XMR-stak-JK 2.10.7 Compiled with no devfee (Updated: 2019-8-18)

[MagicSmoker]

Well, I have to amend my earlier comments about it being painless switching over to 2.5.1... The CN coin I mine - LTHN - just switched to CNv8 and now I am dealing with unrelenting driver and system crashes. I just DDU'ed the 18.5.1 driver and installed/pixel-patched 18.6.1 to no avail. I've also bumped up the core/mem voltages to as high as they were stock (ie - no power saving) and that doesn't seem to help, either.

So I next decided to try SRBMiner and it is crashing the driver, too. I'll keep plugging along for another hour or so, but after that I am just going to stop mining completely.

UPDATE: in a fit of pique I deleted the amd.txt file for both rigs and let xmr-stak automatically generate them. Hashrate is depressingly low - 1020-1060H/s for the RX 570 + Ryzen 5 1600 system (previously ~1180H/s), and <1600H/s for the 4x RX 560 system (previously ~1900H/s), but at least it isn't crashing anymore (and I hope I didn't just jinx myself...).

[1714560883]

1714560883

[1714560883]

1714560883

[1714560883]

1714560883

[1714560883]

1714560883

[1714560883]

1714560883

[JaredKaragen]

Well, I have to amend my earlier comments about it being painless switching over to 2.5.1... The CN coin I mine - LTHN - just switched to CNv8 and now I am dealing with unrelenting driver and system crashes. I just DDU'ed the 18.5.1 driver and installed/pixel-patched 18.6.1 to no avail. I've also bumped up the core/mem voltages to as high as they were stock (ie - no power saving) and that doesn't seem to help, either.

So I next decided to try SRBMiner and it is crashing the driver, too. I'll keep plugging along for another hour or so, but after that I am just going to stop mining completely.

UPDATE: in a fit of pique I deleted the amd.txt file for both rigs and let xmr-stak automatically generate them. Hashrate is depressingly low - 1020-1060H/s for the RX 570 + Ryzen 5 1600 system (previously ~1180H/s), and <1600H/s for the 4x RX 560 system (previously ~1900H/s), but at least it isn't crashing anymore (and I hope I didn't just jinx myself...).

yeah with the upgrade, all the .txt files need to be re-generated (has something to do with formatting and variable fillout for the new variables stored in the file).

My GTX470 is no longer "stable" enough for mining CNv8.   After 4-24 hours, it will cause a cuda crash in the miner.   After removing it from that test rig, no issues.  Odds are that it was just running to hard for its cooling capability;  remember, the 4-series cards were the most power hungry and hottest of any other Nvidia card ever made.


Hashrate seems to be minus 1/4-1/3 or so compared to cnv7 (accounting for my 470 when it was plugged in as well).   So far, no issues with the rx560.



It only made sense that things went the way they did (more power draw, harder on hardware, etc), because its the only way to overwhelm the FPGAs.   They can only have so much responsiveness, and signal timing across a complex FPGA configuration can be quite a chore.    Ive been doing my research and learning on how FPGA's are programmed and work recently;  and.... well, its not an easy task.    My first project personally will probably be making some sort of RISC cpu (maybe a V) on a FPGA and go from there.

notice many cn7 coins raising in value now;    the mass-FPGA holders are probably doing what it takes to squeeze every little pinch out of cn7 that they can before others catch on and they get dumped down to a realistic value... same thing happened during the cn7 swap.

[MagicSmoker]

...
UPDATE: in a fit of pique I deleted the amd.txt file for both rigs and let xmr-stak automatically generate them. Hashrate is depressingly low - 1020-1060H/s for the RX 570 + Ryzen 5 1600 system (previously ~1180H/s), and <1600H/s for the 4x RX 560 system (previously ~1900H/s), but at least it isn't crashing anymore (and I hope I didn't just jinx myself...).

yeah with the upgrade, all the .txt files need to be re-generated (has something to do with formatting and variable fillout for the new variables stored in the file).
...
Hashrate seems to be minus 1/4-1/3 or so compared to cnv7 (accounting for my 470 when it was plugged in as well).   So far, no issues with the rx560.

