Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[rednoW]

Mister Claymore.
Question about Vega eth hashrate.

We already achieved impressive 2000h/s Monero speed using HBCC with blockchain drivers and 2 high intensity threads with xmr-stak-amd.
PM me for details.

Is it possible that with complete miner rewrite, maybe with 2 DAG files in one GPU we can get significant boost with ETH speed?
Or maybe this miner already exists and is used with some private farms?

1. Do you see improvements for enabled HBCC in multi-GPU system? If you use single GPU it may help a bit in some cases, but I doubt that it will help if you have six GPUs.
2. I'm not aware of any faster private ETH miners.

I have no experience with 6 Vega rigs but it certainly works for 3 gpu-s.

Have you tried to start XMR CPU miner in parallel? Perhaps HBCC just use bandwidth of main system memory. Just a guess, I did not check it yet. Though CPU miner mainly uses L3 CPU cache, not main memory, but soon I will check it anyway.
I have finished asm XMR kernels for all GPUs except Vega, I will check Vega in a day.

My system has poor Celeron G3900 cpu and cheapest ddr4 memory running low.
It cannot be boost from system memory.
I guess when you enable HBCC you are enabling some alternative paging scheme allowing effective usage of all GPU memory.

[1714847678]

1714847678

[NavyCuda]

While I haven't touched the rig myself, it's been reported to me that HBCC segmentation can cause crashing on 1x to 16x powered risers.

Myself, both of my vega cards are running 8x in motherboard slots.  Without HBCC segmentation I get 36Mh/s, with I get 42Mh/s.  This is at HBM2 clock of 1100MHz.

[Unclegogi]

Claymore,
I did not check ETH only mode, but in ETH+DCR mode it is easy to reproduce:
1) start mining (no clocks settings in bat file in my case)
2) wait until miner looses connection to pool (you can force it by unpluggung lan cable)
3) wait until miner connects to the pool again (plug in the lan cable)
4) in most cases the graphic card will work in P0/P1/P2 state with low clocks until you restart the miner

I have copyed parts of the log to this post, maybe it will help?
https://bitcointalk.org/index.php?topic=1433925.msg21431939#msg21431939

Then you should be able to see this issue if you press "0" key to disable GPU0 and then press it again two times to enable it. When I do it, everything works fine, I see P0 clocks when GPU is disabled and P7 clocks when it is enabled.

Pressing "0" doesn't force the graphic card to stuck on low P state.
But reconnecting to pool does...
I have reproduced this bug again:
https://image.prntscr.com/image/ODk4rUfXTlCwyS0txlLT5g.jpg (screenshot #1) - internet connection to pool was interrupted, miner reconnected to pool and RX470 is stuck at low P state (466/2000MHz), you can see that pressing "0" doesn't fix this problem, GPU0 is still stuck at 466MHz P state.
https://image.prntscr.com/image/4vZSrEESQU_5WWdpXXz-xg.jpg (screenshot #2) - after screenshot#1 was made, I restart the miner, and GPU0 now works on it's normal clocks (1125/2000MHz).

I can send you log file from miner from that time, when screenshot #1 was made.

[Vlado]

Hi guys, have you ever seen such thing? GPU 0 - 80+MH???

ASUS RX480 4G with custom bios. All others GPUs have the same bios.
Core 1100, mem 1880, 890 mV.

It has been going like this for almost 5 hours.  I noticed that it was doing 27.7 and after a new job the hashrate jumps to 81 MH maybe 5 mins after miner started. Temperature is stable. Pool also gives a total of 222-224 MH for the miner. I will keep it running like this unless it crashes and will pray to have the same bug on more GPUs  

https://imgur.com/a/4M1Ox

EDIT: I checked the shares - GPU0 got only one. So this is some bug that increases speed, but does not increase actual processing of shares. Looks like the GPU revs on neutral.

https://imgur.com/a/x0TMZ

[rednoW]

Mister Claymore.
Question about Vega eth hashrate.

We already achieved impressive 2000h/s Monero speed using HBCC with blockchain drivers and 2 high intensity threads with xmr-stak-amd.
PM me for details.

Is it possible that with complete miner rewrite, maybe with 2 DAG files in one GPU we can get significant boost with ETH speed?
Or maybe this miner already exists and is used with some private farms?

1. Do you see improvements for enabled HBCC in multi-GPU system? If you use single GPU it may help a bit in some cases, but I doubt that it will help if you have six GPUs.
2. I'm not aware of any faster private ETH miners.

I have no experience with 6 Vega rigs but it certainly works for 3 gpu-s.

Have you tried to start XMR CPU miner in parallel? Perhaps HBCC just uses bandwidth of main system memory. Just a guess, I did not check it yet. Though CPU miner mainly uses L3 CPU cache, not main memory, but soon I will check it anyway.
I have finished asm XMR kernels for all GPUs except Vega, I will check Vega in a day.

XMR is good but this is ETH thread ))). Can you try to run 2 gpu threads (with 2 DAG files) with 1 Vega? Current version shows 0 hashrate if I run 2 instances.

[xeridea]

Mister Claymore.
Question about Vega eth hashrate.

