Also, under Share (256/0/0), what do mean the 3 values ? Stales is before rejected ?
accepted/rejected / (stale estimate) 226/0/0 (100% accepted) The mining pools doesn't report the staleshare count, so the estimate might be different from the pool stalerate. |
|
|
|
 2x 8 pin mutex can deliver 300 WATT, this cable have to be special, or it will melt. seems to cost around $4000 |
|
|
|
I was always under the impression that it was best to ensure your miner submits a higher percentage of Accepted Shares vs Stale Shares?
Is that not the case anymore?
Depends on the pool, but less than 5% stales doesn't seem to affect the profit on ethereum/ethereum classic mining. For bitcoin and other coins stales are bad. Increasing the xintensity will increase the hashrate, increase the amount of stale shares and also the power. |
|
|
|
1004m [2021-10-05 15:56:21.147] GPU2 3060 Cuda 469.95 kH/W 50.28 MH/s 50.27 MH/s 223/0/0 (100.00)
The miner will try to estimate the stale shares, but you should always check on the pool as well if they any detect stales. Or change to a pool which doesn't care about stales. accepted/rejected/stale 223/0/0 (100% accepted)
Can you give any example how to use -c --configfile?
https://github.com/sp-hash/TeamBlackMiner#example-configuration-file[moderator's note: consecutive posts merged] |
|
|
|
Depends on the pool. If the pool pay for stale shares the default 256 is fine. Flexpool, ethermine( ? ), nicehash and a couple of other doesn't pay for stale shares, so the miner need a lower --xintensity. Usually when a pool display the stale share amount, they eigther pay nothing or a reduced amount for stale shares. Sometimes the stale payout percentage is documented in the FAQ section of the pool. On flexpool if you see this:  It meens that do don't get payed for around 5% of the shares that you submit (stales), and you have to lower the intensity of the miner. Miningpoolhub.com, 2miners.com, nanopool.org, woolypool.com and many others pay the full amount for stale shares. The ethereum blockchain is paying for stale work. They are called uncle blocks, and this is different from other chains like Bitcoin, Doge, Dash etc. A mining pool can be configured to submit uncle blocks and not credit you for the work. Stealing 5% of your profit. |
|
|
|
Different GPUs
--xintensity 50,1,-1 <--- is it?
We don't support tuning each card like this, but it can be added later. 50 and 1 is too low values for v1.11. On amd cards use without --xintensity. We are working on an improved kernel, but it will take some time the other lines are not needed
With the updated TBM v1.11 and new --xintensity recommended setting (--xintensity 224), I am getting 64+ MH per card (Total 322.14 MH) and with 0 stales/rejects so far (93+ mins running).  Looking good. The default --xintesity on nvidia is 256 in v1.11 but 224 produce less stale shares, so we might change the default in v1.12. [moderator's note: consecutive posts merged] |
|
|
|
I'm using 1.09 with --xintensity 56, 0 stales with this value. Too much stales with -1 (i know that some pools pay for stales...).
Of course removing --xintensity for 1.11 did not help with this situation. I came to the conclusion, that 1.10 and newer need to be used with core-lock or you will have lower hash rate compared to 1.09 or older versions.
in v1.10 and v1.11 you need a higher xintensity than 56. try to multiply by 4 --xintensity 224 . It also helps to add more power. Did a quick test on nicehash.com with the v1.11 cuda 11.4 buiild --xintensity 224 on the rtx 2060 mobile. 100 accepted, no rejected and no stales. The hashrate in v1.11 is the highest of all 11 releases. Gminer and t-rex is doing around 30.5 on the same machine / clocks. 
From a reddit user: 1 RTX 3090 With 24hours hashrate confirmed on the 2miners pool.  [moderator's note: consecutive posts merged] |
|
|
|
|
In v1.11 Remove --xintensity -1 in the batfile on nvidia. run with default
|
|
|
|
|
You have an old version of the NVIDIA-SMI tool 461.40. So you need to download the latest nvidia driver, and reinstall everything with the clean option.
|
|
|
|
|
Reinstall the driver with clean option.
whats the output of
nvidia-smi in the command line?
|
|
|
|
0m [2021-10-03 21:31:55.685] GPU0 0GB added to miner as Cuda device
The program is unable to read the memory size of your card. Can you try with -U [9] ? (NVIDIA GeForce GTX 1060 6GB) |
|
|
|
|
remove a - in -U
TBMiner --algo ethash -U [4] --no-ansi --xintensity 1 --hostname eth.pool.zet-tech.eu --port 8005 --wallet 0x8a1E0bE188865f239bC9866d08ae62d6Fa2b181D --worker_name DESKTOP-820EGVB
works here
|
|
|
|
i add -Y 3 but nothing change... driver is 471.96.. currently i use gminer with no problem
You need to use brackets. and the option was U -U [3] -U [0,1,2,3] and --no-ansi
try with --xintensity 1 and remove --cuda-devices 1 [moderator's note: consecutive posts merged] |
|
|
|
|
Does one card work?
-Y [9]
I see that PCIe bus id: always is 0 did you install the latest driver?
|
|
|
|
win10!
You probobly need to add 64gb virtual memory. Can you run TBMiner --list-devices ? perhaps gpu0 is internal graphics. Users report a bug in the stats with internal graphics, so you can try with --no-stats |
|
|
|
|
windows 7 or 8.1? you need windows 10 or linux to run this miner
|
|
|
|
|
Try the cuda v1.10 11.4 build with --xintensity 256
|
|
|
|
|
core :1950, mem 4400. 90%% tdp 27mhash epoch 444
v1.10
|
|
|
|
|
Show Posts
