ago ?

Ouch, wonder why no one has reported this, i will look into it, thank you

Edit:

I wonder why the driver reports the card has 18 compute units, that card has 36 CU's no ?

I believe XMR stak states the same thing about the compute units.

what i don't get is why cryptonight/gpu from ryocoin works, whats the difference? i mean, i have to start xmrstak in low priority and the mouse freezes every 2 seconds for a half, but that could be the driver...

The mouse freezing problem is probably driver issue. You should reinstall it.

Can someone with RX 5700 try SRB 1.9.3 , it should now compile, but i couldn't test it because i don't own this gpu.
Also if it works then probably the parameters that auto setup gives will be bad. so playing with intensity,worksize,fragments can get better results.

https://bitcointalk.org/index.php?topic=3167363.msg51925083#msg51925083 (https://bitcointalk.org/index.php?topic=3167363.msg51925083#msg51925083)

I was able to get around 1630h/s with my regular 5700 card by running two threads on XMR Stak last night. watts was around 150 off the kilowatt wall meter

Algo was Cryptonight GPU

this is your 3th post and you lie, why

why am I lying?

there was a video of few + rx5700 doing ~1600-1700 and you say the same for one without screens not info, and just troll

thats my 5700

https://i.imgur.com/j9yzru5.jpg

Looks pretty good.

How much power?

My Vega 56 do 1800 each but are fairly power hungry.

Looks like i won't be able to fix anything without having a card :D

Btw, what clocks and settings on a V56 for 1800h/s on Xmr Stak?

He said 150W at wall. GPU-Z can't probably measure much more, it not even shows clocks. They added RX 5700 only preliminary.

Could you post more details maybe config? I've added your post to OP. Thx.
BTW Afterburner can set RX 5700 voltages, so if you didn't play with voltages already, you can do it with that.
https://www.guru3d.com/files-details/msi-afterburner-beta-download.html

My kilowatt wall meter said around 150w or 170w stock. I'm not too good at undervolting so I used the wattman auto undervolt template to get to 150w. doesn't seem to effect the hashrate too bad.
the memory clock can be increased to raise the hashrate to 1750h/s but the watts gets around to 230w at that point

Use the Afterburner instead, enable voltage control and you can start to play with undervolting and OC.
Are you sure that you adjusted only memory clocks? Higher memory clocks will need some extra power, but definitely not 80W more...

What drivers and settings are you using in your config?  I have tried using xmr stak a few times in the past week and have not had any success getting it to mine at all on a 5700 XT.

*hmmm I noticed GPUz has you on 19.7.1 and I'm on 19.7.2

after new lower prices this is not a bad variant i guess

Ill take another look when I get home today.

I believe I took the mem clock from 1750 to 1850 and thats when the jump happened. Ill do another test with afterburner instead and see if i get different results.

---

I downloaded the latest drivers from AMD for the rx 5700 card.

I used XMR STAK auto config for crypto gpu and it started to hash around 1300 h/s

I then went into the AMD config file and doubled the thread to 2. which increased the hashrate to around 1600h/s.

Im surprised the XT is not functioning the exact same way or better.

Got my own 5700 XT and what can I say... drivers are f*cking disaster %) They have finally implemented volatile keyword correctly for OpenCL, this is some kind of "plus", but this removed possibility to control register allocation a bit while compiling OpenCL kernels. reqd_work_group_size broken now, in some cases it can just crash compiler with error "ran out of registers". Memory operations... 10x times slower than on RX series ) AMD once again made dumb as hell OpenCL compiler, so developers should switch to GCN/RDNA assembler(or Vulkan at least), and showed how bad their driver team no matter what for so many years xD

Well, that's a bad news :(

Can you compile any of your CL kernels for 5700 ? so that they at least work  ;D

Yep, and compute power of this card is pretty good, close to Vega56/Vega64 with lower power consumption(can say about simple blake2b and bmw512 speeds) :) But... I don't understand this random nature of how it works xD Here some mind blowers:
- first time I see loop not unrolled: the optimizer was unable to perform the requested transformation... for some kernels
- in some cases compiler just throwing error when structure data types used and stop compile kernel; upd: found a problem, it's "tricky" pointer logic, not structure data types
- simple sha256 kernel fails with AMD HSA Code Object loading failed... wtf %)

Yes, computing part should be even faster than on Vega64, or at least this is what the numbers say :) RDNA looks good on 'paper', time will show is this the case in practice.
I don't have a 5700 card but when i tried to fix some of my kernels blindly, and compile them offline, the compiler was so friendly that it just said 3 errors found, but it won't tell me what are those :D

i could maybe provide u with tw access to my test rig with 5700xt that i got yesterday when i get back from buisness trip tommorow, if u are interested?

Compiled all my main kernels... well, need to say bye-bye to writing any code using OpenCL from now xD Results are bad for complex kernels and playing in game "compiler, please, don't spill so many registers" is not fun anymore.

what do you mean

The only way your going to get any good performance out of this architecture is with ASM kernels...which people are already running privately of course  ;)

Seems like whattomine is stating that ryo is the most profitable coin to mine at the moment for most Nvidia cards.

the hashrate for cryptonightgpu on a nvidia 2060 is comparable to amd 5700 at the moment

new drivers guys 19.7.3
has any one test it

Can anyone mine  cryptonightgpu with the 5700XT version?

Released WildRig 0.19.0 preview with RX 5700 series support.

Here some results for stock Sapphire RX 5700 XT(better results can be achieved with undervolt)
blake2b     - ~2.1GH/s
lyra2v3     - ~72MH/s
lyra2vc0ban - ~65MH/s

Download and Details (https://github.com/andru-kun/wildrig-multi/releases/tag/0.19.0)

Can anyone provide numbers for vega 64 / radeon vii in order to compare?

Vega 64 gives up to 90-100MH/s on Lyra2v3 as I know(with proper undervolt of course). And 5700 XT should lift more than current 72-78Mh/s, probably we will see performance near to Vega56/Vega64.

So let's put it this way, i think navi got alot of potentional, i just got around 65mh on lyra2v3 with undervolt to 850mv and 1650mhz on core, power consumtion is around 110w in gpuz, i can't measure from wall atm, but i will post as soon i can do it
https://i.imgur.com/6y8RohW.png

I've been mining to your VTC address for a little under 2 hours so far with a rx 5700 XT.  So far the average hashrate has been ~72 mh/s on stock settings, using 180 watts and Temp of ~75c (92c Hotspot) at 50% fan.  I have not tried adjusting any clocks or voltages.

Profitability calculator for RX 5700 TX (https://minerstat.com/hardware/amd-rx-5700-xt).

Can anyone offer any assistances trying to mine with 5700xt stock settings STAK 2.10.6 with currency set to cryptonight_GPU getting close to 1600hs but all shares rejected by nanoonpool, switching to Monero or another algo all shares accepted but 854mhs tried the other which is just terrible
any assistance?
SRBMiner works lovely on my Vega but 0mhs on 5700xt
XMR STAK 2.5 also can't get over 850 stock settings
tried 18.5.2 19.7.2 19.7.3 all drivers across all miners

I am very impressed with 72 Mhs for Lyra2rev, these cards seems like the new benchmarks for mining with graphic cards.I think that many new persons will come to mining with these cards soon.

could be a shot in the dark, but maybe try another pool for cn.gpu?

Yeah maybe, know of a decent pool? or I just might mine until I find something more profitable or etherum miner becomes avail....slightly disappointed :( but i'll live the 3900x tho 107 watts 1600mh on same miner super stoked about that and with new algo might go up in October!

1600mh cpu on which algo?
i use pool.ryo-currency.com

cryptonight_r  XmrStak 2.10.6 3900x 4.3 1.32vcore

Try hashvault.pro pool, all my shares get accepted on there. I'm doing around 1600 h/s with a regular 5700 on XMR STak.

are you using default setting or did you modify the amd config file to include 2 threads?