I'm still trying to dial everything in with respect to stability and hashrate but as of now I have gone back to 2 threads on the RX 560 rig, using all the other auto-generated parameters except reducing the intensity to less than half what was specified for 1 thread operation (e.g. - 768 for 1 thread goes to 376 for 2 threads). Right now the 4x rig is doing a little over 1800H/s so a massive improvement from the <1600H/s I was getting before! However, I just made this change so now I need to see if it remains stable as I was still having problems with the 1 thread/768 hashrate config (e.g. - hashrate dropped from 400 to 300 on one of the cards sometime over the night). I'm still also still running the RX 570 system with a single thread, worksize of 8, and 768 intensity for now, mainly because I want to see how the other rig behaves before wasting more time fiddling with this one, but I suspect I will be doing the same tweak in a few hours.

It only made sense that things went the way they did (more power draw, harder on hardware, etc), because its the only way to overwhelm the FPGAs. ...

Totally agree - I have slowly but surely come around to the idea that it is the energy cost that determines how secure a network is to the first degree, not the absolute hashrate or cost of the mining hardware, per se. One of the coins I was an early supporter of that remained on the original CN algo (DERO) has a network hashrate in the range of 400-500MH/s which works out to around 2000 ASICs and 1.2MW total power draw. Prior to being ASIC'ed it had a network hashrate of around 4MH/s which works out to ~5300 RX 570 GPUs drawing ~800kW of power; hardly an improvement in network security, really, at the cost of alienating a massive number of previously loyal GPU and CPU miners, but I digress...


UPDATE: Hashrate dropped on one of the RX 560 (not the same one) yet again, then the display driver crashed so I decided to DDU 18.6.1 and install 18.5.1, then pixel patch and set all the cards to compute mode. I also stopped using OverDrivenTool and switched over to MSI AB, as v4.5.0 is once again able to handle more than 3 AMD GPUs. I have core set to 1200, mem set to 1900, power limit set to -20%, no change as of yet to core voltage, and total draw is at an unhealthy 375W with hashrate at 1740H/s. So a slight drop in hashrate, a rather higher increase in power draw, but if it's more stable I'll consider that a fair trade overall.

UPDATE 2: Well, I finally gave up on xmr-stak on the 4x RX 560 rig because it kept losing hashrate on one card then outright crashing the driver if I ignored that long enough. Even though I had no luck with SRBMiner 1.6.8 before, I decided to give it another go since I downgraded the driver to 18.5.1 and whaddyaknow, it now works. It's only been running about 10 minutes so far, but delivering a fairly consistent 1775H/s at 356W; so, a little better hashrate and a little lower power draw. Neither are enough of an improvement to justify the 0.85% dev fee, but if it can stay up for longer than 4 hours without needing a reboot then that will be worth the fee to me. I'd like to stick with xmr-stak on both my rigs, but v2.5.1 clearly needs some bug-fixes and/or additional tweaking.

[Umister]

The XMR-stak 2.5.2 is out. Will you do a compilaiton? Thanks in advance.

[MagicSmoker]

The XMR-stak 2.5.2 is out. Will you do a compilaiton? Thanks in advance.

Nobody likes a nag. Thanks in arrears.

[JaredKaragen]

The XMR-stak 2.5.2 is out. Will you do a compilaiton? Thanks in advance.

Nobody likes a nag. Thanks in arrears.

lol.


lemme look at the release notes and if its worth going through the rigmarole.....   and at first glance, its a nvidia patch.


Ill try to compile in a little bit.   Hopefully the machine I used to compile last time has no issues.

*edit* compiling.   Seems like its working properly.... but ill know in ~10 min when the machine spits it out for me.  Once that happens, its off to mega and the link will be posted.

[JaredKaragen]

Compile a success.  Uploaded and first post updated.

[newmz]

HEY Nvidia miners!

you should seriously check out CryptoDredge instead of XMR-Stak.

I am mining Loki (CN-heavy) with 1070s and I honestly get 20% more hahs rate. I am not the dev or getting paid or anything -  I just think you should check it out.

