SRBMiner-MULTI AMD & NVIDIA & INTEL & CPU miner

[P00P135]

You should be able to boost up to 2050 core safely and test 100-200 offset from there.

[doktor83]

V3.3.9
+ Added algorithm 'neuromorph' (Cereblix coin) for CPU mining, fee 3.0%*
+ Minor hashrate improvement on algorithm 'pearlhash' for all supported GPUs
+ Now really added support for progpow algorithms, 'autolykos2', 'verthash', 'fishhash' and 'xhash' on Nvidia Volta (SM70) GPUs
+ Reduced poolside rejected shares on algorithm 'qhash'
+ Bug fixes

V3.3.8
+ Minor hashrate improvement on algorithm 'pearlhash' for NVIDIA 3000, 4000 and 5000 series GPUs
+ Added support for progpow algorithms, 'autolykos2', 'verthash', 'fishhash' and 'xhash' on Nvidia Volta (SM70) GPUs
+ Added parameter '--pearl-k2' which uses an optional kernel if available for the selected device*
+ Removed 'pearlhash' support for sm80, sm103 - bad performance
+ Bug fixes

V3.3.7
+ Huge hashrate improvement on algorithm 'pearlhash' for NVIDIA B200 GPU's (~1,1PH/S @ 985W)
+ Fixed broken algorithm 'oggpow' on rplant

[JBoy_32]

ooooh boy! Just bought a hand full of rx 9060 xt's not realizing SRB only supports 9070!? Any chance the AMD RX 9060 can get support in update 3.4.0?!

[techobg]

V3.3.9
+ Added algorithm 'neuromorph' (Cereblix coin) for CPU mining, fee 3.0%*
+ Minor hashrate improvement on algorithm 'pearlhash' for all supported GPUs
+ Now really added support for progpow algorithms, 'autolykos2', 'verthash', 'fishhash' and 'xhash' on Nvidia Volta (SM70) GPUs
+ Reduced poolside rejected shares on algorithm 'qhash'
+ Bug fixes

V3.3.8
+ Minor hashrate improvement on algorithm 'pearlhash' for NVIDIA 3000, 4000 and 5000 series GPUs
+ Added support for progpow algorithms, 'autolykos2', 'verthash', 'fishhash' and 'xhash' on Nvidia Volta (SM70) GPUs
+ Added parameter '--pearl-k2' which uses an optional kernel if available for the selected device*
+ Removed 'pearlhash' support for sm80, sm103 - bad performance
+ Bug fixes

V3.3.7
+ Huge hashrate improvement on algorithm 'pearlhash' for NVIDIA B200 GPU's (~1,1PH/S @ 985W)
+ Fixed broken algorithm 'oggpow' on rplant

Hi doktor83,

Thank you for the continuous work on SRBMiner.

I'm mining PearlHash in SOLO mode with two rigs (~500 TH/s total), and I have one feature request that I believe would be very useful for all solo miners.

Would it be possible to display the difficulty of every accepted share, as well as the best share since miner start?

Example:

Accepted share [247 ms]
Share difficulty: 21.84T
Best share: 38.52T

It would also be great if these values were available through the API:

* last_share_difficulty
* best_share

This information would help solo miners evaluate the quality of their shares and compare miner performance, not only hashrate.

Thank you for considering it, and thanks again for the great miner!

[doktor83]

V3.4.0
+ Improved algorithm 'pearlhash' for (almost) all NVIDIA GPUs and lowered devfee to 2%
+ Hashrate improvement on algorithm 'neuromorph' for AVX2 CPUs and lowered devfee to 1%
+ Added new CPU metrics : power usage and efficiency (linux only)
+ Bug fixes

[dalllas]

V3.4.0
+ Improved algorithm 'pearlhash' for (almost) all NVIDIA GPUs and lowered devfee to 2%
+ Hashrate improvement on algorithm 'neuromorph' for AVX2 CPUs and lowered devfee to 1%
+ Added new CPU metrics : power usage and efficiency (linux only)
+ Bug fixes

I am experiencing an issue with Windows miner and Luckypool mining; the miner stops after receiving a job.