Hi guys again,

so, I've been messing around with the Ethminer Opencl Kernel, and I think I found out what is wrong with the RX 5700 family for mining ETH on OpenCL 2.0:

(Gonna need some help fixing it)

Let's Start:

After you download the new release's source code here:
https://github.com/ethereum-mining/ethminer/archive/v0.19.0-alpha.0.tar.gz
Extract it and find the kernels/cl/ folder,
Edit the ethash.cl file, change this line(236):
to this:
compile and run the miner!

That line is responsible for bottlenecking the hole code!

I am able to achieve 300+ MH/s without that line of code!
(LOOOOL not mining properly, of course, I get only rejected shares!)

So:

My thoughts on this, and after some testing, the problem is MEMORY ACCESS!
let's get back to the code:
if I change for a defined value eg. '0',
It runs @ ~80MH/s (I assume It's the real Speed),
but of course again, REJECTED SHARES, because is supposed to change according to 'lane_idx'!
Bottom line is, fetching this piece of code's value from memory is slowing down the hole thing!

Conclusion:

I need Help, since I'm no ETH nor OpenCL Expert...
We can assume from this the Theoretical Seed of ~80MH/s for mining ETH on the RX 5700 XT!
(not bad, but I was Expecting 100+ MH/s)
(maybe with specific optimizations, driver upgrade, etc, we could get better performance)

I tried I looked up some guides and failed to do it, how do I do 2 threads? I tried just copy paste the the section after the gpu but didn't work
could you show me what I need to edit and edit to what in the amd config file please? I can post back results

I would love to get 1600hs on that or this card maybe some settings you have are better? what's you power consumption

(long time Ethereum miner with claymore, new to XmrStak and Monero)

DDesignE

you expected 100+ for real man....

same with the one that posted cryptonight speed later we see its cryptonight GPU algo speed...

5700 IS PURE CRAP, if you want gddr6 get RTX2060 super its same memory + tensor core don`t beat the dead cat and spread insane expectations just because its damn shite AMD!

Can you copy and paste your amd config here?

Vega64 is doing 104mhs on lyra2v3 at 300$ new card
I doubt navi will pass vega series

They have released their midrange card first. Let's see what the 5800 (XT) will bring to the table.

Its probably still unoptimized, lets wait few months before making statements about hashrates

---

I copied this part

{ "index" : 0,
    "intensity" : 960, "worksize" : 8,
    "affine_to_cpu" : false, "strided_index" : 0, "mem_chunk" : 2,
    "unroll" : 1, "comp_mode" : true, "interleave" : 40
  },
Exactly like they tried intensity 960 890 800 480 all of them cause the amd driver to stop responding
Tried running stock underclocked unfervolted
I might just return this card maybe go for 5700 try it see how it goes
Can get vega 56 gigabyte gaming (worst one based off reviews) for 90 CAD less than 5700 and 170 cad less then the 5700xt I got (if the price dropped on the VII I’d do that but it’s double the 5700 price )
Might go Nvidia super 60 or super 70
Didn’t expect the amd to be so mining problematic

hi guys wanna someone to help this guy

he is mining cryptonight GPU algo

I stated with cpu 3900X and mined monero algo  but shares were rejected by nanopool I switched to cryponight_r algo solved the problem

I’m running another instance of stak  (I have made sure to set cpu to null and vice versa gpu instance have even ran standalone miners to eliminate any conflict same issue )

On both stak miners running cryptonight_gpu does give me 1600mhs on the 5700s however same problem all shares rejected (tested for 13 hrs)
Setting in the pools.txt to cryptonight_r reduces hashes in half but all shares accepted by nanopool
Dual thread attempts resulted in instant lock up even when halving the intensity on each thread

That’s where I am at currently I have about a week to return the gpu and probably pick up vega for cheaper and run srbminer it works flawless just wanted something newer and with more potential and to play a game once every couple months

this is what i have for my normal 5700

 * ],
 * If you do not wish to mine with your AMD GPU(s) then use:
 * "gpu_threads_conf" :
 * null,
 */

"gpu_threads_conf" : [
  // gpu: gfx1010  compute units: 18
  // memory:4048|4048|8048 MiB (used per thread|max per alloc|total free)
  { "index" : 0,
    "intensity" : 864, "worksize" : 8,
    "affine_to_cpu" : false, "strided_index" : 0, "mem_chunk" : 2,
    "unroll" : 1, "comp_mode" : true, "interleave" : 40
  },
  // gpu: gfx1010  compute units: 18
  // memory:4048|4048|8048 MiB (used per thread|max per alloc|total free)
  { "index" : 0,
    "intensity" : 864, "worksize" : 8,
    "affine_to_cpu" : false, "strided_index" : 0, "mem_chunk" : 2,
    "unroll" : 1, "comp_mode" : true, "interleave" : 40
  },

],

What drivers are you using?  I can only get my 5700 XT to mine with 1 thread (815-830 hashrate).  If I try 2 threads it will crash in a matter of seconds.

I also 5700 XT does not want to work in 2 thread. Drivers recent 19.7.4. What could be the problem?

maybe, they are waiting for AIB cards. No one likes blower edtion cards, since they are crap in cooling.

Are we all mining monero  ? Using cryptonight_r algo ?

Do you have 5700 XT? Works in 2 threads?

Nope I do not I’m trying to get that the guy above tho does have 5700 two threads

any new news

I think we all waiting patiently for dev to create miner compatible with new architecture.

Interesting AMD earnings mention the drop in crypto as being a profitability issue for them, but they've deliberately hobbled Navi. I wonder why? Hope to clear old stock to miners maybe?

Added Navi10 support to AMD Mem Tweak XL

would be nice if you guys tested a bit.

Link (https://bitcointalk.org/index.php?topic=5123724.msg52030843#msg52030843)

Cheers!

nice work, will try asap.

https://twitter.com/BenjaminWegener/status/1157178248918421505
talked with andru from wildrig, he found new env vars for radeon navi, tested them, but doesnt't improve much. just fyi

---

first startup ok, testmode works after reboot but then windows hangs like gpu driver killed. win 10 pro newest x64, 19.7.4 driver
everything ok now, wattman profile interfered, only strap injection doesn't work

it hangs if wattman isn't set to default

My test so far, got a 5700.
Running on xmr-stak, RYO, two threads 864. Using AMD Memory Tweak XL (thanks Eliovp), running at  1700mhz @ 905mv, stock mem. 1536h/s stable 5 hours so far. Temp 57C. Gpu only power draw 115.5W (gpu-z).

Any progress guys ?

dag fix driver again? amd love miners if make new driver again

new driver guys Radeon Software Adrenalin Edition 19.8.1
and it mention that
Radeon Performance Metrics may report incorrect VRAM utilization.


any one try it

no change for rx5709 using ryo on xmrstak

Everything points to 1.5mhs eth mining power! :p

Just mining ryo coin with a 5700 at the moment. seems to be the most profitable thing to do with this card

Why does Nvidia cards always work with miners on launch day with full hash power but amd struggles to get working right for many months?

What hashrate are you getting on it? And is it the XT version? Thanks! :)

Hi guys! I am new here and me too suffer from Rx5700 being too new.

After trial and error, using xmr-stak's validating intensity function I found the intensity to fill in the gap and it seems working.

I was bumping around with the interleave and it seems the lower the better until auto-search disappear. 3 or 5 seems to be the magic number.

sorry I am don't know where to host picture so please excuse me!

Now i can mine XMR.

here you go.

Oh yea, Please stand with Hong Kong. lol

https://imgur.com/tX1dd1D
https://imgur.com/Djs17cW

Hardware Settings:

Sapphire RX 5700 8GB enclosed in Razor Core X connected via Thunderbolt 3 to Dell XPS13 9370 i5-8520U.

Hope it helps :)

I get around 1600h/s with a regular rx 5700. I havent tested XT yet

---

Nice! thank you for the good work!

Did anyone try the latest version of Claymore Miner ?

yes, I tried, but nothing has changed. 3,5MH/s on a 5700XT.

