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

[syncro2017]

is there a better config for a RX580 8GB? With cast xmr i can get pretty much stable 1kH/s.

"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 1, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 1, "multi_hash":944 },
]


EDIT: Algo is cryptonight heavy. Also the hashrate with JCE is much more stable. nice work!

For RX580 8GB the best settings i found to be below:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 0, "multi_hash":1200 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 0, "multi_hash":1200 },

[s0ftcorn]

is there a better config for a RX580 8GB? With cast xmr i can get pretty much stable 1kH/s.

"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 1, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 1, "multi_hash":944 },
]


EDIT: Algo is cryptonight heavy. Also the hashrate with JCE is much more stable. nice work!

For RX580 8GB the best settings i found to be below:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 0, "multi_hash":1200 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 128, "beta" : 16, "gamma" : 8, "delta" : 8, "epsilon" : 8, "zeta" : 8, "index" : 0, "multi_hash":1200 },

I get

Code:

Starting GPU Mining thread 0, on GPU 0
Created OpenCL Context for GPU 0 at 0000019eeeb67980
Created OpenCL Thread 0 Command-Queue for GPU 0 at 0000019efc3225f0
Allocating big 4800MB scratchpad for OpenCL Thread 0...
Scratchpad Allocation success for OpenCL Thread 0
Compiling kernels of OpenCL Thread 0, it will be long...
Compilation of OpenCL kernels failed.
Error: CL_BUILD_PROGRAM_FAILURE

When i try to run this for both threads.

[syncro2017]

I get

Code:

Starting GPU Mining thread 0, on GPU 0
Created OpenCL Context for GPU 0 at 0000019eeeb67980
Created OpenCL Thread 0 Command-Queue for GPU 0 at 0000019efc3225f0
Allocating big 4800MB scratchpad for OpenCL Thread 0...
Scratchpad Allocation success for OpenCL Thread 0
Compiling kernels of OpenCL Thread 0, it will be long...
Compilation of OpenCL kernels failed.
Error: CL_BUILD_PROGRAM_FAILURE

When i try to run this for both threads.
[/quote]

Try to lower "multi_hash":1200 in 16 decrements, so first try "multi_hash":1184

[s0ftcorn]

Try to lower "multi_hash":1200 in 16 decrements, so first try "multi_hash":1184

I rebooted and even with "multi_hash":944 which was stable it doesnt compile.

[UnclWish]

Max multihash for heavy algo for 8Gb cards is about 960. With it's near 7.5Gb video memory usage. You both talking about different algo's. One about heavy, other about normal v7.

For heavy algo optimal vaules for 580 8Gb:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":944 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":944 }, 

Also good this:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":896 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":896 }, 

For normal v7:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1152 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1152 }, 

[UnclWish]

I really cannot use a multiple smaller than 16, it would invalidate all my optims. I did a lot of data grouping by chuncks of 16, because that's 4x4. Going down to 2x2 would be impracticable.
About cached compile, that's possible, but more complicated than what Stak does.

And please, switch to the SRB topic for the SRB tips, it's not a rant, no problem to compare with SRB, even for me, but it's out of place and makes the topic harder to follow.

Yes, I got it.
But think, how make more fine tuning...

[JCE-Miner]

Try to lower "multi_hash":1200 in 16 decrements, so first try "multi_hash":1184

I rebooted and even with "multi_hash":944 which was stable it doesnt compile.

i think i found the bug : the miner doesn't supported to be launched from a different dir. said differently, its current dir must be the path where the .exe is

if you call the .bat or exe from a different dir, it fails. i'll fix this in next revision.

[gvb]

I added to my start a directory change to where the miner is.

cd c:\miners\Monero\gpu

then I can use right click start as admin aswell



my GPU mining went fine for hours now it's back to 0H/s even after a quit miner, disable card, enable card, start miner.


any idea what could be the cause?

The driver seemed to crash sometimes but it kept mining (no, it's not my display driver aswell)

[vmozara]

JCE,
You said you have no Vega to test, if you like, I can set up one rig with 1 or 2 Vegas with standard mining cofiguration and share the teamviewer with you if you like to test or whatever configuration is easy for you for testing (windows 10) as long as you don't touch the network or can influence my other rigs

I tried to test this on my Vega rigs but have no idea how to set it up Removed the --auto and tried some settings but then gave up

Now I am back to STAK, because CAST and SRB both proved for me to report mucch higher hashrate than the pool, sometimes it can be up to 20% (no i am not talking about instant hash power but calculated over several weeks)

7x Vega rig on STAK has 13500 and more than 14000 on SRB and CAST but pool reports 13500 for STAK and sometimes as low as 11000 for SRB and CAST

So I have high hopes for your miner and will happily switch all my rigs if the thing is ok

[whotheff]

Tested on my APU - did not work.

Is it going to support Nvidia Cuda in future?

[UnclWish]

Ohhh. For my RX 580 8Gb finded new best intensity for heavy algo:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 }, 

Also finded good parameters for cn-v7 algo for 8Gb cards:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 }, 

Also miner often gives rejected shares "Job not found"...