[JaredKaragen]

HEY Nvidia miners!

you should seriously check out CryptoDredge instead of XMR-Stak.

I am mining Loki (CN-heavy) with 1070s and I honestly get 20% more hahs rate. I am not the dev or getting paid or anything -  I just think you should check it out.

I know that the memory size was not the focus of the CNv8 changes...  Its all about creating functions that will overwhelm FPGA timing and circuit design... for starts....  Its a complex animal to get into.

But my main purpose for releasing/compiling and pushing xmr-stak specifically:  
1) Ive been trying to keep a no devfee miner app available specifically for monero.
2) I understand the code;  the cpp files and their uses.  I have edited them enough now over enough releases;  I get it.  Finally.
3) The source is 100% open and free.
4) less "fluff" associated with having many algos and functions that 99% of people wont use.  Most people who mine CN coins, only mine CN coins.... and would prefer a more optimized app for all of the the others individually.

I can see what angle you are coming from;  but I think you are missing the fact that you can use a cuda 8.0 library when compiling (or replacing my cuda 10 .dll from a separate compilation as I have mentioned in previous posts) and then you would not have to worry about the limits associated with cuda 10.  In response to your other post in another thread (i research shills in this thread before deleting, and you cut it under the wire :  The drivers are cuda backwards compatible (for game compatibility sake in the future for starts) and updating the driver;  is a good thing.  I have yet to have any issues with my 1070 or 1080.  
Only with one game, and one driver version did I have to rollback;  but that was only temporary until the game fixed its bug.

I do 100% understand the power limit you want to set for yourself;  but many people don't set mining limits just based on individual coin power draws;  if they aren't mining to sell immediately.

The official release from githib has: xmrstak_cuda_backend.dll    Just rename it to xmrstak_cuda_backend_9_2.dll and plop it into the folder with my release next to, or in place of my DLL.  See what it does.   Also;  you can just try replacing the file altogether: xmrstak_cuda_backend.dll and it shouldn't have cuda10 libs at the point.     I believe it chooses the cuda DLL based on what driver you have installed;  so if it still tries to load under cuda10 even though you added the cuda8/9 dll's, just replace the file instead of adding along side and it will force it.    I have it added alongside in my personal rigs so that my older generation cards can work alongside my current gen cards in the same machines.

[JaredKaragen]

Ok, so I finally got off my ass and went in and optimized my cards for what I could.

Made my previously ~890H/sec machine into a 1296H/sec machine.... no overclocks or any special configs;  Just modifying the threads, etc.  (poolside i'm seeing 1200-1600h/sec, which is the amount I had on cn7;  Which makes me wonder what it would have been on cn7 if i had done this earlier)


http://www.xmrstak.com   has countless user-submitted configs to give you an idea of what you need to do.  Just compare others' settings to see.      It seems that if I attempt a bios mod on the Rx560 I have, I could go up 30% in hashrate;  but I am not the type of user who does this;  even though I know how to accomplish it and have done it to cards before....   Since my business model isn't based around pinching every penny, or the absolute most efficient;  and the fact that this RX card is not my own as it was graciously lent to me by a fellow board member;  I doubt Ill be doing the bios mod....

But I highly suggest letting the autoconfigs happen, note the hashrates per card, then disable all but one and start trimming the values individually and doing tests;  it's worth the 10-30 min spent.   You can get quite a bump like I did from my 1070; went from ~420H/sec to ~588H/sec with very little effort.  I don't think ill tune it to get the highest possible, this was good enough for me.

[NachoLibre69]

Hello!

Would really appreciate if you found the time to compile the latest 2.6.0 release for us .

Thank you in advance!

[impynick]

I compiled 2.6 already. Its more of an optimization for cnheavy. Didnt notice any difference on my nvidia or amd card on v8 ONLY.

[twotwosix]

Ok, so I finally got off my ass and went in and optimized my cards for what I could.