But if you read Claymore's latest post in the dual miner topic he added a PS to his post:

"PS. AMD still has no public Navi drivers with DAG-fix, so currently 5700XT shows a very bad hashrate. But AMD already solved this issue and they promise to release public drivers soon..."

Why don`t try benchmark mode for DAG 1 or so..

test it with:
EthDcrMiner64.exe -benchmark 2

it will use just ~1GB of ram for DAG

The problem is on driver level and it isn't related to DAG size - as far as I know. I tried to run a benchmark but the result is nearly the same (instead of 3,5MH/s I got 3,6MH/s  ;D ;D )

Yes the problem is AMD intentionally crippled the driver for the ETH algorithm, because they knew it would do nearly 60MH in eth mining and they would not be able to get them in gamers hands.

He who bought an RX 5700XT to mine ETH, GG, you have a very nice gaming card but nothing else. Greed explotion!!!

Did anyone try 19.8.2 driver released 2 days ago?

https://www.amd.com/en/support/graphics/amd-radeon-5700-series/amd-radeon-rx-5700-series/amd-radeon-rx-5700-xt

3.4 MH/s even with 19.8.2 :)

Lol yep!! Amd fan boys got rekkkt!

Team green FTW

its mostl becase there is really not any good AMD mining softeare devs as compared to Nvidia which even further cripplies the Navi cards , REKTD REKTD lol

when Polaris came out, I remember someone modified drivers at the beggining.

I think you remember wrong, there were "blockchain drivers", but those were drivers
made by AMD specifically for mining

Hi,

it's not a good profitable period for mining, and after vacation i'm searching a new GPU to buy.

Are these new gpu good for mining?

Thanks

Think a guy named 'RobinHood' put out a hacked driver just a few hours/day or so before the official blockchain ones. Recall downloading it to test, then seeing the official ones before getting time.

Anyway, doesn't matter now!

here we go again :D
meh, as per last time, eventually it will be allowed by AMD
thats rough tho haha

yup, thats what I remember too, cause I did try them, later than blockchain drivers came out.

19.9.1 is out.  Anyone try it yet?

in claymore's thread a user reported 51Mh/s ETH

Currently running 54 MH, haven't started messing with timings yet :D

Also claymore is still running this as old GCN code, he hasn't updated to Wave32 which should significantly reduce core load needed and reduce power to under 100 watts.

Hi!
What is the miner name??

I am getting 53.8 Mh/s with some brief tuning in Wattman with driver 19.9.1 and Windows 10 Pro.  This is on an RX 5700 blower style GPU.  Using Claymore v15.0 to mine Ethereum.

After a few failed attempts and reboots, I am now stable with memory overclocked to 950, core clock is running between 1700 and 1740 Mhz at 940volts.  Temps are low 60's with fan running at 2400rpm (not loud at all).  It is pulling 180 watts at the wall, however.  Need to work on that!  No setting in Wattman to reduce memory volts, only core volts are available.  Still, not a bad start!

UPDATE> reducing the core clock down to high 1500's with core volts at 870 the watts at the wall are down to 160, temps are down, fan speed slower and it still mines at 53.5 Mh/s

NOTE> Any change to core clock during mining results in an instant freeze and reboot is required.  Other changes in Wattman do not seem to have the same result.  Not a fan of Wattman

I do use wave32 mode. During tests I tried wave64 as well, both showed same speed.

I'll release a new version of AMD Memtweak XL later today with latest driver support (Strap injection support for latest driver)
So you can play around on your Navi card :)

Cheers!


Edit: Here (https://github.com/Eliovp/AMDMemoryTweakXL/releases/tag/1.1.0.2)

http://eliovp.com/screens/1.1.0.2.JPG

Hello.
I am a newbie and thinking of starting mining with these cards and need the help of the community.
Are the common PCI risers adequate for mining with these cards? Can I also use my asrock H110 mobo?

Thanks in advance.

Claymore v15
Mining ETC on ethermine EU server
RX5700 @ 49.75Mh/s

Xmrig AMD
Supportxmr
XMR
RX5700 @ 1.1KHh/s

strap injecting, thats soundsinteresting.  ;D

12 hours trend of Rx5700 mining ETC
https://imgur.com/a/z9P4o9T

Anyone didn't try new driver on cnr?interesting that noone still don'tshare....

Can someone share a ProgPOW benchmark?

I was able to get just over 1100 h/s in monero using XMR-STAK.  TRM didn't even recognize the GPU.  Not a great result for such an advanced GPU in my opinion.  I'll stick with ETH for now.  Would be nice if we could tweak the memory timings - that would certainly help on all mining.

tested 2 miner on windows, no opencldevice found

Need more tests for cards RX 5700/XT

No, we haven't ported our (TRM) asm kernels for CN (and variants) to Navi, it's a pretty big effort, and toolchain support is not there yet. I'm not even sure we'll port the kernels given that we predict a shit hashrate, much like the one you're getting with xmr-stak. The reason is the switch to 128 byte L2 cachelines in the Navi architecture. Modern CN variants (CNv8, CN/r) use 64 bytes, older variants use 16 bytes whenever they touch mem. AMD GCN used 64 bytes cachelines, matching modern CN variants read/write patterns - no overhead.

So, what happens with CN for Navi is that for every mem access, you start by reading 64 bytes, but the mem controller on the Navi card actually reads 128 bytes since it always must pull a full cacheline into the L2 cache. This effectively doubles the consumed mem bandwidth, and you see a hashrate matching a pimped 580 rather than a Vega 56/64. In any case, crap CN hashrate compared to Vegas is the result.

For ethash, you always load 128 byte chunks anyway, matching the new L2 cacheline size, so no problem there.

What about the other algos you do such as mtp, lyra2v3, cuckatoo31 and cuckarood29?

https://m.twitch.tv/videos/458679885

We'll start porting properly as soon as the tools we use as a base for our build system are available for Navi, unfortunately for us and all users the progress is slow on that front :(. For the algos you mentioned, I think we'll start with things like MTP and x16r.

Thanks for the update.  I'm looking forward to it.  ;D ;D

whose mission will it be to use the 5700's 40 compute units?  mining software?

https://wccftech.com/flashing-the-radeon-rx-5700-with-the-rx-5700xt-vbios-for-more-performance/

There is 19.9.2, any increase in the hashrate?

With the upcoming mining algorithm of progpow being released in the release of berlin imminent I would say this is important to know too.
Before forking over alot of money for a card and it only getting half the hashrate after the algo changes over in sometime in early 2020.

Nothing for me. I have three and there's no apparent difference.

It is not wise to spend money right now when the future is not very clear.They have been saying to change to ProgPOW from sometime but they are hesitant to change.I think we need a wait and see approach to the changes,the cards perform well though in Cryptonight algorithm so if you are looking for long term usage you can go ahead and buy these cards.

Has anyone succeeded at getting these cards to work on Linux? Mine won't even fart.

Just covered a quick overview of two streams this past weekend on a 8x Sapphire 5700 XT Build. Watch the 7min recap here: https://youtu.be/hCA603-7C1U

As for the AMD 5700/XT support, the New Linux Kernel looks like it has Navi Support https://www.omgubuntu.co.uk/2019/09/linux-5-3-kernel-release-features

So basically we RX 5700XT users are stuck with ETHASH by Claymore now.
Using PhoenixMiner, somehow one of my two RX 5700 will crap out after sometimes and the one that is still mining will still be showing fully loaded after mining is stopped.

I have three XFX 5700XT cards and mining with PhoenixMiner 4.6c
No problems here. Hashing at ~52 MH/s each.

-tt 58 -cvddc 750 -cclock 1300 -fanmin 0 -mclock 900 -amd -mode 1 -clKernel 1 -clNew 1

good for you. Maybe because I force copy PhoenixMiner 4.6c latest exe into nicehash miner plugin folder, it's not stable.

At this point why would anybody buy a rx 5700XT over a regular 5700 if they almost produce the same hashrate?

Did you try running it without Nicehash?

Personally, I don't use Nicehash, I might might on their pool if it's more profitable but I don't use their program. For profit switching I use Awesome Miner.

