Topic: [XMR] JCE Miner Cryptonight/forks, now with GPU! - page 84. (Read 90814 times)

Sincerely your miner is best as it's the one gives me highest hashrate.pls add gpu temp , backup pools (just in case) and the ability to automatically reboot the rig in case of gpu failure or miner crash.

Awesome, thank you!! Getting 620 now (cpu at 3.8GHz with really bad memory timings, i need to tweak)

Now all my rigs and home PC are officialy JCE

I also mine with a 1700.
My settings for Crytponight v7 is one thread per physical core, all using cache.


The result is about 700 h/s.

I am trying to set the JCE miner on my home PC, that is Vega 64 and Ryzen 1700

When I do the "auto" my CPU is hashing 600+ H/s, but when i remove auto and put -c config.txt to use the GPU as well, no matter what config i try, my CPU always ends up hashing 300-400 or even less. I am trying to mine CN7. Anybody have some good configuration for 8-core ryzen to copy paste into the config for CN7 mining?

Question 2. How to see why is the difference between blue hashrate and effective hashrate? Only few times so far I have seen stale rejected share or duplicate, but sometimes there is quite the difference between blue and effective hashrate and I would like to see if I have some hardware errors or something

Question 3: Why is it not recommended to run the log file?

Question 4: I am mining on moneroocean pool, how to know if the miner is mining properly if the coin is switched by the pool?

I started the rigs more than 40 hours ago and so far there was 0 issues with stability. Whenever I had srbminer uptime for more than few days, some cards would lose few H/s, up to hundred or two. Here, so far, perfect. I didn't even do any tuning, justa copy paste of config with 1936 intensity that fellow forumer put here, and 0 issues with best hash rate I have ever seen from a miner.

Is there any chance you get a Vega card for better development?

There's much more big time miners with GPU rigs running on Linux than you think. They don't like to babysit 24/7 like you have to do with Windows. That's also where most of the hashes could hit your miner because they do like to scale things way too much

You can't see them on the CPU version because they don't mine with CPUs to get 100 more hashes. People who you can see on the CPU version are just using old hardware to squeeze a few extra hashes from their old laptop, not a serious miners.

There's so many miners saying if they can get 2K hashes on Cryptonight from Vega they would switch back to Linux in seconds. I personally use my Vegas on another algos just because I'm not planning to use Windows even if I can make a few more bucks in a week. No thanks.

I don't have any vega to test on, but I experienced the same problem with my HD7950s, they were more stable alone than together.
I fixed this by increasing the input PSU voltage to 12.20V so the Vdrop caused by all the cards mining at max could be compensated. However this can be done only if you have control over your 12V rail voltage.

I'm building a minor update of the GPU version now, with per-GPU pause and the OpenCL side effect regression found by UnclWish with --no-gpu fixed. (thanks uncl btw).
I will also make back ETN a CN-Classic coin by default.
And add the double-check by CPU of found shares, à-la Claymore 10. Disabled by default, enabled by parameter --doublecheck

I dont think it have anything with undervolting or intensities, because if i am able to use Mem clock up to 950 if its only 1 vega per rig... But if i add another vega, i need to lower both Mem clock to 935... and if i add another vega... i need to lower all vegas again to 920, and if add another vega.. i need to lower all the vegas to 910.. thats way its odd....

 and it happen to both rigs that i have...

Are you undervolting?

Also check your intensities on the miner software. Miner software can affect how far you can clock

Unfortunately all vegas are different. Every single one of mine has a different max clock on memory.

Hi,

I have 2 mining rigs :
1st rig  = 4 RX 580 8 GB (MSI) and 2 RX Vega 56 (Power Color Red Devil)
2nd rig = 5 RX Vega 56 (Power Color Red Devil)

Both installed win 10 1709, AMD driver 18.5.1, using Hard disk & mining normal_v7, why in rig 1 (2 Vegas), I can set my Vega 56 with Mem P3 Clock to 920, but in Rig 2(5 Vegas), the max Mem P3 Clock I can only set is 895 without any error with Overdriventool??

Any ideas why this is happening??

Any help appreciated.

Thanks

really?
i read he said he already had a better code. i really see no good ideas in the changes, just some trivial fixes.
alas, my code is different and yes i've some ideas to improve it

Ok, just wondered if it gave any clues where and how to get heavy more efficient.
Interesting that Doc made some improvements to SRB based on it, just saying.

I'm the most late, a lot of simple features are missing like temperature monitoring or autoconfig.



if you're talking about this
https://github.com/imperdin/xmr-stak


well, first it applies only to xmrstak, i use my own code, based on jce cpu 32 Bits
and... it's mostly a bunch of changes that fix some goofs of the Wolf0 code, add some new ones, leave most of them in place. Nothing interresting. and since jce is close source, it would be illegal to take his code.

Thanks! You're the most hard working dev around here

Good idea, i'll add parameter --doublecheck to make CPU re-check the share, like Claymore 10.

i ordered a A8 APU on socket FM2 to make JCE compatible with APUs

Hi JCE,

When using your GPU miner, can you advise as to whether or not you are currently verifying the GPU result by cross checking with CPU that is is valid before submitting to the pool? Some other miners perform this check and will avoid submitting if there is a discrepancy, merely noting it as a 'Compute Error'. If you aren't, would you consider adding the functionality, even if only via a config or switch? If it slows down performance, it can still be beneficial when testing more aggressive clocks and multi_hash/intensity settings to have this check before submitting when trying to find the performance vs stability sweet spot and not wanting to get banned by the pool for invalid shares.

I received this error in JCE while testing the reasonable limit of each GPU so it got me wondering about this. My static difficulty is definitely appropriate as it's not even 25x my hashrate. I am sure that lowering my clocks will reduce or eliminate this issue, but I thought I'd include this result for informational purposes.


Thank you for considering this. As things stand now only having the ability to specify one pool, it can result in complete downtime if invalid shares are submitted that are outside the proper difficulty level and the pool issues a ban.

You should add "with mod bios and optimized clock/power/ram settings". @Stock the performance is significantly lower!  

Heavy is not good optimized yet. But on 570/4 speed will be about 750-800, 580/8 - about 1000 h/s.

FYI: Electroneum is back to ASIC-friedly CN original, too! 
https://www.facebook.com/electroneum/posts/2030562537205714

I'd go for CN-Lite, AeonCoin for example. There the CPUs' advantage over GPUs is max. With a scratchpad size of 1 MB per thread you can start 32 threads on your 2 socket system and the data still fits into the L2 cache of your Opterons 6376. When you activate double-hash, you can still run 32 threads, filling the L2 and L3 victim cache for the last extra percent of performance. "Bulldozer was without doubt an unmitigated failure" (Andrew Feldman, Corporate Vice President and General Manager of the Server Business Unit, AMD), but at mining CN, it is a beast!      

Hi Guys,

Are there some estimated Hashrates for CN_Heavy?

I am interested in:
RX570, 4 GB
RX580, 8 GB

I am actuallly uses SRB Miner

Thank you