[JCE]Fast & stable CN/v8/Heavy/Tube/XHV miner, CPU+GPU, Vega56 1800+ RX580 1200+

[dingdongtobias]

what is gut conf for vega56/64 ?

"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 16, "alpha" : 64, "beta" : 16, "index" : 0, "multi_hash":1792 },
     { "mode" : "GPU", "worksize" : 16, "alpha" : 64, "beta" : 16, "index" : 0, "multi_hash":1792 }
]

only 1720hs, teamred give 1980hs

[JCE-Miner]

Is it on CN-v8?

Try:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 16, "index" : ..., "multi_hash":1920 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 16, "index" : ..., "multi_hash":1920 },

or, alternative

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : ..., "multi_hash":1936 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : ..., "multi_hash":1936 },

TeamRed is a very good Miner that just came out, specialized into CN-v8 on Vega, so i'm not jealous if you find them to be faster. We all were impressed by their performances. Just keep in mind they have 2.5% fees, i've 0.9%.
Well configured, JCE should be almost on par with TeamRed, maybe 1% or 2% slower.

[dingdongtobias]

Is it on CN-v8?

Try:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 16, "index" : ..., "multi_hash":1920 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 16, "index" : ..., "multi_hash":1920 },

or, alternative

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : ..., "multi_hash":1936 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : ..., "multi_hash":1936 },

TeamRed is a very good Miner that just came out, specialized into CN-v8 on Vega, so i'm not jealous if you find them to be faster. We all were impressed by their performances. Just keep in mind they have 2.5% fees, i've 0.9%.
Well configured, JCE should be almost on par with TeamRed, maybe 1% or 2% slower.

yes sorry v8

i try your configs and report thanks

[mk111]

Thanks for all the info everything has been giving!
Latest version of JCE.

4 x Vega 56 Reference with V64 BIOS
1475 / 915 and 1100 /900

Getting 1650h/s per card

Current Config:

     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta":4, "index" : 1, "multi_hash":896 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta":4, "index" : 1, "multi_hash":896 },


Driver :24.20.13011.1009

Tried various configs, cannot get more than 1650h/s.
Any advice?

And is it normal for JCE not to show temp and fan speed?

Thanks

[MadSilence]

OMG JCE MAN!

Just tested your miner on two 7870 cards. and im getting what I got before with claymore and even better! 400 per card!! NICE!!!!!!!!!!!!!

Now I really need a GPU linux version and watchdog temp asap:D Then your miner will be on par or better than Claymore even!! NICE JOB! =)

[rudefyet]

How does the autoswitch work? Using moneroocean.

When I download their fork of xmrig, it does a quick benchmark and saves my hashrates for each algo, then goes from there. How does JCE know which algo is most profitable on my hardware? Does it save hashrates over time?

I gave it a quick shot and it chose cnv8 which is less profitable on a dual core i5 than cryptolight.

[JCE-Miner]

i read your reports about hashrate stability, i'll focus on stabilize Heavy hashrate as the top priority

Hi. Miner not work with format "user wallet + paymentID.diff". Can you fix it? Please)

What pool did you use?
I'll do a test on OptimusBlue, they support this syntax

OMG JCE MAN!

Just tested your miner on two 7870 cards. and im getting what I got before with claymore and even better! 400 per card!! NICE!!!!!!!!!!!!!

Now I really need a GPU linux version and watchdog temp asap:D Then your miner will be on par or better than Claymore even!! NICE JOB! =)

Linux: very unlikely, because of lack of time. I dev alone and have all the CPU versions to update too. Cannot afford Linux GPU port from the Adrenalin.
Temperature watchdog: added, now testing, for 0.33b6
syntax will be: --max-temp M with integer M>0, in °C

How does the autoswitch work? Using moneroocean.

When I download their fork of xmrig, it does a quick benchmark and saves my hashrates for each algo, then goes from there. How does JCE know which algo is most profitable on my hardware? Does it save hashrates over time?

I gave it a quick shot and it chose cnv8 which is less profitable on a dual core i5 than cryptolight.

It exists only for the CPU version for now. Then, it just connects and let MoneroOcean switch the algo on the fly. The miner itself takes no decision. MoneroOcean tend to mine mostly on CN-v8, at least when i tested in end of october. When mining such way, with --variation 16, to get more logs about what algo is mined.

example:

Code:

22:08:37 | Connected to pool. Now logging in...
22:08:37 | Algo-autoswitch Pool gets proposed:
22:08:37 | * Original Cryptonight
22:08:37 | * Original Cryptolight
22:08:37 | * Cryptonight V7 fork of April-2018
22:08:37 | * Cryptolight V7 fork of April-2018
22:08:37 | * Cryptonight-XTL fork of May-2018
22:08:37 | * Cryptonight-Alloy fork of May-2018
22:08:37 | * Cryptonight-MKT/B2N fork of May-2018
22:08:37 | * Cryptonight-Arto fork of May-2018
22:08:37 | * Cryptonight-Fast MSR fork of June-2018
22:08:37 | * Cryptolight-Red fork of July-2018
22:08:37 | * Cryptonight V8 fork of Oct-2018
22:08:38 | Successfuly logged as .......
22:08:38 | Auto-select algo switched to: Cryptonight V8 fork of Oct-2018
22:08:38 | Pool changes Difficulty to 100.
22:08:38 | CPU Thread 2 for Cryptonight/forks wakes-up.
22:08:38 | CPU Thread 1 for Cryptonight/forks wakes-up.