Only reason would be if the RX can mine other coins, might be a better at other algos.  The XT can get 77 on vert coin.

Yeah I would agree with you there. I was just checking the dag file size for mining ethereum and it is at 3.23gb so anything below 8gb wont be able to mine the current algorithm when it reaches the 4 gb size limit.
Maybe mining another algorithm such as cryptonight you recommended would be a better option for the longevity of these cards thinking of future proofing the equipment you are purchasing now.
Best thing would be to wait to see where it all goes for the ethereum algo for now.


There has been someone I was watching on a youtube video which mentioned they were able to flash a 5700 to 5700xt and it worked fine for mining ethash. After the flashing it had all the same memory clocks as an xt version after checking them compared to before and after.
So it worked! :)

please make OCLMembench test

Can anyone perhaps give us some benchmarks with the XT. I seems like everyone wants to keep quiet about the results lol

Good question, i too would like to know, WHY?

But if i had to speculate, some people will buy it just because it is more expensive and makes it seem more premium. I know some people who are like this and are fixated on proving to the world that they are rich. All we can do is pray for them. lol.

I bought 3 open box xts, hoping that when miners are developed other algos might be faster on an xt.  A gamble buy we will see.

How much did you get them for? It's sad that you can't buy open box items from Amazon or Newegg if you live outside the US. On the other hand, in our country there aren't very many open box opportunities. Probably because companies here are too cheap to offer discounted prices even on items with damaged boxes.

has anyone been able to mine the new ravencoin algo with these cards?

As far as I know there isn't a miner that supports X16R (any variant) on Navi GPUs.  Overall it's been very slow getting any software for these GPUs to mine with, so I'm glad I only bought 1 of them to game test.  It is a nice GPU for gaming and does well on Eth at least.

I don’t have high hopes that it’ll do well on RVN.  AMD was poor on x16r and from what I can see hash rate wise there’s not much different between x16r and x16r2. 

What's the guess on eth performance on the new rx 5500? My guess is similar to the 5700/XT.

128bit bus,lower bandwidth than stock 580, i can see something around 30mh +/-

Indeed, I expect 35 top...some architectural changes would maybe help,
but its very bandwidth limited compared to 5700

But more realistic, probably 25-30 range

Seeing 110w stock TDP i was kinda dissapointed, since i was expecting that card to come without 6pin or 8 pin, but if there is possiblity to underclock it all the way to 750mv like u can do on 5700 series, 50-55w for 30mh would be really nice. that would be half power of 580.

If priced at 150 bucks, requiring 55 w for 30 Mh/s, with the current bitcoin price, the break even point would be 2 years. Much better than Navi 10's 3 years.
Hoping that BTC/USD will go to 15000 again soon :P

5500 is for now only for OEM.

I had hoped for a 5600 with a 6pin and something like 130 watt with GDDR6 and 256 bus which would be an ideal card for many scenarios including mining. Who knows, maybe AMD will surprise us all

There was never such slow adoption of new graphic card by mining devs, I think

Navi family is intentionally crippled by drivers, I suppose

Is this card only suitable for ethash?

Yes, AMD targets gamers and TMSC 7nm nodes are all 100% full. If the drivers are not crippled, miners would buy more 5700s and price would go up by a lot.

Ethash right now most profitable, but 5700/5700XT capable to mine mtp on pair with 2060/2070

How's the hashrate for MTP?

currently 2.9MH/s per 5700XT and 2.5MH/s per 5700 using wildrig multi.

I'll have to give it a try. I only have the XT versions.
Do you use F2Pool? What are optimal OC settings?
I just entered 2.9 in whattomine and I'm not impressed at all with the results.

The problem of AMD cards is the power consumption issue, I hope that this latest release has the technology to reduce the power consumption.

Anyone else able to squeeze a little more juice out of these cards?

modify timings/straps on the fly or bios mod but rx 5700 at the moment  are unsoported  :'(

I'm worried that we're all going to be screwed if ETH/ETC goes POS... or even ProgPow. I must have have to sell my 5700 cards.

Can it get less than 800mv in GPU and less than 850mv in VRAM? 5700 No XT.

U can get core to 712mv, but u need to modify registry files of amd drivers with Amd Tweaker tool or More power tool, it is pretty much complicated and can make card go unstable, since you will be lowering SoC voltage and some other voltages that does not matter for mining eth.

lolminer can mine grin31 and 29, beam a few other coins.  I was getting .95 gphs on grin 31 on my 5700s and 1.15 on my 5700 xts.

what's your power draw for grin 31?

What hashrate 5700 and XT on BFC coin? Supported by Gminer and NBminer

If we talk about Ethash algo, AMD has much better hash per watt ratio than Nvidia cards...both in previous and current card generation

eth is inevitably going to transit over to pos consensus in a year or two
it is delayed for so long but i am sure it will happen
once it is done
a lot of ethash based coins will follow the suit too
if by that time there is no other pow coin that can take over the load for gpu mining
than it will suffer a lot and it is also suffering right now because of the price

any tests on the rx 5600 XT yet?

sadly few months later of your post, 5600 is 192 bus and price is 290 usd, amd is not pricing it based on mining, that is for sure, i mean the 5700 can be bought for $300.

before the price leveled out, I was able to grab 2 visiontek 5700 cards from dell.com for 238 a piece. the price was set at 279 on front of site, I contacted dell chat support to see if they can give me a further student/military discount on top and the rep online approved a 15% discount for me.

Thinking about it now I wish I could of maxed out my order and resold a bunch on craigslist/ebay to come up on a instant profit but I was focused on finishing up my rx5700 rig instead.

Great prices for them 5700's!!

Once the 5600 has been out a few weeks it will drop to sensible levels and be a very efficient card for the hash. Something between the 580 and vega but less power so pretty decent.

People held on to purchase the 5700, hoping the 5600 would be as good and a lot cheaper, funny, now the 5700 seems much more expensive than used to be and people are buying it. As a rule, anything less than a 256 bit memory bus is trash. I hope nvidia returns to a competitive approach on the nm again, 12nm x 7nm is not working anymore. AMD knows it and is not lowering prices, i mean 560x used to be $120 ~ $140, amd is asking double of that because nvidia has nothing of same performance and price, this blame is on nvidia, success on amd for the elaborated plan. As it stands, i see amd winning in all fronts, cpus and gpus.

RX 5600 XT goes 40 Mh/s ethash at 90w

20% improvement over rx 580 which does 32 mh/s with 130w, rx 580 is been sold for $130 or so I think, no point in either one for mining other than data research.

How to undervolt RX5700 below 800mv? In MSI Afterburner 800mv is the minimum.

My results with 5700: https://youtu.be/0NfqtrsILYY

https://www.igorslab.de/en/radeon-rx-5600-xt-with-new-bios-and-stable-ram-without-flash-instructions-amd-limits-and-benchmark-morepowertool-tutorial/

I'm glad someone take the step to show how capable these new AMD GPUs are, I've looked every where online and I can't find a single information, now that I've seen the results I'm not impressed at all, CPUs can actually do better if mining Monero

Newsflash: algorithm that is designed to run better on CPUs than on GPUs actually run better on CPUs than GPUs

They say there will be a new video card from AMD Navi x2-probably it will bring 120 Mh

the new amd drivers is making the rx 5700 mining kind of buggy. i would avoid for now.

I don't know if this is of any interest to you (no experience with mining), but I made a GUI for easy powerplay table editing of Navi 10 cards on Linux. It supports multiple GPUs (though I only have one card myself so it's not tested, report issues if there are any, and let me know if you think some important settings are missing). Code and more information at https://github.com/azeam/powerupp (https://github.com/azeam/powerupp)

Dude that looks great! What framework did you use for the GUI?

Thanks! It's GTK3, built with Glade.

Nice what kernel/driver version are you running? Lost track of what's going on with AMD on linux side with their open source vs closed drivers.