[2026-06-25 07:48:26] Hashrate represents the 60 second average value - max hashrate after few minutes
[2026-06-25 07:48:27] GPU0 |A0|pearlhash           |
[2026-06-25 07:48:27] GPU1 |A0|pearlhash           |
[2026-06-25 07:48:29] Connected to pearl-eu1.luckypool.io:3360 [pearlhash]
[2026-06-25 07:48:29] Job received [019bd1aa_888888   ] block height 77857 [diff: 888888] [pearlhash][0]
Press any key to continue . . .

[doktor83]

V3.4.1
+ Fixed 'proof validation failed' error for algorithm 'pearlhash' introduced in previous version [3.4.0]
+ Pearlhash 3060TI and 3080 10GB hashrates improved
+ Bug fixes

[doktor83]

V3.4.0
+ Improved algorithm 'pearlhash' for (almost) all NVIDIA GPUs and lowered devfee to 2%
+ Hashrate improvement on algorithm 'neuromorph' for AVX2 CPUs and lowered devfee to 1%
+ Added new CPU metrics : power usage and efficiency (linux only)
+ Bug fixes

I am experiencing an issue with Windows miner and Luckypool mining; the miner stops after receiving a job.

[2026-06-25 07:48:26] Hashrate represents the 60 second average value - max hashrate after few minutes
[2026-06-25 07:48:27] GPU0 |A0|pearlhash           |
[2026-06-25 07:48:27] GPU1 |A0|pearlhash           |
[2026-06-25 07:48:29] Connected to pearl-eu1.luckypool.io:3360 [pearlhash]
[2026-06-25 07:48:29] Job received [019bd1aa_888888   ] block height 77857 [diff: 888888] [pearlhash][0]
Press any key to continue . . .

if you get DLL issue or miner just crashes install : https://aka.ms/vc14/vc_redist.x64.exe

[dalllas]

if you get DLL issue or miner just crashes install : vc_redist.x64.exe

Thank you. It works perfectly!

[incognito001]

Error CL_INVALID_VALUE when getting pcie topology for device 0 on platform 1


How to solve the error?

[fenomenyaa]

Hi Doc,

I'm experiencing an issue with version 3.4.0 and newer on some of my Windows mining rigs.

The miner starts normally, generates the DAG successfully, but as soon as it connects to the pool, it immediately closes without any error message.

The problem does not affect all of my rigs—only some of them. The same rigs work perfectly with version 3.3.9 and all earlier versions, so the issue seems to have been introduced after 3.3.9.

Do you have any idea what could be causing this, or is there anything I can test to help identify the problem?

Thank you!

[doktor83]

V3.4.2
+ Improved algorithm 'pearlhash' for (almost) all NVIDIA GPUs
+ Added (back) 'pearlhash' support for NVIDIA A100
+ Bug fixes

[doktor83]

Hi Doc,

I'm experiencing an issue with version 3.4.0 and newer on some of my Windows mining rigs.

The miner starts normally, generates the DAG successfully, but as soon as it connects to the pool, it immediately closes without any error message.

The problem does not affect all of my rigs—only some of them. The same rigs work perfectly with version 3.3.9 and all earlier versions, so the issue seems to have been introduced after 3.3.9.

Do you have any idea what could be causing this, or is there anything I can test to help identify the problem?

Thank you!

yeah, have some bugs on windows, some of em are solved in 3.4.2, test it.

[Mordunkus]

Hello and thanks, Doc.

Testing 3.4.2, Windows 11, 4090 with updated driver, Meowcoin (I'm not the OP).  Coincidence that 262 is exactly a quarter of 1048?  Will run forever doing nothing, but doesn't crash:

[2026-06-27 13:40:37] Connected to pool.us.woolypooly.com:3116 [meowpow]
[2026-06-27 13:40:37] Epoch    : 262 [meowpow]
[2026-06-27 13:40:37] Generating cache - please wait ...
[2026-06-27 13:40:42] Difficulty 0.0000000000 [meowpow][0]
[2026-06-27 13:40:42] Job received [0000c043          ] block height 1970885 [meowpow][0]
[2026-06-27 13:40:42] [GPU0] Auto tune started - wait a few minutes after it finishes for best hashrate
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:42] Couldn't set epoch 1048 on GPU0 [meowpow]
[2026-06-27 13:40:51] Job received [0000c044          ] block height 1970885 [meowpow][0]
[2026-06-27 13:41:04] ============================= v3.4.2 =============================
[2026-06-27 13:41:04]

•     [meowpow]
[2026-06-27 13:41:04] GPU0 : 0.00 H/s          [       0|     0|    0|             -]
[2026-06-27 13:41:04] Total: 0.00 H/s          |       0|     0|    0|

[fenomenyaa]

Hi Doc,

I'm experiencing an issue with version 3.4.0 and newer on some of my Windows mining rigs.