We already achieved impressive 2000h/s Monero speed using HBCC with blockchain drivers and 2 high intensity threads with xmr-stak-amd.
PM me for details.

Is it possible that with complete miner rewrite, maybe with 2 DAG files in one GPU we can get significant boost with ETH speed?
Or maybe this miner already exists and is used with some private farms?

1. Do you see improvements for enabled HBCC in multi-GPU system? If you use single GPU it may help a bit in some cases, but I doubt that it will help if you have six GPUs.
2. I'm not aware of any faster private ETH miners.

I have no experience with 6 Vega rigs but it certainly works for 3 gpu-s.

Have you tried to start XMR CPU miner in parallel? Perhaps HBCC just use bandwidth of main system memory. Just a guess, I did not check it yet. Though CPU miner mainly uses L3 CPU cache, not main memory, but soon I will check it anyway.
I have finished asm XMR kernels for all GPUs except Vega, I will check Vega in a day.

My system has poor Celeron G3900 cpu and cheapest ddr4 memory running low.
It cannot be boost from system memory.
I guess when you enable HBCC you are enabling some alternative paging scheme allowing effective usage of all GPU memory.

I get extreme instability and bugs with blockchain drivers.  Does this work on 17.9.1 drivers?  Is this an unreleased version of miner? I got ~1050H/s with blockchain driver, 950 mem (Vega 56).

[smychok.smychok]

Hello all

I cannot use key during mining
The letter is displayed in the console and nothing else happens.

I use Windows 10 pro

What I do wrong ?

Thanx

Same problem, but Windows 10 Home

[ingyenfrag]

Hi!

I've experienced a weird bug on my new rig (6 * vega 56) . Everything works fine when I use 5 cards, but when I add the last one it boots, but 1 gpu always fails at the start of the mining and it freezes in seconds.
I've tried all the cards before one by one and they are perfect. Risers are fine as well. The weird thing is that when I try to start mining only 1 card fails, and it's not the same, it seems its a random stuff which one fails.

Config: Asrock h81 btc r2.0, g1840, 4gb 1600Mhz CSX, 2*750W EVGA GQ, WIN 10 64 pro, [Suspicious link removed]pute driver 2017-08-23, PSU1--> system, 3 cards and their risers, PSU2--> 3 cards and their riser.

I've tried to run 3 cards and the system on only 1 PSU (for days) and it worked perfectly --> ~620 W from wall. PSU shouldn't be the bottleneck.
I've tried to reinstall driver. I've tried to lower the power targets of all card by maximum 50%. I've tried stock settings. I've tried stock settings with 100% fan speed.
As i mentioned only 1 card fails per each start, but it looks random, it's not even connected to the same PSU.

5 cards work perfectly, 39Mh/s/card in solo mine, consuming 1010W from the wall.
Is it a software or hardware bug? I can't even guess, and I'm out of ideas what to try.

Any ideas?

Thank you!

[xxcsu]

updated all of my mining rigs to the latest Win10-64Bit-Crimson-ReLive-Beta-Blockchain-Workloads-Aug11 driver
updated the mining software on all of my rigs to Claymores 10

- rigs are much more stable under dual mining ( ETH/SIA )
- slightly better hashrate on ETH side with same setting from previous Claymore's Software

[PPB]

I tried to reproduce the issue with clocks and low speed until miner restart.
Windows 10, latest blockchain drivers, stock RX580 Nitro+.
1. No overclock: If I start miner in ETH-only mode, it shows about 16MH/s only (and low memory controller load in GPU-Z) and I see this speed for about 20 seconds, then it shows 24.6MH and works fine. For dual mode it shows full speed immediately. Unfortunately, I have no idea why GPU works slowly within first 20 seconds, it is something related to drivers.
2. I close the miner and apply some changes in clocks: memory 2000 -> 2100, core 1300 -> 1100. When I start miner, I see that it applies new clocks in about 10 seconds, but still shows 18MH/s only for first 20 seconds. Then it shows about 25.5MH.
Note that I use default ROM so these speeds are not high.
3. I close miner and apply new clocks again: -cclock 1000, -mclock 2050. Same behavior: it shows low speed at start, then normal speed for these clocks. GPU-Z shows changes in clocks in about 10 seconds after start.

So right now I cannot reproduce the issue with new clocks: in my tests new clocks are applied immediately, I don't need to restart the miner or reboot the system.

I have an update regarding this issue:

- First I told you it happened on my 480 ASUS DUAL 4GB that was having an OCL thread hang, then at restart the GPU would stay in a lower performance mode, albeit having the clocks applied at miner start.

The problem is that this issue with non-rounded milivolt values with the miner makes the GPU have an OCL thread hang AND reset it's clocks. Then with the clocks reset when you open the miner back after the restart changing the default clocks/voltages makes the GPU enter in this lower performance mode.

With this ASUS DUAL 480 4GB, 1050core 1755mem and custom strap it hashes 25,35mh/s normally, 21,9mh/s on this lower performance mode. Restarting the PC makes it go back to the higher hashing value.