I'm using Ubuntu 20.04 with kernel 5.6-rc1 and drivers from Oibafs ppa right now, but the card has been working just fine for me since kernel 5.4 (the first I tried iirc, got the card before christmas) and during the 5.5 release candidates (with Oibaf drivers).

How to Undervolt/OC and mine Ethereum with Navi 5700 (Tutorial) - https://www.youtube.com/watch?v=nhuDw8ui16A

How to mod Navi 5700's BIOS (Tutorial) - https://www.youtube.com/watch?v=LhT74CBcFvM

Sapphire Pulse 5700 - 56 MH/s Hashrate - https://www.youtube.com/watch?v=mUhno8pVRhI

what is best settings to mine ETH
on hiveos
phoneix or claymore

and how to know memory manfucturer

Depends on who you ask.

Depends on which version of card you have.

I prefer PhoenixMiner.
You can get your memory details using GPU-Z.

GPU-Z doesnot show memory manfucturer

Do we have figures of the 5600XT hashrate ?

Look under Memory Type.
What does it say?

It seems to vary.

If you use the MorePowerTool and Red Bios Editor you can manage between 54MH/s and 58MH/s.

I personally use these settings:
-mi 14 -wdog 0 -cclock 1250 -cvddc 743 -mclock 1800 -mvddc 850 -tstop 75 -tstart 60 -hstats 2 -tt -30 -mt 2 and result in ~55MH/s

There are a lot of videos and guides.
I just watched this one: https://www.youtube.com/watch?v=LwGUJjQgj38&t=608s

Thanks but I'm looking for the 5600, not the 5700.

Sorry, my bad.

Navi 5700 - Phoenix miner vs Claymore miner comparison - https://www.youtube.com/watch?v=qzbwVkW5QyI

5700 to 5700XT Bios Flash - Ethereum Mining Comparison - https://www.youtube.com/watch?v=H05gGdQHgFU

It's 34Mh @ stock 36Mh with mem OC, but i have the slower card with 12gbps ram.

Thanks, that's the wattage please ?

aren't all 5600 with 14gbs ram, well after you update the bios. The first release were 12gbs, but later than they gave out bios update. at least there was a lot of talking

Yes, there are Bios updates for certain cards but not for all.  Unfortunatelly - despite the fact that mine is an OC card - it does not have Bios update.

According to GPU-z its 77W.

Hello,

i am building a 12x5700 MSI EVOKE.

So far so good, I modded the timings and I am getting 56Mh on all cards.

I am using the morepowertool to undervolt the core to 750mv.

GB Fintech B250 MB with G3930 and 4GB RAM
HP Server PSU 1200W 94%
Phoenix 4.9C on Windows

Problem is a am getting about 140W from the wall per card. I see people doing under 100W. What I am missing?

Thank you!

Set core state to 2 and core clocks between 1180-1300. Then VDDC 750-900 range

How do I set the core state?

Its set to state 2 and vddc is 750mv. Anyway I guess thats the best my cards can do.

Btw I googled minerstat because of your post. Awesome stuff, trying it out now on 1 rig.

130-140W is normal consumption, this GPU cannot be under 100W from the wall!

Agree. Same results from my rigs even with vddc 706mv...

Thank you gentlemen.

My observations is that under 740mv core clocks start dropping under 1200mhz and hashrate suffers quickly. 56.3Mh requires 1280Mhz Core or more.

Sapphire Pulse 5700 - KAWPOW Hashrate - https://www.youtube.com/watch?v=S7_k6iwcK1M

One more question, after I added all the GPUs I cant reach the 56Mh from before.

Does it have anything to do with the RAM or CPU? Any ideas?

Running on 4GB ram and G3930 CPU

How many do you have?  Are they all MSI EVOKE?  What hashrate are some of the card hashing at?

Yes, MSI EVOKE.

When I was setting up the frame I had 7 installed and achieving 56Mh at 1280Mhz Core, 910 Mem, 750mv on the core.

Now I installed all 12 and with these settings I am getting 53.5Mh.

I get the same results on Windows and HIVE OS

Is it the situation when you start the miner they achieve 56mhs but after 4-5 minutes they reduce to 53.5?  Or is it immediately when you start the miner the hash at 53.5 and remain there?

As soon as I turn it on all are stable on 53Mh.

I think I will unplug 5 cards and see what happens as soon as get to the mine.

Gotcha.  I was going to say if your cards start out hashing at 56 then drop to 53.5, the likely culprit is memory temperature is too high and the card is throttling back.  But this would not be the case if it immediately starts at 53.5 and stays there.

Temps are good.

I wonder if it needs more ram or a faster cpu to handle 12GPUs.

amd release new card rdna2 in this year?

Guys, is RX5700 better than Radeon VII ? Because presently the Radeon VII is the highest AMD graphics card with better hashrate when mining cryptonight coins, I haven't have the privilege to try them head to head

Better have a Vega 56 with Vega 64 bios. VII is ways too expensive.

I had 12 Radeon VII's and 9 burned up and had to be RMAed.  Radeon VIIs are notorious for failing, I wish it wasn't so or I would have lots of them.  RX 5700 is great I have 18 with zero problems.

RedPandaMining had to RMA all of his VII's in one of his videos because he had them connected to breakout boards.

I bought a new card, RX 5700XT Gigabyte, I can do 53.5 with 105w 1280 cc, 735mv, Memory at 1870mhz
The core temperature is very good, 60c to 65c, but the memory temperature is 102c  ???

is it normal to have such temperature at memory? I know these new AMDs are hot, but never seen this temperature on my GPUs

Its probably one of those early cards with faulty coolers that have some distance between cooler and memory

This is my temps running Furmark

https://i.imgur.com/bQcUQyr.gif

I'm doubt, because this card was bought last week, it's a new batch of cards, I'm aware that some of cards have that problems with thermal pads not touching dissipators
Anyone with 5700XT can tell me the normal memory temperatures?

I search in a lot of places, there is some talk about this temp being only the junction temperature of the memory, not the surface temp as we are used to measure in GPUs

You can try increasing the fan speed. If mem temperature barely decreases, it could be bad contact between the memory modules and the heatsink.
That being said, memory temps on reference cards are around 90º with auto fan.

Also, my experience with Gigabyte cards is that memory cooling is not great. I wouldn't be surprised that is the case.

I already did some adjustments on the fan curve and updated the bios of this GPU
I didn't disassembly this card because it's under warranty, but looking to the backplate it looks good, the thermal pads are in touch with the heat sink and there's thermal pads between the board and the backplate too

I tested with some games, and it looks good, 74°c on the vram, but with mining or under stress it's 100°c

I'll wait for more tips, and if someone have this same model, please share your temps

Thanks

Good morning to everybody

I have modified the original bios of my Gigabyte 5700XT, only copying the value of the VRAM that appeared in the 1550 field in the rest of the fields. I have flashed that BIOS and now I get 56/57 MHS with a consumption of 123W.

I am a newbie to AMD and I ask for help, how can I modify the BIOS to obtain less electricity consumption? Is there a way to reach 61Mhs as I have seen in some Internet video?

I have no AMD experience, I have never modified BIOS, can I download a Bios Mod for my card from somewhere?

Thank you very much and sorry for the inconvenience.

You are already mining with excelent hashrate, do you know the temperature of your vram? (memory temperature) while mining?
I have the same card as you, but I didn't touch in the bios and mine are doing 53mhs with 110w

With this tool: More power tool : https://www.igorslab.de/morepowertool-amd-radeon-rx-5700-und-rx-5700-xt-tweaking-and-overclocking-software/

You can decrease the core clock voltage to 725mv and you'll reduce the power, usage it's easy to use and you will only change the parameters that your windows reads of your card, leaving the bios untouched

And about that videos with 61mhs it's only a command you put on the .bat of your miner, but it's not stable, your card will submit almost all invalid shares

Thank you very much for answering. I was going crazy trying to find a way to get 61 Mhs, and in the end it turns out I'm having a good hashrate ...

