Pages:
Author

Topic: Team Black Miner (ETHW ETC Vertcoin Ravencoin Zilliqa +dual +tripple mining ) - page 7. (Read 34492 times)

newbie
Activity: 27
Merit: 0
TBM 1.75 averaging much lower hashrate at the pool than reported in the miner. Says 1.74 but it is really 1.75:

Miner: https://i.ibb.co/J7mydNS/image.png (31mh per rig)
Pool: https://i.ibb.co/WWNk7ct/image.png (29mh per rig)

Going to switch back to verthashminer until this is improved
newbie
Activity: 27
Merit: 0
https://i.ibb.co/xs4Pyyz/Screenshot-from-2022-11-21-18-53-05.png

Reproduced after a difficulty change. This is a problem with the pool. If you change to miningpoolhub.com or hashalot.com you should get 100% accepted. There are some things you can do to improve it. Reducing the intensity might help. Also please check that you have a low ping on the connection. Changing the mining server to a closer pool will help.

I have 20-22ms ping to woolypooly server on my shares (Wooly is the lowest ping of all VTC pools), tried reducing intensity to 256 with same problem. Even after 8h uptime, still 3-4% rejected share rate:

https://i.ibb.co/nRKpvxc/image.png

It's not just at diff changes, either. You can see in this picture the time stamps, there was no diff change, just the block was solved and the miner submitted the share after:

https://i.ibb.co/S3KSJtp/image.png
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer


Reproduced after a difficulty change. This is a problem with the pool. If you change to miningpoolhub.com or hashalot.com you should get 100% accepted. There are some things you can do to improve it. Reducing the intensity might help. Also please check that you have a low ping on the connection. Changing the mining server to a closer pool will help.
newbie
Activity: 27
Merit: 0
Testing new 1.75 version and it seems to have fixed both the miner restart errors on VTC as well as the miner stalling with 0 difficulty. However, it seems rejected share count is high as you can see here:

https://i.ibb.co/3kw7bg6/image.png

2 rigs of 12 cards each both have 3-4% rejected shares. The shares seem to all be from "job not found" errors when a share is submitted after block was solved already
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Linux builds added.


We have tested v1.75 the cuda 11.8 on the RTX 3080 stock clocks using the TRM hashrate tester in the Ethereum classic algo. (1 second per job)

Here are the results.

TeamBlackMiner: 92.04 MHASH
LOLMiner 91.10 MHASH
T-Rex 90.45 MHASH
NBMiner 90.05 MHASH
Gminer 90.04 MHASH
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
v1.75
1. Removed 2 bugs in verthash mining.
2. Reduced the rejected shares and improved the poolspeed in ETHW/ETC
3. Improved performance on ezil.me / zil switching pools / ethproxy. 0 rejects on dag switch.


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

TeamBlackMiner_1_75_cuda_11_8.7z
https://www.virustotal.com/gui/file/2bb379aab6ce9e5d9000295de65c7272b9a884e94c17732e0cdab1c7869b77a8?nocache=1

TeamBlackMiner_1_75_Ubuntu_18_04_Cuda_11_8.tar.xz
https://www.virustotal.com/gui/file/1a66dfcf8739bea4a00e773a0e8677e2e79249490933d75c7e467c8b921a983f?nocache=1

TeamBlackMiner_1_75_Ubuntu_18_04_Cuda_11_6.tar.xz
https://www.virustotal.com/gui/file/cfe71b03eb9719143e7ce5ec23fc68d3e1d6aa51b8d2523fa2ca651c0f2f6247?nocache=1

TeamBlackMiner_1_75_Ubuntu_18_04_Cuda_11_5.tar.xz
https://www.virustotal.com/gui/file/646561c6a70c627b776dddde51a648baba7dc43e94455df8eb0e7d9f1481415c?nocache=1


sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
I talked to you before about the 170HX rigs mining VTC and getting double free errors which cause the miner to restart and this is happening now again on version 1.72, 1.73, and 1.74 very frequently mining to WoolyPooly. I even adjusted the xxintensity down from default to 1024 to 512 to 256 and I am still getting these problems. Restarts happen in less than 1 hour apart usually.