[johndoe2]

I'm selling my Gtx 1060s because they are not very good at mining. Whether it was the Hynix memory, or the bad components used, I only got  17Mh/s out of each card after OCing without touching the voltage. I cut my fingers twice with the fan blades, and whenever the network difficulty changes some of the cards hashrates drops to 1Mhs. I'm getting 22 Mhs from 570 8Gb and want to know how to increase hashrate other than increasing the memory clock, thanks.

Just get out now while you still can. Those GPU's going to start taking toes.  

Stop takin a P00P at him

To the OP : Look up Bios modding, your 580 8GB should be able to get 29mh easy peasy at low overclocks.

His "570 8GB" should hit 29.5 with alt ethash coins or 29 if he is using windows 10+ Claymore 10 + Blockchain drivers. Also dual mining LBR he should get about 65 mh's with LBR or about 730 with DCR
All he has to do is A. Up the memory clock B. Run Polaris 1.64 and click the Memory Timing button to load the specific memory timing for the card. Nice new feature of the Polaris fork" and he is good to go, Also I would put the following in the Start.bat for his 570   -cclock 1175 -mclock 2050 -cvddc 900 mvddc 900 -dcri 20 (with LBR)   -dcri 25 (with Decred)

Thanks, will try this. I'm using MSI Armor 4Gb and 8Gb version of the 570, and it kept crashing when I synched settings on all cards, so I'll try it with different settings on AB.  I'll take stability over hashrate anyday of the week.

[Yohow]

-dcri 25 (with Decred)
[/quote]

Which is the best "-dcri" value for a RX 480 8GB rig? I'm trying with 21 since 25 it wasn't stable. Claymore automatically lower the intensity of Decred mining, it could be that one GPU reaches max temp value?

[VladimirT]

-dcri 25 (with Decred)

Which is the best "-dcri" value for a RX 480 8GB rig? I'm trying with 21 since 25 it wasn't stable. Claymore automatically lower the intensity of Decred mining, it could be that one GPU reaches max temp value?

Best for me -dcri 30, but rx580. ETH ~29.5 / DCR ~890 . My temp's ~ 65 for MSI Armor , 55-55 for Gaming X , and more then 70~ for Asus Dual . Asus poor cards, i will never buy them again.
I do not use them for dual - only solo.

[se32786]

I'm probably the only person here who uses Ethman, but does anyone else have an issue where it constantly closes itself? Very frustrating.

[golden0707]

To Mr. Claymore.

Just a wish . Im monitoring my rigs through corp. FW. Because of that i have to use only common ports on -mport option, like 80,443.
Third rig on same IP works on 8008 port. If i user browser - it show fine. If i use your Miner Manager ( ver. 3.1 for examp.) it cannot watch that rig.

I suppose it is becouse of proxy settings for browser.

Could you add an option to use proxy in Miner Manager's next release ?

Thanks.

[l30z1n]

I know it's kind of off topic. Does anyone know this store? It's trustable?

https://boston-electronics.us/

[okiagr]

I tried to reproduce the issue with clocks and low speed until miner restart.
Windows 10, latest blockchain drivers, stock RX580 Nitro+.
1. No overclock: If I start miner in ETH-only mode, it shows about 16MH/s only (and low memory controller load in GPU-Z) and I see this speed for about 20 seconds, then it shows 24.6MH and works fine. For dual mode it shows full speed immediately. Unfortunately, I have no idea why GPU works slowly within first 20 seconds, it is something related to drivers.
2. I close the miner and apply some changes in clocks: memory 2000 -> 2100, core 1300 -> 1100. When I start miner, I see that it applies new clocks in about 10 seconds, but still shows 18MH/s only for first 20 seconds. Then it shows about 25.5MH.
Note that I use default ROM so these speeds are not high.
3. I close miner and apply new clocks again: -cclock 1000, -mclock 2050. Same behavior: it shows low speed at start, then normal speed for these clocks. GPU-Z shows changes in clocks in about 10 seconds after start.

So right now I cannot reproduce the issue with new clocks: in my tests new clocks are applied immediately, I don't need to restart the miner or reboot the system.

Same here, Gygabyte RX580 custom, w10, aug 11 drivers, ETH only. Everything stable. What I also noticed is that if I reboot, after the initial low hash seconds, sometimes 1 min or so, it recovers to the normal hash but the voltage at the wall is 50W lower when normal speed is restored. Then, after another minute or two, the power consumption goes up again with the same normal high hash speed.

[mrtrust]

What is the hashrate of NVIDIA Quadro M5000  8GB?

[se32786]

What is the hashrate of NVIDIA Quadro M5000  8GB?

2

[bomberb17]

I have 4x Rx470 running Claymore v.9.0, only for Ethereum mining.
I have observed that their hashrate is slowly decreasing over time.
I remember back 3 months ago they were at 28Mh/s each, now they are at 25Mh/s each (without changing anything on my configuration of course).
Here's how I run Claymore:

Code:

./ethdcrminer64 -epool eu2.ethermine.org:4444 -ewal <eth address> -epsw x -mode 1 -tt 68 -allpools 1

Any thoughts?