Temp of my card is 46º with 60% FAN speed

I have tried to lower the core voltage with MSI Aferburner to 750mV, but when doing this the card always returns 0Mhs, and I must restart the computer to return it to normal.

Are you saying that with MorePowerTool I can lower the power without having to flash the BIOS again? Could you tell me where is the value that I should modify?

Many thanks.

No problem
It's better if you watch this video: https://www.youtube.com/watch?v=X77N4sDXP9g
Because you can understand better

And I'll add this: forget MSI Afterburner, you will only use Radeon Software to set your card
With Radeon you can decrease the clock voltage, the clocks of core and vram, fans and power limit
If you decrease the power limit you can save power (do it slowly until you see a hashrate decrease)

And with this software More Power Tool you can "trick" windows and GPU to let you decrease even more the voltage clock of your CPU, because it's limited to a minimum of 750mv, and you will probably need 725mv, reducing the power usage

And yes, with this software you don't need to flash bios again, it's awesome

hmm i tried all these tricks, MPT mpt , copying straps from 1500 down , even tried 1800 on these new 5700 MSI cards i grabbed to play with no luck on anything

any ideas ?

Those temperatures are seen on 5700 line of cards which have the thermal pad problem and even had asus do a revision on theirs to the evo line due to customer complaints.

I had a question for you concerning the memory speed. Does the boost clock memory affect mining in any way? There are cards which can do 2035mhz in their boost but have around the same as the one you listed above for their regular core memory speed.
Now as I was discussing here: https://bitcointalk.org/index.php?topic=5244869
Could you post your response there because I asked someone what manufacturer of cards they were using but did not reply back.
Thanks. :)

Ok, but what are you trying to achieve?
You said you tried these things but what's your objective?






I find that it's normal temperatures
There are no problems with this model, I opened my rig and put the fans on 60% and the temperature now is 85c, completely normal, because the memory is on 95% load
I tested in games and the temperatura is normal too

I'll read your thread

my stock hashrate is 49 max and power use is high around 140 watts in gpuz

tying to get to the 56 or even 53 mhs and 100 watt in gpu z but the straps amd bios mod seem to do nothing

Ah ok now
You can use More Power Tools to decrease the minimum voltage to 725mv on the clock core

Now you go to Wattman (Radeon Settings) and try the following: 1280mhz core clock, 735mv, 1870 memory speed and decrease the power limit to -35

This is the best settings for me, my card is doing 54mhs stable, without any bios mod

Your memory clock is the most important to achieve more hashrate

So bios strap mod is not importan ? you dont need to flash the MPPT settings ?

I prefer not to use bios mod to not lose warranty if I need

You have to understand that with bios mod you are flashing the bios of your card, on hardware level and with MorePowerTool you'll modify the values of the software on Windows, so your Windows will read the values of MorePowerTool, without need to flash the bios, you just need to use the program and restart your computer

I suggest you to read: https://www.igorslab.de/morepowertool-amd-radeon-rx-5700-und-rx-5700-xt-tweaking-and-overclocking-software/

There's plenty of information on Youtube too

In my case, I didn't touch the bios and my card is doing 54Mhs, it's good enough for me, I don't want to take risks to achieve 56Mhs, but it's up to you

Update MSI 5600XT Mech OC just got Bios update to 14gbps. with +100Mhz Ram oc it is hitting 38,6Mh/s.

With some timing edit it would reach 40Mh/s. However there is no way yet to do as i know.

do you have to use Phoeniex to get over  50 mhs ?

I'm using Phoenix miner, but this hashrate can be achieved with Claymore too

Which model of card do you have?
And what's your settings of overclock?

I'll try to help you, because it's usually easy to achieve 50mhs or more on those cards

Oh, man you are awesome. I literally made an account to tell you that. I was struggling with my new build, and going livid this morning about not being up and running and all these "help" sites/videos with settings bluescreening my new build SMH. All I wanted were some simple settings where I could use the Radeon software and get something reasonably decent in hashrate but super stable. I'd do this as a hobby on the side, and there is no way in heck I'm soaking assloads of time into all this for 4 gpu and 10 extra mhs.

Anyway, I started a good bit more conservative than you:

In Wattman I lowered mV to about 800ish, memory up to 1800, power limit like -10ish and seems really stable at 51mh/sec for my RX5700. Good enough for me. Finally, something that is actually stable after I struggled a good bit getting going. Mining software says 120 watt per card. I do have a fairly aggressive fan curve. I notice if the memory gets too hot then the core will drop and your hash rate will drop off. So, I made the fan curver quite a bit steeper. Perhaps, I will get a little more aggressive later with the memory clock as I get comfortable. I should mention these settings are working great for my 3 RX5700s for 8 hours so far at a temperature around 55C. Though, I do have one RX5700 XT that seems nowhere near optimized as it's reporting 49mh/sec with the same settings.

Anyway, thank you for sharing.

Non XT 5700 is not being produced anymore.

So what, Ebay is full of them. What seems to be the problem, do you really need to get them brand new, simply buy them used on Ebay or other used goods sites. Now that the Nvidia 3000 series specs were released, prices for older AMD and Nvidia gpu's should come down in price also...

I need some help. I bought 2 5700 XT's yesterday and I'm trying to figure out why is the hashrate only 22 Mh/s on stock settings, and only around 26 Mh/s when I tweak it. Tried both Claymore and Phoenix both give a similar result. Trying to find information online but can't seem to find anything specific.

Did you set your video driver in compute mode ?

Yeah, even though it isn't needed with this architecture. No difference at all.

Did you plug in both pciex cables in? also is the power supply good enough?

Yeah none of those simple mistakes. I've been mining for several years but only with Nvidia. It's funny how simple setting up Nvidia cards is, simply plug and play. Here with my first AMD cards and here I am...

PSU has plenty of power. It's 1600 watt EVGA P2 Platinum.

my gigabyte 5700 xt was doing 54 out of the box...even undervolted it does well, maybe if there was nvidia driver on the pc try to ddu, also if the miner command line has intensity settings maybe delete them

Done all that DDU stuff multiple times in slightly different ways, still no change... also tried playing with intensity. I'm going to wipe my rig clean and install win10(currently running win7 but I doubt this is the cause of my problem) and try it from scratch. I'm starting to think it has something to do with OpenCL.

Do miners see your cards as 2 or as 4 cards? Recently, miners started to see my RX580 as two openCL cards, both mining at exactly half hashrate,
but also overheating while doing so. No amount of DDU or reinstalling drivers helped

Maybe it's a driver issue, what driver version from AMD are you using when that happened? A lot of people are complaining about poor and buggy drivers from AMD in the last few years. Maybe trying a different driver version would help getting rid of the problem. What do you think?

I just did complete windows and driver reinstall (latest win10 build, latest drivers). Now it works normal, but remains mystery why it happened.
I tried many drivers, even old ones that previously worked, but nothing helped except full reinstall

Yeah just came to say that I did the same I forgot to say that I also fixed my problem also with win 10

You should deduct idle system consuption, which is about 40W usually.

You should deduct system power consumption once the miner is running.

Easy
https://youtu.be/Gn6iFQ7J6Ys
55.5Mh/s 105-110W from the wall (87 by sensor)

Moжнo пoлyчить и мeньшe ))) Haпpимep нa кapтax c caмcyнгoм peaльнo пoлyчить дaжe 85W c poзeтки. Hyжны пpaвильныe кapты и пpaвильныe нacтpoйки.

What driver/kernel versions have decent hashrate using phoenixminer under Ubuntu 18.04 or 20.04?

Rx5700 and RX5700XT are one of the best AMD graphics card for mining nowadays and I still have some respect for AMD VEGA cards as well but the problem with RX5700s are high temp, for longer life these cards aren't the best and also if you are from a hot country Nvidia cards are best for you, I prefer RX480 or RX580 Amd cards over these RX5700s, they are way too hot like 80oC to 90c that's too high.