This issue has been solved in v1.75. will release soon.
newbie
Activity: 27
Merit: 0
Another bug I found mining VTC on woolypooly is that sometimes a rig will reach a point where the miner stops submitting shares and the difficulty is reduced to basically 0. The miner will not restart and will sit idle until I restart it which is a big problem: https://i.ibb.co/bvgWvwH/image.png
full member
Activity: 1372
Merit: 216
Quote
Team Black Miner v1.73
1. Submit solutions in a seperate thread. Faster poolspeed hashrate in ETHW/ETC and Kawpow.

Quote
v1.74
5. Cpu verification moved to submit thread. (faster poolspeed on low difficulty pools and simulators)

I was wondering if someone would implement that. No need for the GPU to wait for verification and submission, just keep hashing.
newbie
Activity: 27
Merit: 0
1.74 final is out with linux

I talked to you before about the 170HX rigs mining VTC and getting double free errors which cause the miner to restart and this is happening now again on version 1.72, 1.73, and 1.74 very frequently mining to WoolyPooly. I even adjusted the xxintensity down from default to 1024 to 512 to 256 and I am still getting these problems. Restarts happen in less than 1 hour apart usually.
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
1.74 final is out with linux
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Team Black Miner v1.74

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

v1.74
1. Removed rejected shares bug on ethproxy pools.
2. Fixed a bug in poolspeed on ethproxy and vardiff pools.
3. The vardiff poolspeed in the console now reset on every difficulty change.
4. Fixed the zil dagcache for AMD. Faster switches between zil/etc/eth
5. Cpu verification moved to submit thread. (faster poolspeed on low difficulty pools and simulators)

TeamBlackMiner_1_74_cuda_11_8_fix.7z
https://www.virustotal.com/gui/file/cab45c06f01be6675a0e2659c48d73f6ffc49729ea304369bbc542aa81c77365?nocache=1

TeamBlackMiner_1_74_Ubuntu_18_04_Cuda_11_8.tar.xz
https://www.virustotal.com/gui/file/b44bdee3d4165d4de8c2d5b9190434cb1a21de3428e0428222209d130ca0020f?nocache=1
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
We have found a bug in v1.73 with pools using the eth proxy protocol. ezil.me / hiveon.net etc.. Will release next version soon.
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Team Black Miner v1.73

1. Submit solutions in a seperate thread. Faster poolspeed hashrate in ETHW/ETC and Kawpow.
2. Cuda 8 tuned ETHW/ETC/ETF kernels, abit faster on RTX cards.
3. Fix nonce problem that caused rejects on some pools. (Ethash/etchash)
4. Added Ethereum Fair pools.
5. API is now running at init.
6. Test on the Verthash filesize, and recreate on wrong size.



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

TeamBlackMiner_1_73_cuda_11_8.7z
https://www.virustotal.com/gui/file/916f27fb8fc1da8bfd29aa637367dd4bcabf62d57e6410d4dbd2c25f2aa2899f?nocache=1

TeamBlackMiner_1_73_Ubuntu_18_04_Cuda_11_8.tar.xz
https://www.virustotal.com/gui/file/470bbb27162417396e2e36bba7591e6ca61f5ce9895bb02c6de77751ce3cd097?nocache=1
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Driver, windows version and build? The cuda 11.8 build should have nvrtc-builtins64_118.dll in the same folder. Cuda 11.7 nvrtc-builtins64_117.dll and cuda 11.5 nvrtc-builtins64_115.dll
member
Activity: 221
Merit: 12
Did you try the cuda 11.5 build? With the cuda toolkit installed you can remove nvrtc64_112_0.dll  from the folder. The wrong lib. Will try a better way to do it in the next release

Yes, same error.

try the 1.72 version. Different dlls for different toolkit versions, so make sure you use all the files in the folder of the build


Yes, I tried with 1.72. Maybe I need something to do, I am not sure, I just downloaded the miner and started the RVN bat file. With Vertcoin bat (on 1.69 version) everything was working fine.
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Did you try the cuda 11.5 build? With the cuda toolkit installed you can remove nvrtc64_112_0.dll  from the folder. The wrong lib. Will try a better way to do it in the next release

Yes, same error.

try the 1.72 version. Different dlls for different toolkit versions, so make sure you use all the files in the folder of the build

member
Activity: 221
Merit: 12
Did you try the cuda 11.5 build? With the cuda toolkit installed you can remove nvrtc64_112_0.dll  from the folder. The wrong lib. Will try a better way to do it in the next release

Yes, same error.
sp_
legendary
Activity: 2884
Merit: 1087
Team Black developer
Pages:
Jump to: