SRBMiner-MULTI GPU & CPU Miner 0.9.4 - page 57.

KolordZ

newbie

Activity: 28

Merit: 0

Hi doktor83,

I am doing a lot of tests on CPU mining with RandomX algo.
I am using version 0.1.6 on a Fujitsu Primergy RX2560 M1 server.
I have two Intel XEON E5-2630 V3 @ 2.40GHZ 16 cores, 32 threads available

I would like to share my results with you and discuss about hyper-threading option.

1st test (WITH hyper-threading option enabled in BIOS, default option) :

- 16 cores, 32 threads available
SRB uses only 16 threads with auto config with a speed of 4400 Hr / s

1 in 2 thread is not used when analyzing processors activity.
I tried to play with the CPU Affinity and CPU Thread options but could not get better than 4400. If I force 32 threads (or 24, or 30) with the CPU threads option I get a ridiculous speed .

2nd test (WITHOUT hyper-threading option activated in the BIOS) :

- 16 cores, 16 threads available.
SRB uses 15 threads in auto setup. The speed increases a bit, 4900Hr / s

So I would like to know if you knew about it and if you intended to add in your code the hyper threading. If you do not plan to work on it maybe you could warn people not to use the hyper threading option to have a better Hr / S.
I think if hyper-threading option can be manage in SRB we can get maybe 1.6x more speed, what do you think ?

Thank you for your feedback on the subject

Nico

dragonmike

hero member

Activity: 1274

Merit: 556

Quote from: doktor83 on November 18, 2019, 11:19:15 AM

Quote from: dragonmike on November 18, 2019, 10:14:27 AM

Any hashrate indics for k12 on commonly used GPUs?

auto intensity settings, memory is not important for K12 i just benchmarked for you :

Rx560 [14 cu] 1170/1500 : ~340MHS
Vega56 [56 cu] 1450/940 : ~1700MHS

Oh dear. That one's gone to the ASICs/FPGAs it seems.
$0.15/day per Vega... Tongue

doktor83

hero member

Activity: 2548

Merit: 626

Quote from: dragonmike on November 18, 2019, 10:14:27 AM

Any hashrate indics for k12 on commonly used GPUs?

auto intensity settings, memory is not important for K12 i just benchmarked for you :

Rx560 [14 cu] 1170/1500 : ~340MHS
Vega56 [56 cu] 1450/940 : ~1700MHS

dragonmike

hero member

Activity: 1274

Merit: 556

Any hashrate indics for k12 on commonly used GPUs?

doktor83

hero member

Activity: 2548

Merit: 626

V0.1.7 beta

+ Added 'm7mv2' algorithm (Magi coin - XMG)
+ Cpu auto configuration should be better than on previous versions
+ Added CPU data to API, also rearranged a few things in API
+ Console window 'Quick Edit' mode disabled on miner start, so users won't accidentally pause miner by clicking in the console window

doktor83

hero member

Activity: 2548

Merit: 626

Quote from: KolordZ on November 13, 2019, 04:00:24 PM

Hello @doktor83 !

This is my test... What a poor result as an i7-8750H :/

Code:

