Topic: Team Black Miner (ETHW ETC Vertcoin Ravencoin Zilliqa +dual +tripple mining ) - page 6. (Read 34492 times)

Kawpow is supporting  --xintensity . The default is 32, but you can run with 64,128, 256 or more. or less if you have little power.


--xintensity 256  on the RTX 3080ti give a hashrate of 66MHASH , but >300Watt

Two different systems (same motherboard for both (ASUS Z590)), both with the same version of Windows 10 and drivers.

Did a new test, DDU drivers on 3090 system, re-installed the 527.56 drivers, same issue, ran 11.6 version, worked fine, retested 12 version and now it's working fine, so no clue as to why that cuda runtime error the first time round.

Strange to say the least.

As for speed, yes, not much between the two versions.

Are the Zotac abd EVGA cards in the same rig or are ypu using them in different rigs. I in different rigs try switching the cards from rigs, check drivers and the actual cuda version. You do know 12 was only officially released last month?

I doubt there is a difference in hashing rates between cuda 11.6 and 12.

@sp_
Could you clarify this?

Cheers

@sp_

Bug report.

error.log file

`00:00:00 [2023-01-03 13:45:06.290] OpenCL driver version: 527.56
00:00:00 [2023-01-03 13:45:06.301] Cuda driver version: 12.0
00:00:00 [2023-01-03 13:45:06.301] Cuda runtime API version: 12.0
00:00:00 [2023-01-03 13:45:06.301]
00:00:00 [2023-01-03 13:45:06.301]                                
00:00:00 [2023-01-03 13:45:06.301]
00:00:00 [2023-01-03 13:45:06.301] Welcome to Team Black Miner 1.78
00:00:00 [2023-01-03 13:45:06.301]                                
00:00:00 [2023-01-03 13:45:06.301]  
00:00:00 [2023-01-03 13:45:06.363] GPU0 NVIDIA GeForce RTX 3090 25GB added to miner as Cuda device
00:00:00 [2023-01-03 13:45:06.363] Api server started at 127.0.0.1:4029
00:00:00 [2023-01-03 13:45:06.364] Trying kawpow.na.mine.zpool.ca:1325 (kawpow)
00:00:00 [2023-01-03 13:45:06.968] Pool responded successfully to extranonce subscribe
00:00:00 [2023-01-03 13:45:06.983] Pool responded successfully to authorize
00:00:00 [2023-01-03 13:45:07.031] GPU0 Starting generating kawpow dag (Epoch: 22)
00:00:00 [2023-01-03 13:45:09.180] GPU0 DAG generated in 2.15 seconds (1228799kb)
00:00:00 [2023-01-03 13:45:09.180] Job 9a02
00:00:00 [2023-01-03 13:45:10.937] GPU0 Accepted (1/1/0) Reply: 78ms
00:00:00 [2023-01-03 13:45:14.869] Job 9a06
00:00:00 [2023-01-03 13:45:19.468] GPU0 Accepted (2/2/0) Reply: 76ms
00:00:00 [2023-01-03 13:45:20.510] Job 9a03
00:00:00 [2023-01-03 13:45:20.606] C:\code\tbminer\source\sp_kawpow\cuda_kawpow.cpp:cuda_kawpow_init_epoch():155: A cuda runtime error occured: invalid argument
00:00:00 [2023-01-03 13:45:20.607] Shutting down threads
00:00:00 [2023-01-03 13:45:20.619] Control thread exited`

This is with the Cuda 12 version, Cuda 11.6 works fine without the error.

Update...

Cuda 12 version on an RTX 3090 Ti (Zotac RTX 3090 Ti AMP Holoblack) - runs fine fine on Kawpow, only my RTX 3090 (EVGA RTX 3090 FTW3 Ultra) has this cuda runtime issue.

Look forward to a fix in the next release.

cuda 11.6 update worked, cuda 12 didn't work. testing 11.6 now. ill come back with results/bugs

latest driver with  cuda 12 support?

try the 11_6 build
TeamBlackMiner_1_78_Ubuntu_18_04_Cuda_11_6.tar.xz

linux update does not work. ran manual update and get this looping error: https://i.ibb.co/XttnMyR/tbm.jpg

linux bins have been added. TBM 1.78 is the fastest miner for KAWPOW on NVIDIA RTX cards. Beating t-rex with a few percent on the POOL..
1.78 has no job not found rejects on wolypooly on default intenstity settings.

Linux today also or no?