The miner starts normally, generates the DAG successfully, but as soon as it connects to the pool, it immediately closes without any error message.

The problem does not affect all of my rigs—only some of them. The same rigs work perfectly with version 3.3.9 and all earlier versions, so the issue seems to have been introduced after 3.3.9.

Do you have any idea what could be causing this, or is there anything I can test to help identify the problem?

Thank you!

yeah, have some bugs on windows, some of em are solved in 3.4.2, test it.

Thank you for the update and for all the work you've been putting into the miner.

Version 3.4.2 resolved most of the compatibility issues on Windows, which is greatly appreciated. However, the NVIDIA A2000 series is still not working. Hopefully this can be addressed in one of the upcoming releases.

Thanks again for your continued efforts and support!

[P00P135]

Hi Doc,

I'm experiencing an issue with version 3.4.0 and newer on some of my Windows mining rigs.

The miner starts normally, generates the DAG successfully, but as soon as it connects to the pool, it immediately closes without any error message.

The problem does not affect all of my rigs—only some of them. The same rigs work perfectly with version 3.3.9 and all earlier versions, so the issue seems to have been introduced after 3.3.9.

Do you have any idea what could be causing this, or is there anything I can test to help identify the problem?

Thank you!

yeah, have some bugs on windows, some of em are solved in 3.4.2, test it.

Thank you for the update and for all the work you've been putting into the miner.

Version 3.4.2 resolved most of the compatibility issues on Windows, which is greatly appreciated. However, the NVIDIA A2000 series is still not working. Hopefully this can be addressed in one of the upcoming releases.

Thanks again for your continued efforts and support!

A2000 work great on hiveOS 32TH 70w

[fenomenyaa]

Hi Doc,

I'm experiencing an issue with version 3.4.0 and newer on some of my Windows mining rigs.

The miner starts normally, generates the DAG successfully, but as soon as it connects to the pool, it immediately closes without any error message.

The problem does not affect all of my rigs—only some of them. The same rigs work perfectly with version 3.3.9 and all earlier versions, so the issue seems to have been introduced after 3.3.9.

Do you have any idea what could be causing this, or is there anything I can test to help identify the problem?

Thank you!

yeah, have some bugs on windows, some of em are solved in 3.4.2, test it.

Thank you for the update and for all the work you've been putting into the miner.

Version 3.4.2 resolved most of the compatibility issues on Windows, which is greatly appreciated. However, the NVIDIA A2000 series is still not working. Hopefully this can be addressed in one of the upcoming releases.

Thanks again for your continued efforts and support!

A2000 work great on hiveOS 32TH 70w

Windows side broken.

[doktor83]

V3.4.4
+ Improved algorithm 'btx' for AMD/NVIDIA GPUs*
+ Improved algorithm 'pearlhash' for NVIDIA 5000 series GPUs
+ Improved algorithm 'pearlhash' for AMD RDNA2 GPUs
+ Minor improvement for 4090 GPU on algorithm 'pearlhash'
+ Added optional 'pearlhash' code for 3070, 3070ti which can achieve higher hashrate (use --pearl-k2)
+ For 'pearlhash' on 3060 12gb the default code is now the one from v3.4.2 (more power efficient) , but setting --pearl-k2 uses code from prev. v.3.4.3

V3.4.3
+ Added algorithm 'btx' (BTX coin) for AMD/NVIDIA GPUs, fee 1.0%*
+ Pearlhash related fixes

***
'btx' can be mined on pools that support 'ninja' stratum (luckypool, ninjaraider ...) [ btx-eu.lproute.com:8660 ]
Website : https://btx.dev

[energy_xxx]

hi doctor

https://github.com/optiscaler/OptiScaler/releases/tag/v0.9.3


Is it possible to integrate FSR/DLSS code into the miner software for AMD cards, specifically for Pearl cards, to run games?

I tried it for SRB, but it didn't work because the miner couldn't find the necessary arguments. It works in games. Is it possible to create a model that calls FSR/DLSS codes?

[angelbbs]

Hi!
And for PearlHash DP4A → PTX IMAD and 16F_PEDANTIC for NVIDIA CMP, please.