Topic: Team Black Miner (ETHW ETC Vertcoin Ravencoin Zilliqa +dual +tripple mining ) - page 31. (Read 35090 times)

Try v1.42 cuda v11.4 build. 11.5 might have issues. As reported in the previous post. 3108 might not be the optimal config for your rig. This setting works good on 1 rtx 3090.

I tried on 2miners , on a 12x3080 rig, Hive OS cuda 11.5 driver 495.46 cuda 8 xint 3108 , I have 1321mh in the miner window. Average per pool per day 1.24 ghs. It works well for the first hour, then it starts to fall behind. I can put it for you again for a test for a day, so that you can be convinced of this yourself. Maby in hive i must use 11.4 cuda , i dont know, but it not work ....

Now I will turn it on on cuda 11.4 And I will throw off the test result in 2-3 hours here. Cuda 11.5 doesn't work on Hive os, probably because autotuning always chooses core 1 for my cards on any settings.

Which version did you test. Which mining pool did you test? Let it run for 5000 shares. TBM should be faster if you configure it correctly.. What speed are you getting in the miner window. Screenshots please.

Hello, I tried this miner with different settings with different driver versions in Hive os, left it for a day, but G miner on a similar rig from 12 x 3080 always gives more. I don't know how to set it up and I haven't really seen anyone who was able to consistently get a good result on it. At first it gives good speed, but after a couple of hours Gminer overtakes it even without a stail share, so draw your own conclusions.......Maybe there are some miracle settings that I don’t know about, but it loses outright to other miners. All these photos with measurements on 1 card for an hour are not a test, but prophonation.
I tested it on 2 identical rigs and it loses outright ..... There are no miracles and fairy tales))))

I spent about 4 days on tests, but I could not achieve at least an equal result with Gminer. I tested all the cores and xintensity from 1124 to 9000. I installed different drivers cuda 11.5 -11.4, but there was no result. Only once I managed to get an increase on the pool, but it lasted for an hour, I think it was just luck, then Gminer went ahead by about 30-40 shares per hour. Maybe on AMD it gives an increase, but on NVIDIA it is useless.

I don't know if anyone else here is running a 3090 / 2060S / AMD 6-series type combo (or a non-lhr Nvidia AMD mix), but it seems to me that in linux (HiveOS ubuntu variant anyway), that the CUDA 1.15 driver (495.46) _REALLY_ nerfs performance, and I'm wondering if anyone is seeing that as well.

With my 3090 under TBM 1.42 for CUDA 1.14 I am seeing 145 - 160 MH/s, but with 1.42 CUDA 1.15 I am seeing 128 - 133 MH/s.
With my 2060S under 1.14 I see 43 - 45 MH/s, but with 1.15 the best I saw was 37 MH/s.


Now some notes on that 3090; it's running in a semi-insulated environment and with the temperatures being about -18 degrees Celsius right now, I can clock the memory up to 11375 without any crashes, and the GPU is running at about 23 degrees (memory is probably in the mid 40's but I'm guessing based on previous ratios I have from windows readings since linux doesn't provide memory temps on Nvidia cards), so I am able to push this beyond the normal speed just due to it being damned cold.

So, does anyone else have this experience with the 1.15 CUDA implementation in the 495.46 Linux drivers?

TBM 1.42 @1.14 CUDA
https://imgur.com/a/EEFk3Rf

TBM 1.42 @1.15 CUDA
https://i.imgur.com/xddqbko.png

Thank you so much! Testing it right now!

---

Not bad, but not many shares  

On ezil good like 2 miners

No he posted in a public thread on reddit. Memclock higher. I think he had modified the cooler on the memory so that it could be clocked higher. I am unable to find the post now. A few months back. But TBM works better on 2miners than on ezil.me.
On 2miners you can use --xintensity 3000 to 5000 without stale shares or rejects. (every share payed).

I have bad eanglish , it's a secret? I want use you miner,  but don't understand,  core clock musbe low or higher? Now I use 1440 on hive , on g miner I use 1100

I run 2 hours , then restart , and kernel now 1 .