[2019-11-13 21:49:19] Startup monitor attached
[2019-11-13 21:49:19] Miner version: 0.1.6
[2019-11-13 21:49:19] Windows version: 10.0 build 18362
[2019-11-13 21:49:20] CPU: Intel(R)Core(TM ) i7-8750H CPU @ 2.20GHz [L3:9216 KB][L2:256 KB][L1:32 KB]
[2019-11-13 21:49:20] Algorithm: randomx
[2019-11-13 21:49:20] Gpu mining: disabled
[2019-11-13 21:49:20] Cpu mining: enabled
[2019-11-13 21:49:20] Huge-pages: enabled
[2019-11-13 21:49:20] HW-Aes: available
[2019-11-13 21:49:20] Startup monitor: enabled
[2019-11-13 21:49:20] Socket init done
[2019-11-13 21:49:20] Starting init of worker threads
[2019-11-13 21:49:20] CPU worker thread 0 bound to PU 0 on node 0
[2019-11-13 21:49:20] CPU worker thread 1 bound to PU 2 on node 0
[2019-11-13 21:49:20] CPU worker thread 2 bound to PU 4 on node 0
[2019-11-13 21:49:20] CPU worker thread 3 bound to PU 6 on node 0
[2019-11-13 21:49:20] CPU worker thread 4 bound to PU 8 on node 0
[2019-11-13 21:49:20] Thread[0] : CPU
[2019-11-13 21:49:20] Thread[1] : CPU
[2019-11-13 21:49:20] Thread[2] : CPU
[2019-11-13 21:49:20] Thread[3] : CPU
[2019-11-13 21:49:20] Thread[4] : CPU
[2019-11-13 21:49:20] Thread[5] : CPU
[2019-11-13 21:49:20] Finished init of worker threads
[2019-11-13 21:49:20] CPU worker thread 5 bound to PU 10 on node 0
[2019-11-13 21:49:20] Startup monitor detached
[2019-11-13 21:49:20] 6 CPU core/s - 12 threads available
[2019-11-13 21:49:20] Using 6 threads for CPU mining
[2019-11-13 21:49:20] Node 0 total memory: 5626 MB
[2019-11-13 21:49:20] CPU extension : SSE4.1 | AES
[2019-11-13 21:49:20] Gpu-watchdog started [triggers after 5 rounds: 150 sec]
[2019-11-13 21:49:20] HTTP API disabled
[2019-11-13 21:49:20] json_send[201]: {"id":1,"jsonrpc": "2.0","method":"login","params":{"login":"4897Wc5pexE25twCvQ1bpyBEyMZCrdeARSFqfYQBtPfSD6FZrCW6L6zfAB6PnhJJdaVi2Q3r1AgcViBPrMoZGqRj6hhKLUC","pass":"x","agent":"SRBMiner-MULTI/0.1.6"}}
[2019-11-13 21:49:20] json_receive[412]: {"id":1,"jsonrpc":"2.0","error":null,"result":{"id":"500772239118811","job":{"seed_hash":"3b901ea2f0f825a5d3cf7bc0a7e0f7faa8d6efc157c36256b3cc3398f3ee89c7","height":1341994,"blob":"0c0cf4d6b1ee05a4adb049ba70b7cc094b3e07b13933374c8bae270a141b22768c64b5f9ee605100000000dcaf44f9d65426de835141c45f46f13c10b6ed26e385467f9b6f435b34dbb72501","job_id":"899071441393928","target":"b2df0000","algo":"rx/0"},"status":"OK"}}
[2019-11-13 21:49:20] Connected to rx.minexmr.com:4444
[2019-11-13 21:49:20] Initialising dataset & cache & vm
[2019-11-13 21:49:21] memory_manager: Allocated memory with huge page support [2129920 kB]
[2019-11-13 21:49:21] memory_manager: Allocated memory with huge page support [262144 kB]
[2019-11-13 21:49:26] Dataset & Cache & VM initialized in 5649 ms
[2019-11-13 21:49:26] Worker thread 0 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 0
[2019-11-13 21:49:26] Worker thread 1 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 1
[2019-11-13 21:49:26] Worker thread 2 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 2
[2019-11-13 21:49:26] Worker thread 3 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 3
[2019-11-13 21:49:26] Worker thread 4 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 4
[2019-11-13 21:49:26] Worker thread 5 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 5
[2019-11-13 21:49:26] New job received, block height 1341994
[2019-11-13 21:49:58] hashrate: CPU: 995.50 H/s
[2019-11-13 21:49:58] hashrate: Total: 995.50 H/s

I used your guided-setup", SRBMiner MULTI 0.1.6

I missed something ?

Thx for what you've done !

Hi, you could manually set the number of threads and the affinity, maybe it can get out more hash.
New version in a few days, it should have a better auto setup.

KolordZ

newbie

Activity: 28

Merit: 0

Hello @doktor83 !