Nah they are good, more heat output more efficient they are in winter :) Love em when its turning under minus 0 C

https://www.anandtech.com/show/16150/amd-teases-radeon-rx-6000-card-performance-numbers-aiming-for-3080 (https://www.anandtech.com/show/16150/amd-teases-radeon-rx-6000-card-performance-numbers-aiming-for-3080)
new rx6000 series aka big navi teased, perform between rtx 3070&3080, at this point looks the gpu same route as rx5000 series
hope they running wider memory bandwith, fpga gaining more advance in the recent years, compete with gpu in mining, specially in compute performance rather than memory algo

Same reason why I'm against this RX5700XT, I ordered one from Newegg with double fans but they are too hot for my liking, it feels like the card will die very soon, I have no confidence mining with this card around the clock ( full 24hrs/ day ), some said to get the variants with three fans but they are way too overprice, got mine around 337$ but double fam variants are 449$ or more.

Looking at Moore's Law is Dead video, it seems the 6900XT will basically be double the 5700XT. So I would imagine double the processing units. The bus should be wider by 64 bits at the minimum and it will have 16GB of memory, but it probably won't be GDDR6x nor will it be HBM2.

So before tweaks it would probably run 100-110MH/s and with best optimizations after a month or 2 perhaps 120MH/s if pushed on a card with excellent cooling. The revised architecture should help decrease wattage but on a card that "big" you'll still need a massive cooler, so I would guess $800+ for a decent card.

According to his sources 5700XT is currently stopped in production for this big Navi.

That's if the memory bus allows it.

Hi DDesignE

Did you get Ethminer to work with RX 5700?  I tried v0.19.0-alpha miner but the miner shut down.  Please help.

Thanks

Vin

I have an XFX 5700XT Triple Dis, power consuption 130W in gpu-z no matter what i do. If i turn TDP down more than -28, than hasrate drops a lot (with gpu clock) and of course Power consuption as well.

My Sapphire 5700XT do work -40TDP with 50+MH.

Anyone has the same card (with same issue)

solved

I've just signed up to answer this - recently bought a Gigabyte RX 5700 XT OC for a good price in order to mining with it, and I get exactly the same memory temperatures under mining load when running the default fan profile of the card. I can get the temps down to 96 degrees if I raise the fan profile up to 80% (which keeps the core temp (non-junction) at around 58 degrees) but it sounds like a jet engine. Having done a fair amount of research on the subject through various forums and articles, it does seem as though this is standard behaviour for these cards when mining or performing memory intensive workloads, regardless of the model, and that the memory temperature reported is a 'junction temp', not the edge temp. I've noticed that under more normal gaming loads (e.g. 3dmark) the mem temps get up to 80 degrees or so with the standard fan profile so it does seem to be the mining that really pushes it higher.

Having owned a number of other graphics cards I still think these temperatures are too high versus what I'm used to seeing normally but maybe the true junction temperatures in other cards aren't reported in the way that they should be. What rx 5700 xt models are other people mining with? Do you see the same temps on sapphire and powercolour models? From the anecdotal evidence I've seen on other forums, these cards have similar temps in the same scenario.

@screamopilla

Hey there...Just picked up a Gigabyte 5700 XT OC...Currently running 850 1300 (120 watts in Phoenix miner 5.3e software) and the mem at 1820...I got the mem temps down to about 88 on the highest card by forcing it to run at about 45% constant.   Problem I'm having is that I can't get a stable setup over 1820 ish on the mem...Which gives me about 52.6 mh.  I am running a rig that had an nvidia driver in it but did the DDU 2x.  Curious what your clocks are and what miner you're using.  Are you guys on Windows?  

Have you tried the Bios mod?  Appears the links to the 3.04+ flashing software on Igors site are dead...

@rdlduffy try a fixed 45-50% on the fans and see how it looks.

ive got core at 1390 mem at 920 (for windows that would be 1840...im on linux) voltage at 0.750 (may show as 750) mvdd 1325 and mvddci 0.8 (or 800) nets 53.25mh at 104 watts. been stable for days...i have not yet modded the cards bios.  Mem temp is 88 and fans are going  30-50% Asic temp is 58.  wattage can probably reduce a little more say to 90 but for now happily mining away. 

You need to mod the bios - https://www.youtube.com/watch?v=LhT74CBcFvM

Thanks

With my Gigabyte I'm using 70% on fans to keep memories at 92°  :D It's very hot in my country and the only solution I found is to put fans to work hard
My Gigabyte is rev 1 and I think this is the best version, I'm getting 55Mhs right now on Phoenix, stable and without any restart, with 107w on software (no biosmod)

I bought an MSI 5700 Mech (non xt) and I'm trying to tweak, but it's not good as my first gpu, only getting 52Mhs, with 119w and not very stable yet
My gigabyte accepts 715mv but my MSI only 800mv to be stable

Thinking about modify the bios on this one

MSI 5700Mech 730mv =+53mh/s=103W logiciel
XFX 5700        750mv=+53mh/s=  84W logiciel
--no XT


very difficult to understand this GPU.

https://image.noelshack.com/minis/2020/48/7/1606657806-msi.png (https://www.noelshack.com/2020-48-7-1606657806-msi.png)

https://i.ibb.co/6rr3qkC/XFX.png (https://ibb.co/GnnKrqS)

Thanks for that Dowper, is that the method you used?   I've seen other vids of them talking about using 3.04+ to flash from Igor's site, but those links appear dead at the moment.

The mem temps I reported earlier were when running the card all at stock, in a case with not great airflow and mining on Phoenix miner. I've now put the card in my mining case which has better airflow and am running it underclocked w/ power tuning. New clocks are as follows:

gpu core 1247mhz @ 836mv
mem 875/1750 (default) @ stock voltage

power limit -35
fan speed @ 67%
power consumption: 119w
hashrate: 50.5 MH/s

Using the above settings the memory temp stays at 90 degrees under load. I can get it down to 88 degrees by raising the fan speeds above 80% but the noise starts to become too loud. This is the best I can get on the memory temps on this card (gigabyte rx 5700xt OC) when mining. Given that I think these temps are too hot I don't want to raise the memory clocks any higher and I'm happy with hashrate and power consumption as it is. I've just upgraded from an rx480 4gb (which won't support the DAG size at the end of the month) so it's a big step up from that.

As mentioned previously running the card under normal gaming loads the memory doesn't get near these temps with a custom fan curve so obviously mining is taking it's toll heatwise here. The GPU hotspot temps are max 65 degrees so more than fine there. I imagine you'd get better memory load temps with the powercolor red devil or sapphire nitro cards but for the price I paid for this one (£330 brand new) I'm happy with the performance.

Update: The addition of two top mounted 120mm exhaust fans has brought memory temps down by an addition 2 degrees to 88 degrees under mining load (with a slight increase in gpu fan speeds). This was my target temp so I'm happy  ;D

Update 2: Now down to 84-86 degrees at long term load with a lower ambient temperature, this is with 66-71% fan speed. I've been able to bump up the memory to 1800mhz for an extra 2 mh/s with minimal change in temps, so 52mh/s is my new hashrate. Don't feel the need to push it any further for the moment

Hi, im wondering what is the max memory frequency you guys managed to get with a 5700 xt?
Im mining with flashed bios 1445 core 1920 memory. It seems memory clock is maxed out and i manage to get 58.7 Mhs in teamred

I have mine slightly overclocked to 1804 which gives me around 52m/hs with peak 88 degree memory temps. This is on a gigabyte rx5700xt OC (3 fan version). Above that and it gets up to 90 degrees which is too hot for me

1350 core / 920 mem / 775 volt = 53.22mhs on my 5700xt cards :)

Only the Thicc XFX unit got 1400 / 910 / 775. 52.72 mhs.

Power shows 105watts.  Memory temp 80 and under Asic Temp 60 and under.

No bios mods.  MMPOS

 

Yes, you can use 3.04+.

Hi to all.
Hive recently added soc clock and soc Voltage to change.
I remeber few month ago, i had nice russian XML with bios and all settings for 5700.
Can any1 help me find it aggain ??

