Topic: [XMR] JCE Miner Cryptonight/forks, now with GPU! - page 9. (Read 90832 times)

JCE, I've posted the Stellite test mining pool, did you find time to test your miner with the new algo ?

@pbfarmer: i'll redo some tests, but when you mine with the CPU+GPU, do you get some message like

If yes, so the shares went to the pool, and were accepted.
I've a Ryzen + AMD RX rig so i will be able to do the exact same test.

Try pressing R key to get the effective hashare of each device. CPU should not be zero, unless it has a negligible hash speed compared to the GPUs. But a Ryzen is quite poweful.

Not that the pool connection and fee sessions are global, not per-device, so it cannot be something like CPU disconnected or stuck in fee session. Otherwise you would have zero hashrate on all devices.

please tell what coin you mine on what pool, so i can do an more precise test.

webchain: no support planned, that's not a good algo for CPU, and i don't like how they changed the CN netcode conventions just because they could. Same for Hycon. Both coins can be mined with the CPU reference xmrig fork (possibly recompiled with zero fee), or on GPU with SRB.
I may add them later, but i'm getting less dev time those days and i need to choose between what i support or not. Currently i'm finishing the Stellite v8 fork.

@JCE-Miner - I think there's a bug w/ mixed GPU + CPU mining.  The CPU portion seems to not only not be counted in your effective total, it doesn't show on the pool side either - seems like it's just missing altogether, other than in the hashrate report section.  I've monitored this over several weeks as a combined process, and have witnessed the problem over multiple rigs.  In the past couple days, I've isolated the two miners into separate processes, which now results in the proper effective and pool-side hashrates.  This may just be a Ryzen issue - I can't definitively say it happens w/ Intel CPUs, as the one I'm using has such a low h/r, it's hard to tell if it's included or not.

I also realize the hashing power difference between CPU threads and GPU threads could cause the CPU to have lower output on a difficulty suited to a GPU rig, but again, I've monitored this over weeks (if not months,) over multiple versions, and its pretty reliable that the CPU h/r simply never appears in the actual results.

thx for your answer.
my antivirus quarantined it immediately, so my first thing was to add to the exceptions. now i downloaded again, but nothing changed.
the cpu 32 bit version works, but after I close the program, crash.

the gpu version available only 64 bit, right?

*update
problem solved. av exceptions isn't enough. I added the miner to the trusted files, now everything ok.

What about webchain???

Hello all:

crash: please look at your antivirus and related (antispyware...) it may have corrupt/quarantined the binary since it's always detected as a virus, like all other miners. The attrib trick is documented and works this way since day one.
Maybe try to run the CPU 32-bits versions, it's a lot less detected. Just to test, as it mines slower.

ETH: above any financial consideration, it just would be hopeless to compete with Claymore, who's just the best referenced, and technically the best. I would have like zero users, assuming i could make a working miner.

The GPU miner can mine cpu-only on Win7, even if pointless. At least it should start.

Yeah that's simple and nice to have, good idea.
I'll add the new Stellite, but no urge. I'd like to get a test masari pool to release both, every upload is 10M and i'm flooding my github history

Please, add key to rig identifier for pool-side statistics.

I think it is less profitable as the emission is changed from 3 to 2 ETH per block.

GPU version of miner works only on Windows 8.1+

I'm really hoping for this to get nvidia compatibility, that would be awesome. Great software and nice updates

i tried the gpu version on my laptop (win7 64 bit) but the attrib.exe always crashed on start. then i tried the cpu version, but the result are same.

this is the win's log:

Happy new year !

Here comes the stellite test pool:
https://testxtl.cryptopool.space/#/home

No height announced yet for the fork.

Hoo, i've spent some time to try to reproduce the bug, but fine you figured it out.
Yeah i reworked a lot my memory model in the recent versions, that's very normal the good mem config got different.

The bad share come from memory overflow, i guess your GPU returns zeros and a zero share has an infinite value (while completely invalid) so it gets proposed, and rejected.
When the 1st share is rejected, i display a help message about the config, since it's a very common problem.

Eth miner: i would compete with very very good opponents here, including Claymore Dual, and i would enter the game so late that i don't think i could get any non-symbolic amount of users. Not saying i just don't have time to work on a third miner, after CN cpu and CN gpu.
Claymore did because he's 1) very good 2) entered the competition first, more or less.

0.33m CPU Linux is online

JCE, didn't you thought about making ethereum miner? About a week it's more profitable than any cn algo on 470/480/570/580 cards...

on both my platforms(one is igpu other is non igpu) it says the same thing, low diffuculty share found, not accepted by pool, please use correct algo......BUT im using correct algo - 15 for GRAFT, also tried leaving 0 for auto recognition, same problem. And this problem is only here when using custom config for gpu, when set to auto no problem.

please if someone can copy/paste their custom thred config for rx470/480/570/580/vega 64 for cn v8 algo(monero/graft), and also the last few command lines in start.bat so i can compare.



EDIT: located the problem, something is wrong with my worksize settings(cannot use older worksize settings 944 for 4gb rx580., and 1680 for 8gb ver.)