The record result was posted by a reddit user, the clocks and settings are hidden in one of the Team Black Miner threads there. Pro miners rarely share their settings because they are the only one who will survive if the price continue to fall.

What settings do u use , core clock and memory ? And kernel ?

Low kernel number is usually better if you run on low power settings, or if the card trottle. The autotune is running when the card is cold. You can let the rig run for 30mins, and restart the miner. Does it still end up on 1? Only way to do a better accurate measure is to let it run for a longer period of time.

The RTX 3080 can do up to 111MHASH poolside on the 2miners pool on high --xintensity (100% mem usage / copyspeed

Hi why on auto tuning  I always have cuda kern 1 ? on my 3080 no lhr rig .Maby I use bad settings ? 1440 fix core , 2450  memory , pl 240 . What settings I must use ? Hive os cuda 11.5 , tbm 1.42 xint 284 pool ezil me . maby you say what the best kernel for this card , and same like 3070 3060ti no lhr ? Kernel 1 is the best ? Or 8 the best ? For best result I must use high core or low , like 1100 ?  On 3070-3080 ?
https://i.postimg.cc/dLHcJG3x/Screenshot-20220119-231607-Chrome.jpg

Normal for this card, but your hashrate is low. You should get around 47MHASH. My rtx 2060 laptop is at 88% utilization.

Perhaps run with  --lock-cclock [1250,1250] or similar can help.

Try 1050 core, and 8100 memory.

So I've been running various tests in HiveOS with different versions of the miner and different video drivers for the Nvidia cards (I messed around with the AMD drivers but keep on killing the OS, so I'm just going to leave that alone for now).

One thing I have noticed with the 1.42 Cuda115 release with the 495.46 Nvidia drivers (I'll check with 11.4 drivers later, but I don't recall seeing this), is that my 2060Super is only at  85-86% memory utilization (not the amount used but the time it's being used), while the 3090 is moving along at a solid 100%.

Here's a snippet of the test output:

root@local:~# nvidia-info
 
Wed Jan 19 12:08:16 EST 2022

=== GPU 0, 37:00.0 GeForce RTX 3090 24268 MB ===
  Bios 94.02.42.80.27, PCIE Link Gen 2, PCIE Link Width 1x
  S/N -, UUID GPU-e3c74e57-e377-7cda-4fd7-71caa2700bf2
  Power Limit:  Min 100.0 W, Default 420.0 W, Max 500.0 W, Current 500.0 W
  Frequency:  GFXCore 1590 MHz, MEMCore 11214 MHz
  Memory:  Total 24268 MB, Used 5025 MB, Free 19242 MB, Micron GDDR6X
  Utilization:  GPU 100 %, MEM 100 %, Throttle: None
  PSTATE P0, PWR 335.0 W, Temp 38 °C, Fan 89 %, RPM 2523, RPM 2536

=== GPU 1, 3D:00.0 GeForce RTX 2060 SUPER 7982 MB ===
  Bios 90.06.44.80.41, PCIE Link Gen 3, PCIE Link Width 4x
  S/N -, UUID GPU-24bf0276-6c6c-87d1-7922-8927a3d0b845
  Power Limit:  Min 125.0 W, Default 175.0 W, Max 215.0 W, Current 215.0 W
  Frequency:  GFXCore 1350 MHz, MEMCore 8062 MHz
  Memory:  Total 7982 MB, Used 4860 MB, Free 3122 MB, Samsung GDDR6
  Utilization:  GPU 100 %, MEM 86 %, Throttle: None
  PSTATE P0, PWR 97.9 W, Temp 35 °C, Fan 80 %, RPM 3040, RPM 3040

root@local:~# nvidia-smi stats -i 1 -d memUtil
1, memUtil , 1642612102256629, 86
1, memUtil , 1642612103259833, 86
1, memUtil , 1642612104263012, 85
1, memUtil , 1642612105266008, 85
1, memUtil , 1642612106270243, 85
1, memUtil , 1642612106270393, 85
1, memUtil , 1642612107273472, 86
1, memUtil , 1642612108276433, 86
1, memUtil , 1642612109280346, 86
1, memUtil , 1642612110283657, 86
1, memUtil , 1642612111286904, 85
1, memUtil , 1642612112289849, 86