My 5700 any brand are set 1400/912 after rbe bios mod - gives me ~55Mhash for ~115W - Micron memory - i`m not lucky guy :/

Is anyone measuring their usage at the wall (whole rig)? Mine is by no means a legit mining setup (with rack etc.) but after further tweaking I've managed to get down to 186w load at the wall, with around 113w reported on software. The Gpu fan speed has been reduced to around 60% which keeps memory temps at around 84-86 degrees under load. I was able to get nearly 20w reduction in total consumption by running gpu core voltage of 0.751mv @ 1280mhz (previously had it left on 0.830mv from my first pass at undervolting which took 200w @ wall).

Complete rig component setup is as follows: asrock z170m socket 1151 motherboard running a pentium g4560 dual core cpu and 8gb ddr 2133mhz + 120gb sata ssd. I also have 5 120mm case fans in the case to keep the gpu cool plus a usb wifi dongle. PSU is a Gigabyte Aorus Pro 850W Gold Modular. As previously mentioned I have a gigabyte rx5700xt OC 3 fan gpu. This is running windows 10 but I might try one of the miningOSs to see if that makes any difference

Interested to hear what others are getting in terms of total rig power consumption at the wall, pretty sure I can get this lower with a bit of tweaking. I've read that power consumption on these gigabyte models can be higher than other manufacturer versions, especially if the card has Micron memory. GPU-Z doesn't say which memory I have so I'm not able to tell

sounds like you are close to optimum under/overclock settings.  im using msos(OS) on a 10 card 5700 rig, power limit set @110 watts per card achieves 51.5mh/s.  strongly advise moving to a linux based OS way better to tweak and run, used win10 in the past and what a pita!!

used to run on these manually "fixed" settings/values very well

GPU brand...ASUS ROG Strix Radeon RX 5700 OC Edition(Micron memory)
mining software...LOLMINER
OS...MSOS(minerstat)
core @1300/core mV@770
mem @900mhz
fans set 80-90%  very hot summer weather where i am :(
memory chip temps during the day between 82-86...@night drops around 78-80

as of late i don't even worry to set fixed value tweaks as setting the "power cap/limit" gpu's self adjust core clock+ core mV automatically.  so now the only values i set is power(110w/per card)+memory(900Mhz)

hope this helps

Added a 3060 Ti FE to my mining rig today - works fine with the dual cards on windows 10 using phoenix miner (I'm using the onboard gpu output for windows itself).

Had to reconfig them for best temps (moved rx 5700xt from main slot to second pcie slot, and rtx 3060 ti into main pcie) but after tweaking the 3060 ti FE I'm now doing 108mh/s for 318W at the wall and its only added 2 degrees on to the rx5700xt memory temps are tweaking fan profiles a bit.

Very happy  ;D

software reported watts are meaningless for 5700xt, some brands report 100w, some brands report 135w, watts at wall = same.  If you are running at .75v and 1280mhz realistically there is improvement to be made but that improvement will not be 20w.  So your gpu seems to be around 150-160w which is good.  i know some people claim (dont trust verify) much better but consider you have a gold psu probably on 120v and that's that.  you are good.

Just to share my numbers:

XFX RX 5700 XT THICC III

out of the box: 49.x MHs

settings to reduce power consumption and a bit of OC (no bios mod)

Core Clock, Mhz = 1400
Core Voltage, mV = 790
Memory Controller Voltage, mV = 790
Memory Clock, Mhz = 1800
Memory Voltage, mV = 1350
Fan, % = 65%

Results:
51.17 MHs
Core 44 Celsius degrees
Mem 62 Celsius degrees
Power consumption 112 watts

hi

please guide me

is  power supply 550 is enough for two 5700 gpu  ? and which type power supply is best and have many connectors for add more gpu  for later

thank you

what is the minimum motherboard needed? Is it ok an old pentium dual core on PCIe 2.0?

Yes, no problems with old motherboards.

Is power consumption measured at the wall or software? If software you need to add an extra 30 or so watts being drawn from the pci-e slot that isn't reported. I have my core voltage all the way down to 725mv on my rx 5700xt (gigabyte OC) via MorePowerTool but total system power consumption at the wall is around 180w running phoenixminer 5.5c. I seem to have hit a plateau in terms of power consumption using the AMD radeon tuning settings there (it's a single miner rig, with not the most efficient motherboard and cpu pairing (ddr3, haswell i3)). Around 35-40w of that is the base system and case fans.

I have another single gpu rig with a single rtx 3060 ti (again lots of case fans but this time with ddr4) and that does around 168w at the wall. I am working on making both more efficient but I can't adjust the memory voltage down on the rx5700xt because it's running on windows at the moment. Apparently you can access this setting on HiveOS so I'm going to create a dual boot and see if I can get the total power consumption down further. Annoyingly you have to mine on HiveOS mining pool or you have to pay $3 per month (using Ethermine with my current setup) but I will give it a go

In the spirit of trying to keep this thread alive and anyone interested in efficiencies...I swapped out the psu from my rx 5700xt single gpu miner rig y'day (built entirely out of spare/second hand parts) and have saved at least 12.5w usage at the wall in the process.

Previously I was running an old corsair hx620w psu (just retired from my main rig after about 10 years of use) and having read up on it it wasn't the most efficient psu, particularly at partial load and given it's age. But it was what I had spare.

I replaced it with a silverstone strider 650w platinum unit from another rig and power consumption at the wall is now at 172w with the tweaked rx5700xt. This was versus the 182-185w I was getting with the old corsair unit. For reference the total system spec is as follows:

-Intel Core i3 4130 socket 1150 (2core 4 thread, Haswell)
-2x4gb ddr3 1600mhz (mixed crucial & hynix sticks), running at 1333mhz
-Gigabyte Socket 1150 GA-H81M-DS2V mATX motherboard
- 1x200mm Thermaltake case fan
- 2x140mm Arctic F14 case fans
- Silverstone Strider ST65F 650w Platinum Modular PSU
- Thermaltake Core V21 mATX case
- Gigabyte Gaming RX 5700XT OC 8GB

I've run separate tests without the gpu in the system and was getting 45-50w idle with the old psu which was quite high. I haven't re-run the test with the new psu yet but I'm guess it will now be around 40-45w for the system now with the more efficient psu, so this means the rx 5700xt is pulling around 130w total at the wall which is pretty decent. AMD Software reports 102-103w (which excludes power draw from the pci-e slot, as I understand it).

Next thing to try is a dual boot with HiveOS installed so I can tweak the memory voltage and see if I have a Samsung or Micron card (GPU-Z wouldn't show me). I also have a cheap Pentium G3220T dual core cpu on the way to see if the lower TDP rating helps with consumption any. Given that I have the i3 4130 downclocked already it may be negligible

New hashrate settings for my Gigabyte Gaming RX 5700XT OC 8GB card, for anyone who has one and may be interested. This is using windows 10 64 (onboard graphics), the latest WHQL AMD gpu drivers and all tweaks made through the AMD software (tuning section):

Hashrate: 54mh/s

Core clock: 1287mhz (gpu-z shows 1251 actual)
Core voltage: 725mv (limit lowered via MorePowerTools)
Vram clock speed: 1872mhz (gpu-z shows 1868mhz actual)
Vram voltage: default (cannot change this in AMD software)

Custom fan curve: 60-72%, depending on gpu temp

Gpu temp: 43 degrees
Gpu temp (hot spot): 51 degrees
memory temp: 80-86 degrees (depending on ambient temps)

power consumption, software: 107w
power consumption, wall (total system)*: 178w

*Spec for total system is show in the post above (approx 40w overheads).

Notes: with a wall meter installed on the plug, the above settings are unstable (crashes/hangs during auto-tuning). Without the wall meter installed these settings are stable

while rx 5700 has similiar hashrate with vega series, but more efficient nice on bear market, at this point we need cheap gpus with big number, and rx 6700 similiar bandwith with rtx 3060 only 10% faster, expected 10% faster compared rx5700 which is no interesting in expensive value at 479$ while fancy infinite memory dont utilize well in current miner