Made my previously ~890H/sec machine into a 1296H/sec machine.... no overclocks or any special configs;  Just modifying the threads, etc.  (poolside i'm seeing 1200-1600h/sec, which is the amount I had on cn7;  Which makes me wonder what it would have been on cn7 if i had done this earlier)


http://www.xmrstak.com   has countless user-submitted configs to give you an idea of what you need to do.  Just compare others' settings to see.      It seems that if I attempt a bios mod on the Rx560 I have, I could go up 30% in hashrate;  but I am not the type of user who does this;  even though I know how to accomplish it and have done it to cards before....   Since my business model isn't based around pinching every penny, or the absolute most efficient;  and the fact that this RX card is not my own as it was graciously lent to me by a fellow board member;  I doubt Ill be doing the bios mod....

But I highly suggest letting the autoconfigs happen, note the hashrates per card, then disable all but one and start trimming the values individually and doing tests;  it's worth the 10-30 min spent.   You can get quite a bump like I did from my 1070; went from ~420H/sec to ~588H/sec with very little effort.  I don't think ill tune it to get the highest possible, this was good enough for me.

Can you compile 2.6.0 I would greatly appreciate it

[JaredKaragen]

Ok, so I finally got off my ass and went in and optimized my cards for what I could.

Made my previously ~890H/sec machine into a 1296H/sec machine.... no overclocks or any special configs;  Just modifying the threads, etc.  (poolside i'm seeing 1200-1600h/sec, which is the amount I had on cn7;  Which makes me wonder what it would have been on cn7 if i had done this earlier)


http://www.xmrstak.com   has countless user-submitted configs to give you an idea of what you need to do.  Just compare others' settings to see.      It seems that if I attempt a bios mod on the Rx560 I have, I could go up 30% in hashrate;  but I am not the type of user who does this;  even though I know how to accomplish it and have done it to cards before....   Since my business model isn't based around pinching every penny, or the absolute most efficient;  and the fact that this RX card is not my own as it was graciously lent to me by a fellow board member;  I doubt Ill be doing the bios mod....

But I highly suggest letting the autoconfigs happen, note the hashrates per card, then disable all but one and start trimming the values individually and doing tests;  it's worth the 10-30 min spent.   You can get quite a bump like I did from my 1070; went from ~420H/sec to ~588H/sec with very little effort.  I don't think ill tune it to get the highest possible, this was good enough for me.

Can you compile 2.6.0 I would greatly appreciate it

Looks definitely worth doing.

It will be compiled with cuda 10; and the recommended blockchain SDK version from fireice-uk's github instructions.

It could be tomorrow, it could be in a few days.   I have been in and out of it for a bit;  but ill try and get it done.

[impynick]

since updating ive been getting these errors on my 580.


  24 | AMD Invalid Result GPU ID 0      | 2018-11-19 09:31:03 |

24 times since yesterday! I might revert back to 2.5.2 if it keeps up.

on cryptonight v8

[NachoLibre69]

since updating ive been getting these errors on my 580.


  24 | AMD Invalid Result GPU ID 0      | 2018-11-19 09:31:03 |

24 times since yesterday! I might revert back to 2.5.2 if it keeps up.

on cryptonight v8

Just revert to (AMD Driver)5.1 or 6.1. Anything above that gives errors.


Cheers!

[JaredKaragen]

since updating ive been getting these errors on my 580.


  24 | AMD Invalid Result GPU ID 0      | 2018-11-19 09:31:03 |

24 times since yesterday! I might revert back to 2.5.2 if it keeps up.

on cryptonight v8

I was about to say;  2.5.2 is the last version I have released;  so what are you using right now if not 2.5.2? lol

[impynick]

I compiled 2.6 on my own. Its giving me errors on my 580.

I am using blockchain drivers (17.xx) as adrenaline drivers (18.xx) was causing instabilities.

[JaredKaragen]

I compiled 2.6 on my own. Its giving me errors on my 580.

I am using blockchain drivers (17.xx) as adrenaline drivers (18.xx) was causing instabilities.

2.6.0 compiled and released.

Let me know if there are any issues.

Same as before, compiled under Cuda 10, i suggest the developer's cuda libraries for older devices as noted in my other posts in this thread.


Enjoy.

[The?DEvil]

Can you pls compile it to "tar.gz" and upload? Thx a lot. I would try using it in HiveOS.