[syncro2017]

Ohhh. For my RX 580 8Gb finded new best intensity for heavy algo:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 }, 

Also finded good parameters for cn-v7 algo for 8Gb cards:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 }, 

Also miner often gives rejected shares "Job not found"...

Nice settings, thanks for the tip!

[syncro2017]

Ohhh. For my RX 580 8Gb finded new best intensity for heavy algo:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":832 }, 

Also finded good parameters for cn-v7 algo for 8Gb cards:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 },
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash":1696 }, 

Also miner often gives rejected shares "Job not found"...

Just to add more data, for RX570/580 4GB cards, the best settings so far for me on cn-V7 are:

Code:

{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash": 944},
{ "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "gamma" : 4, "delta" : 4, "epsilon" : 4, "zeta" : 4, "index" : 1, "multi_hash": 944}, 

[UnclWish]

Author, too often miner finds rejected shares "Job not found" on heavy and v7, ssl and nossl... nicehash...

[Bazzaar]

Really well done with this gpu version, such good results with a "beta" release.
I have 3 Vega56 and a Vega64. On MoneroV7 I get ~1900 h/s with the 56's and ~2000 on the 64.
No bad shares or errors.
One thought, with cpu and gpu combined isnt the difficulty going to end up way to high for the cpu to be effective?

Baz

[bragmining]

Is there a version of this for Nvidia 1080Tis?

[hammuh]

JCE,
You said you have no Vega to test, if you like, I can set up one rig with 1 or 2 Vegas with standard mining cofiguration and share the teamviewer with you if you like to test or whatever configuration is easy for you for testing (windows 10) as long as you don't touch the network or can influence my other rigs

I tried to test this on my Vega rigs but have no idea how to set it up Removed the --auto and tried some settings but then gave up

Now I am back to STAK, because CAST and SRB both proved for me to report mucch higher hashrate than the pool, sometimes it can be up to 20% (no i am not talking about instant hash power but calculated over several weeks)

7x Vega rig on STAK has 13500 and more than 14000 on SRB and CAST but pool reports 13500 for STAK and sometimes as low as 11000 for SRB and CAST

So I have high hopes for your miner and will happily switch all my rigs if the thing is ok

My Vega settings are on page 28. There are two entries for each gpu in the config file. Sometimes the index starts from "0" and sometimes from "1". Basically have to run it and see if a GPU isn't working or recognized. Would be nice to get more opinions because there is probably room for improvement.

In the bat file, instead of "auto" change to this:
-c confgpu.txt --no-cpu

[vmozara]

Awesome, thanks, I will try when I get back home, cheers

[JCE-Miner]

Vega: thanks but both for practical and security reason, I cannot work this way. Practical because debugging OpenCL crash the GPU very often (GPU memory is not protected by OS like main memory) and i cannot press reboot on a remote rig. And security because it would involve moving my code through network, and that's too dangerous, even with a secure connection.

I already refused such offers for Xeon optims or Threadripper or Xeon Phi for exact same reason. This is very friendly, but i've more reasons to say no than yes.
However i could invest into a Vega if JCE GPU is successful, even if my favorite GPU is RX550 (1/4 speed of vega for 1/8 of price)

Effective hashrate : i observed the same thing with Claymore 10+ (but not 9.x), a pool hashrate a lot lower (~10%) than the displayed one. Even fees deduced. Curiously, this problem occurs only on close source miners, on open ones, where you can check in code there's no cheat, the reported rate is 100% the same (minus fees).
JCE does no tweak, and you can double check pool side.
This is why I counter-argument with Unclewish about Claymore 11.3 being faster than JCE on HD7800 : the displayed number, maybe. The displayed hashrate on my HD7850 rig was 2200 on Claymore 11.3, and 1950 pool side. With JCE, it's 2080 both side. Even after 24h.

"APU not working" : i need more info. Miner crash ? APU not detected ? hashrate is zero ?

Hashrate drops to zero : I already experienced that, and most often on the GPU where the screen is plugged (real or dummy plug). That's when we use too much memory, close to max, the miner starts, then the display driver gets some of the memory (it has higher priority) and never give it back, locking the miner.
Fix : use a slightly lower multi-hash value, or plug no screen on that card.

"Job not found" ok i get it, and that's because JCE is a GPU miner based on CPU miner netcode. A round of CPU is so fast that the job never timeout during mining. A GPU round is often 1 full second, a lot more than the ping. So you can get a stale share. I'll fix that in next version. No need to fix on PU version, only GPU.

I'll add your best configs on the Github page, thanks all (especially UnclWish) for the reports.

nVidia : i didn't even test on nVidia cards. nVidia provides a OpenCL driver so it may work, but JCE is made with OpenCL, not Cuda, so i don't expect it to be that fast.

New! Debian packages
Fellow dev eLvErDe ported the Linux zip into two debian packages (32 and 64 bits). No change on miner itself, but should be more Linuxish.
https://github.com/jceminer/cn_cpu_miner

[syncro2017]

However i could invest into a Vega if JCE GPU is successful, even if my favorite GPU is RX550 (1/4 speed of vega for 1/8 of price)

Can i suggest also invest into rx570/580, i think the majority of people (myself included) have those cards and would really appreciate any further optimizations
I really like the miner, running it already for 2 days on both rigs. Keep up the good work!