V1.78

1. Kawpow +5% on NVIDIA.
2. Added lib to run on amd only rigs.
3. Option --ssl-verify-none added to disable SSL pool certificate authenticity.
4. Reduced stale shares. (Vertcoin)
5. Fixed bug in average hashrate.

https://github.com/sp-hash/TeamBlackMiner/releases/

TeamBlackMiner_1_78_cuda_12.7z
https://www.virustotal.com/gui/file/72d0ac9dda65ba40f1e4af2a9bf6518bb3f140ed6b677390e5909b0c6c7b6695?nocache=1

TeamBlackMiner_1_78_cuda_11.6z
https://www.virustotal.com/gui/file/b6dcb85a387301d9ef82302b1db88c4effc436b349315b626026da50eab42c10?nocache=1

TeamBlackMiner_1_78_Ubuntu_18_04_Cuda_11_6.tar.xz
https://www.virustotal.com/gui/file/94bb7def4c366f86b5b7b45c63c4b6ce4cda7a75c6c0e057613bef59f6cfcd81?nocache=1

TeamBlackMiner_1_78_Ubuntu_18_04_Cuda_12.tar.xz
https://www.virustotal.com/gui/file/f07e3235bda7d8f49a6b5ee3df8d1d92d1c430abbcdad96f8664ba904336e136?nocache=1

Yes. It has been fixed in the next version.
We have 2 bugs left to fix in the ssl implementation and in the improved ravencoin implementation.
hope to release soon.

Also just wanted to add that woolypooly VTC has "job not found" rejects still fairly frequently in version 1.77. Does not happen on hashalot

Sounds good, any chance you can do VTC + zil dual mining? Then would be perfect

In v1.78 verthash will have staleshare fix as well with higher poolspeed.

Tried VTC on Wooly with the same "possible stale" errors a lot, as well. And they result in stale shares. Still also getting about 2-3% rejected on Wooly, and about 0.5% rejected on Hashalot

Using 1.77 on Hashalot VTC, and I get a lot of these "Possible Stale Job" messages:

https://i.ibb.co/4FTH9N5/image.png

Other than that, this version seems good so far. If you could add VTC + zil dual mining using zil ZMP protocol like Flexpool has, it would be perfect. bzminer can do ERG + zil that way for good profits

v1.77

1. Reduced stale shares
2. Added ca.pem file to the Windows archive. (fix ssl mining)

https://github.com/sp-hash/TeamBlackMiner/releases

TeamBlackMiner_1_77_cuda_11_6.7z
https://www.virustotal.com/gui/file/5b88ec0c1549bd03e11db99b3059b32f00760a1ac4933b6df497eb9f4278b4ec?nocache=1

TeamBlackMiner_1_77_cuda_12.7z
https://www.virustotal.com/gui/file/6f0a15e3326cad125f1dd5bc339e6a7b522171b14320f97bc5454b8eeae84a53?nocache=1

TeamBlackMiner_1_77_Ubuntu_18_04_Cuda_11_8.tar.xz
https://www.virustotal.com/gui/file/90e84b1ef1c87a994cb6978867d19fecd7a73997c983fc5fb94b5261b2d1921d?nocache=1

TeamBlackMiner_1_77_Ubuntu_18_04_Cuda_12.tar.xz
https://www.virustotal.com/gui/file/759a4c6a712367e90fe8b3045629ce4b50b62df17632e74aa0f32f9237d5f3a0?nocache=1

TeamBlackMiner_1_77_Ubuntu_18_04_Cuda_11_6.tar.xz
https://www.virustotal.com/gui/file/949ccf565e2b874443b8872883d3b6dffe8d6e5b3c3d1f7ca18a23f8e7c0504c?nocache=1

v1.76
1. Improved the the stratum implementation for ezil.me, hiveon.com and others (ETC/ETHW). Higher poolspeed, fewer rejects.
2. Faster Verthash mining  (Submit solution code in seperate thread)
3. Removed job not found rejects for verthash (woolypooly.com/zerpool.com)
4. Fixed an issue when submitting shares while validating the Dag. (AMD/NVIDIA)
5. Fixed an issue in the zil cache for 8GB cards on AMD.
6. Fixed set powerlimit (TDP)
7. Fixed SSL for Nicehash and others.
8. Updated thirdparty libraries.

https://github.com/sp-hash/TeamBlackMiner/releases

TeamBlackMiner_1_76_cuda_12.7z
https://www.virustotal.com/gui/file/211e73bf279c3a4d7e6d271a8411a51f200c0db0880f90f4eb95f5f05b9e31fc?nocache=1

TeamBlackMiner_1_76_Ubuntu_18_04_Cuda_12.tar.xz
https://www.virustotal.com/gui/file/42bbd85b677963db5a94d62d7b22f6022e947db6aeaa2be02504ec59978e8064?nocache=1

Woolypooly have changed their stratum code so some adjustments are needed in the code. The next version v1.76 will fix it.

You should consider switching to another pool because this pool now have over 59% of the total hashrate and reject 4% of your shares.

This is a common complaint