This is auto settings and it works....
"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1008 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1008 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":864 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":864 },
],


This is from my older 033b6 configuration and it doesnt work on new miner(only worksize is the diffrence)
"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1680 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1680 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":944 },
],




Would appriciate if someone can post config for rx470/480/570/580/vega 64 for cn v8 algo(monero/graft)

its not related to igpu problem because same problem appears on amd 970 chipset(no igpu platform), so its something else

the post above configurations is from non igpu platform, also tried on intel mining rig by shifting indexes(first thing i tried) no working, says something thsat it cant recognize gpu at highest index(3)

--legacy doesnt help on either igpu and non igpu platform

you use gpu 0 in your manual config, and you have an Intel igp on your second rig.
In JCE Intel gpu/igp are gpu 0, so just shift your indexes by one and that should fix.

note the list of gpu and their indexes ate listed at startup, in green, in such case, this is the very first thing to check.

the autoconfig skips the intel gpu by default, hence why it works.

otherwise, i advise you run as auto, note the config it generates, and use it as a base to make your manual config

another tip: add parameter --legacy it may help in case of regression.

Hi all and wish u a happy new year.

dindnt update miners for some time and now have problems with setting up latest ver. of jce gpu miner 033b15 on my rigs(last ver. that i use is 033b6 and it works fine).

Problem is when i try to use gpu configuration file settings from 033b6 ver. to newest version 033b15 miner crashes(says low diff share and then crashes), but when i use auto settings it works in 033b15 it works fine.

Can someone please help and point out what am i doing wrong(amd 970 chipset-no igpu platform), these r my settings from configuration gpu file and start.bat(4*rx580 4gb and 1*rx580 8gb)


/* The CPU configuration. This is a simple EXAMPLE with one core.
   Replace by configuration that matches your CPU. */
"cpu_threads_conf" :
[
     { "cpu_architecture" : "auto", "affine_to_cpu" : 0, "use_cache" : true },
],


/* The GPU configuration. This is a simple EXAMPLE with two low-intensity threads on the first GPU.
   Replace by a configuration that matches your GPUs. */
"gpu_threads_conf" :
[
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 0, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1680 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 1, "multi_hash":1680 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 2, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 3, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":944 },
     { "mode" : "GPU", "worksize" : 8, "alpha" : 64, "beta" : 8, "index" : 4, "multi_hash":944 },
],





Start.bat conf:


@echo off

rem /!\ This is an EXAMPLE of miner start script, replace all EXAMPLE values with yours /!\

rem Pool address, can be a name or a dotted IP
rem Replace that EXAMPLE value with yours
set POOL=pool.graft.hashvault.pro

rem Pool port, can be standard or SSL.
rem If SSL, see note below.
rem Replace that EXAMPLE value with yours
set PORT=5555

rem If you use a SSL port, change to: set SSL=--ssl
rem Otherwise leave empty
set SSL=

rem Your wallet. Some pools let you add extra information at the end like .PAYMENTID or .FIXEDDIFF
rem no problem, that's supported.
rem Replace that EXAMPLE value with yours
rem This is the donation XMR wallet.
set "WALLET=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"

rem 'x' almost always works, some pools let you put a Worker name here. Rarely important.
set PASSWORD=x

rem  0 is "automatic", you can force one of:
rem  1 = Original Cryptonight
rem  2 = Original Cryptolight
rem  3 = Cryptonight V7 fork of April-2018
rem  4 = Cryptolight V7 fork of April-2018
rem  5 = Cryptonight-Heavy
rem  6 = Cryptolight-IPBC (obsolete)
rem  7 = Cryptonight-XTL fork of May-2018
rem  8 = Cryptonight-Alloy
rem  9 = Cryptonight-MKT/B2N
rem 10 = Cryptonight-Arto/Fest
rem 11 = Cryptonight-Masari/Fast
rem 12 = Cryptonight-Haven
rem 13 = Cryptonight-Bittube v2
rem 14 = Cryptolight-Red
rem 15 = Cryptonight V8 fork of October-2018
rem 16 = [reserved]
rem 17 = Cryptolight-Dark
rem 18 = Cryptonight-FreeHaven/Swap
rem 19 = Cryptolight-uPlexa

set FORK=15

rem The legacy mode may provide a better hashrate on some combinations of
rem cards and algorithms. Uncomment the next line to enable it.
rem set LEGACY=--legacy

rem the GPU environment variables
set GPU_MAX_HEAP_SIZE=100
set GPU_USE_SYNC_OBJECTS=1
set GPU_MAX_ALLOC_PERCENT=100
set GPU_SINGLE_ALLOC_PERCENT=100

rem All is good! Let's mine
@echo on
jce_cn_gpu_miner64.exe --no-cpu -c config.txt --any --forever --keepalive --variation %FORK% %LEGACY% --low -o %POOL%:%PORT% -u %WALLET% -p %PASSWORD% %SSL% %*
@pause


also have the same problem on my other rig that is with vega 64 and rx570 on intel based platform with igpu.


tnx for answers and help

very unlikely, and same for hycon. those forks are rather netcode forks more than algo forks, and i probably won't have time to do the needed netcode refactoring, for such obscure coins. if their algo were cpu friendly i would be more motivated, but that's not the case
the cpu mining is possible with the official xmrig forks, and gpu on recent Srb