Topic: [XMR] JCE Miner Cryptonight/forks, now with GPU! - page 90. (Read 90841 times)

Strangely i dont need to. It changes directory fine. I think you only need to use quotes in the command line if your path contains a space, but in a batch file its fine. windows lol.
Anyways: still compile error for the opencl kernel :/

If path is long-name you must set it in quotes:
W:
cd "W:\Miner\JCE CN GPU Miner"

the .bat is in te same dir, sadly. Also giving the absolute path to config instead of a relative doesnt change the behaviour.

This is my current start-batchfile:

ok for apu, i'll probably buy a little used apu to test that. on fm2.

shares : i know the stale share loss, and the -10% i observed was with same difficulty than with JCE (120000). really i never had anything above 90% of displayed hashrate pool side. while claymore 9.7 was perfect. on same rig.

A10-5800к APU:

Detecting OpenCL-capable GPUs...
UNKNOWN_ERROR

When i check with GPU-Z, gpu is detected as 7660D and has the OpenCL tick.

The answer for reduced hashrate on the poolside is here:

All GPU miners looses effective speed on rejected shares because of job was gone while GPU findes a share.
It's a lucky thing. And pool difficulty. On high pool diffilulty rejected share have big amount, and speed lowers significant.

That's why on the same miner different peoples often have different poolside effective results.

I agree, 550s are great but only when you have space to put 12GPU rigs and a lot of them.
In my case, its all about the density, but VEGAs are too expensive so i rely on 570/580s

i don't know why, i just bought two rx550, same speed and half the price, and consumes less power.

ho, and to answer a previous question: gpu and cpu combined. i think it's good because it allows a cpu to participate to big share mining without being kicked. sure it will find one share per hour but it will worth a lot. i like to see my core2 find a 120000-value share, in the middle of my gpus

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!

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

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

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

Is there a version of this for Nvidia 1080Tis?

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

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

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

Nice settings, thanks for the tip!

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


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


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

Tested on my APU - did not work.

Is it going to support Nvidia Cuda in future?

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