This is my test... What a poor result as an i7-8750H :/

Code:

[2019-11-13 21:49:19] Startup monitor attached
[2019-11-13 21:49:19] Miner version: 0.1.6
[2019-11-13 21:49:19] Windows version: 10.0 build 18362
[2019-11-13 21:49:20] CPU: Intel(R)Core(TM ) i7-8750H CPU @ 2.20GHz [L3:9216 KB][L2:256 KB][L1:32 KB]
[2019-11-13 21:49:20] Algorithm: randomx
[2019-11-13 21:49:20] Gpu mining: disabled
[2019-11-13 21:49:20] Cpu mining: enabled
[2019-11-13 21:49:20] Huge-pages: enabled
[2019-11-13 21:49:20] HW-Aes: available
[2019-11-13 21:49:20] Startup monitor: enabled
[2019-11-13 21:49:20] Socket init done
[2019-11-13 21:49:20] Starting init of worker threads
[2019-11-13 21:49:20] CPU worker thread 0 bound to PU 0 on node 0
[2019-11-13 21:49:20] CPU worker thread 1 bound to PU 2 on node 0
[2019-11-13 21:49:20] CPU worker thread 2 bound to PU 4 on node 0
[2019-11-13 21:49:20] CPU worker thread 3 bound to PU 6 on node 0
[2019-11-13 21:49:20] CPU worker thread 4 bound to PU 8 on node 0
[2019-11-13 21:49:20] Thread[0] : CPU
[2019-11-13 21:49:20] Thread[1] : CPU
[2019-11-13 21:49:20] Thread[2] : CPU
[2019-11-13 21:49:20] Thread[3] : CPU
[2019-11-13 21:49:20] Thread[4] : CPU
[2019-11-13 21:49:20] Thread[5] : CPU
[2019-11-13 21:49:20] Finished init of worker threads
[2019-11-13 21:49:20] CPU worker thread 5 bound to PU 10 on node 0
[2019-11-13 21:49:20] Startup monitor detached
[2019-11-13 21:49:20] 6 CPU core/s - 12 threads available
[2019-11-13 21:49:20] Using 6 threads for CPU mining
[2019-11-13 21:49:20] Node 0 total memory: 5626 MB
[2019-11-13 21:49:20] CPU extension : SSE4.1 | AES
[2019-11-13 21:49:20] Gpu-watchdog started [triggers after 5 rounds: 150 sec]
[2019-11-13 21:49:20] HTTP API disabled
[2019-11-13 21:49:20] json_send[201]: {"id":1,"jsonrpc": "2.0","method":"login","params":{"login":"4897Wc5pexE25twCvQ1bpyBEyMZCrdeARSFqfYQBtPfSD6FZrCW6L6zfAB6PnhJJdaVi2Q3r1AgcViBPrMoZGqRj6hhKLUC","pass":"x","agent":"SRBMiner-MULTI/0.1.6"}}
[2019-11-13 21:49:20] json_receive[412]: {"id":1,"jsonrpc":"2.0","error":null,"result":{"id":"500772239118811","job":{"seed_hash":"3b901ea2f0f825a5d3cf7bc0a7e0f7faa8d6efc157c36256b3cc3398f3ee89c7","height":1341994,"blob":"0c0cf4d6b1ee05a4adb049ba70b7cc094b3e07b13933374c8bae270a141b22768c64b5f9ee605100000000dcaf44f9d65426de835141c45f46f13c10b6ed26e385467f9b6f435b34dbb72501","job_id":"899071441393928","target":"b2df0000","algo":"rx/0"},"status":"OK"}}
[2019-11-13 21:49:20] Connected to rx.minexmr.com:4444
[2019-11-13 21:49:20] Initialising dataset & cache & vm
[2019-11-13 21:49:21] memory_manager: Allocated memory with huge page support [2129920 kB]
[2019-11-13 21:49:21] memory_manager: Allocated memory with huge page support [262144 kB]
[2019-11-13 21:49:26] Dataset & Cache & VM initialized in 5649 ms
[2019-11-13 21:49:26] Worker thread 0 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 0
[2019-11-13 21:49:26] Worker thread 1 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 1
[2019-11-13 21:49:26] Worker thread 2 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 2
[2019-11-13 21:49:26] Worker thread 3 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 3
[2019-11-13 21:49:26] Worker thread 4 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 4
[2019-11-13 21:49:26] Worker thread 5 allocated L3 scratchpad [HP]
[2019-11-13 21:49:26] Created VM for worker thread 5
[2019-11-13 21:49:26] New job received, block height 1341994
[2019-11-13 21:49:58] hashrate: CPU: 995.50 H/s
[2019-11-13 21:49:58] hashrate: Total: 995.50 H/s