Auto-select algo switched to: Cryptonight V8 fork of Oct-2018

Here that's MoneroOcean that chose the algo, among those listed above "gets proposed"

[Pennywis3]

How's the warm up for heavy on 0.33b6? Did u manage to improve it?  

[JCE-Miner]

still testing...

[vmozara]

JCE, I have vega rigs only and was thinking to move all my rigs from cn v8 (TMR) to heavy algos to see if i can reduce power consumption a bit for similar profit. Can I go with current version of miner, or wait for b6? You plan to introduce some changes for Vegas or I can start changing slowly?

[Lermite]

JCE, I have vega rigs only and was thinking to move all my rigs from cn v8 (TMR) to heavy algos to see if i can reduce power consumption a bit for similar profit. Can I go with current version of miner, or wait for b6? You plan to introduce some changes for Vegas or I can start changing slowly?

The b5 brings a big improvement on CH Heavy but its hashrates are unstable.
That's why you should wait for the coming b6.

[JCE-Miner]

Right, better wait a bit for b6 if switchingall your rigs is not trivial.

[vmozara]

I just tried b5 on one of my rigs and got this:

Abort was called at 767 line in file:
D:\qb\workspace\19992\src\vpg-compute-neo\runtime\os_interface\windows\wddm.cpp

i used settings that worked at my home pc with vega 64.



Anybody knows what is this?

I tried these:
 { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 8, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 0, "multi_hash":960 },
 { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 8, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 0, "multi_hash":960 },

then tried mutlihash 944 but same

[JCE-Miner]

hello!

Documentation:
https://github.com/jceminer/cn_gpu_miner#troubleshooting

Last question:

Q. I get error D:\qb\workspace\19992\src\vpg-compute-neo\runtime\os_interface\windows\wddm.cpp
Intel GPUs are now enabled among AMD GPUs, ensure you're not applying a previous AMD configuration to an Intel GPU

I'm about to release a minor update with more stable hashrates for Heavy (I hope...), no more autoconfig on Intel GPU (causes a lot of troubles) and smoother autoconfig on AMD ones. And a little optim for Monero v8.

Question asked just above.
The 0.33b6 and above will ignore the Intel GPU in autoconfig (but not in manual config)
Some Intel chip has probably shifted your GPU indexes

[JCE-Miner]

Hi. Miner not work with format "user wallet + paymentID.diff". Can you fix it? Please)

Code:

23:11:51 | Connecting to mining pool mox.optimusblue.com:3333 ...
23:11:51 | Connected to pool. Now logging in...
23:11:51 | Successfuly logged as Xwmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm.26000+9999999999999
23:11:51 | Pool changes Difficulty to 10000.

Just test and works fine. Better give me your full command and log.
Wallet and pay-id anonymized of course, but tested with a real one.

[vmozara]

Question asked just above.
The 0.33b6 and above will ignore the Intel GPU in autoconfig (but not in manual config)
Some Intel chip has probably shifted your GPU indexes

Sorry, I was not following the topic (lots of work )

I shifted the index and tested with 1400mhz core 1100hbm on mixed vega 56-64 rig with custom 880mv power tables. Hashed 11300 with super small consumption 1050W.
But hashes were all over the place, with some cards hashing 1800 and some 1500.

I will wait for new version 

[JCE-Miner]

Thanks for report, i think there's a lot of potential, and that current version, even with unstable hashrate, has the current lead (I expect TeamRed to provide a good one soon, of course).
Currently testing on my most unstable rig (with some crappy RXs) and b6 looks more stable, and also a lot faster than b4 whatever.

The next version after b6 will probably be a CPU version with some extra features like Cryptolight-Dark and more MoneroOcean config.

[bucah]

I just tried b5 on one of my rigs and got this:

Abort was called at 767 line in file:
D:\qb\workspace\19992\src\vpg-compute-neo\runtime\os_interface\windows\wddm.cpp

i used settings that worked at my home pc with vega 64.



Anybody knows what is this?

I tried these:
 { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 8, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 0, "multi_hash":960 },
 { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 8, "gamma" : 8, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 0, "multi_hash":960 },

then tried mutlihash 944 but same

Try 896

[UnclWish]

Please, share optimal multi_hash parameters for RX 580 8Gb for heavy algo.

[Aerocool]

Hi. Miner not work with format "user wallet + paymentID.diff". Can you fix it? Please)

Code:

23:11:51 | Connecting to mining pool mox.optimusblue.com:3333 ...
23:11:51 | Connected to pool. Now logging in...
23:11:51 | Successfuly logged as Xwmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm.26000+9999999999999
23:11:51 | Pool changes Difficulty to 10000.

Just test and works fine. Better give me your full command and log.
Wallet and pay-id anonymized of course, but tested with a real one.

Sorry it's my mistake in syntaxis .bat file . Just started test CN Heavy at RX580.