root@local:~# nvidia-smi stats -i 0 -d memUtil

0, memUtil , 1642612118885645, 100
0, memUtil , 1642612119091821, 100
0, memUtil , 1642612119298340, 100
0, memUtil , 1642612119504453, 100
0, memUtil , 1642612119710978, 100
0, memUtil , 1642612119917647, 100
0, memUtil , 1642612120123717, 100
0, memUtil , 1642612120329955, 100
0, memUtil , 1642612120536173, 100
0, memUtil , 1642612120742582, 100
0, memUtil , 1642612120948902, 100
0, memUtil , 1642612121155221, 100
0, memUtil , 1642612121361320, 100
0, memUtil , 1642612121567466, 100
0, memUtil , 1642612121773858, 100

The 2060S is hashing at 40 MH/s but the client is reporting 44.75 (after 1 hour, so luck might be heavily influencing the result); do you know of any reason why utilization would be so low for the 2060S?

I included a cuda 11.4 build in v1.42. It's abit slower than 11.5 on the latest driver, but perhaps work good on your driver. Stales can be removed by reducing the default --xintensity

How to configure a windows rig to autoboot in 5 minutes:

1. Copy a shortcut of the bat file to the desktop and move it to shell:startup
2. Run netplwiz to set a username/password for automatic login
3. Use the latest AMD drivers and setup fast memory timings. (custom menu)
4. Setup MSI afterburner to remember the clock setting on startup.
5. Disable the AMD watchdog service
6. Disable fast startup mode in windows.

The Bios should also be setup with Power setting to automaticly startup on powerloss

Thanks!  I dropped my cards memory down to 1066 (2132 Windows), and it's been running for 10 hours now, so it looks like the memory was the cause of the failures, despite the Memory junction temps being between 43 and 48 degrees (now the cards are running at between 40 and 45 degrees).  I'll experiment with different values for Memory speeds and intensities and switch over to 1.42 and try the same to see if there's more stability to be found at higher Memory clock speeds.

BTW, I'm very impressed with how my troublesome 6900XT is running in windows.  12 hours solid and averaging 67.2 MH/s on 2Miners. (334/1/159 @ 12:00:00 runtime).  normally this card runs 1-2 MH/s slower than the 6800XT cards, it's just too bad that windows doesn't have a very good automation tool for maintaining the miner (like if the system shuts down / reboots having the miner run as a service and re-start), or a driver that supports low-level changes to voltages and speed of the chips other than just the GPU core.

Thanks again, this is all great work!

To reduce stales you need to reduce the --xintensity or change the miningpool  . Try build v1.42 with --xintensity 48? There where a couple of timeout bugs in the previous version that can cause the miner to reconnect, recreate dag and loose hashrate.


We have many users now and they are good in finding bugs. The Team Black Miner is a new miner with only a few months since the first public release.

Here is a snapshot from our git.

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

v1.42

1. Fixed gpu timeouts on AMD rigs.

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

TeamBlackMiner_1_42_cuda_11_5.7z
https://www.virustotal.com/gui/file/863dbba4dc5812801de7c8e024c91f0ccd2ee6308051f625d77e9b1cbf639912?nocache=1

TeamBlackMiner_1_42_cuda_11_5.7z
https://www.virustotal.com/gui/file/863dbba4dc5812801de7c8e024c91f0ccd2ee6308051f625d77e9b1cbf639912?nocache=1

TeamBlackMiner_1_42_Ubuntu_18_04_Cuda_11_4.tar.gz
https://www.virustotal.com/gui/file/bc1a737f0a19312b8f6a7d7d8f3e536fcb0a260d5056cb37d7428b8866f2bb69?nocache=1

TeamBlackMiner_1_42_Ubuntu_18_04_Cuda_11_5.tar.gz
https://www.virustotal.com/gui/file/9bf3af2aea444445e8c2c4acee119a5e3ea108c5b9522208e995c865fa7a6a87?nocache=1