I used your guided-setup", SRBMiner MULTI 0.1.6

I missed something ?

Thx for what you've done !

doktor83

hero member

Activity: 2548

Merit: 626

Still need to improve the auto selection of best config for randomx Angry

dragonmike

hero member

Activity: 1274

Merit: 556

RX Vega 56@64, core clk @ 1500mhz

CKB: ~ 673 MH/s

doktor83

hero member

Activity: 2548

Merit: 626

Quote from: dragonmike on November 12, 2019, 10:03:58 AM

This thread lacks highly important hashrate reports!...

Come on fellas, share your findings.

i5-2500K CPU @ 3.30GHz @ 3 threads

RandomX: ~ 1490 h/s
RandomXL: ~ 1520 h/s
RandomWow: ~ 1815 h/s
RandomArq: ~ 7620 h/s

Celeron(R) CPU J1800 @ 2.41GHz @ 1 thread Grin

RandomXL: ~ 63 h/s
RandomWow: ~ 88 h/s

doktor83

hero member

Activity: 2548

Merit: 626

I fixed & re-uploaded, download again from github please :

https://github.com/doktor83/SRBMiner-Multi/releases/tag/0.1.6

if you set the number of threads yourself, you should also set the coresponding affinity.

doktor83

hero member

Activity: 2548

Merit: 626

Quote from: ?? on ??

After the add ---cpu-affinity 0x00003FFFFFFFFFFF closes with such an error

terminate called after throwing an instance of 'std::out_of_range'
what(): stoul
SEGFAULT ?

Shutting down miner...

Delete ---cpu-affinity 0x00003FFFFFFFFFFF, works on 24 threads

Ouch, you found a bug, thanks

dragonmike

hero member

Activity: 1274

Merit: 556

This thread lacks highly important hashrate reports!...

Come on fellas, share your findings.

doktor83

hero member

Activity: 2548

Merit: 626

Quote from: ?? on ??

Hello
Yes it's better but I have x4 Opteron 6234 48 threads.
When I load on 46 threads it can load only 24, I can use the utility to load on 46 but it should be done manually, or 2 miners on 24 threads.

[2019-11-12 16:56:18] OpenCL not found, gpu mining disabled
[2019-11-12 16:56:18] Startup monitor attached
[2019-11-12 16:56:18] Miner version: 0.1.6
[2019-11-12 16:56:18] Windows version: 10.0 build 17763
[2019-11-12 16:56:18] NUMA binding: enabled
[2019-11-12 16:56:18] CPU x 4: AMDOpteron(TM) Processor 6234 [L3:6144 KB][L2:2048 KB][L1:16 KB]
[2019-11-12 16:56:18] Algorithm: randomarq
[2019-11-12 16:56:18] Gpu mining: disabled
[2019-11-12 16:56:18] Cpu mining: enabled
[2019-11-12 16:56:18] Huge-pages: enabled
[2019-11-12 16:56:18] HW-Aes: available
[2019-11-12 16:56:18] Startup monitor: enabled
[2019-11-12 16:56:18] Starting init of worker threads
[2019-11-12 16:56:18] Finished init of worker threads
[2019-11-12 16:56:19] Startup monitor detached
[2019-11-12 16:56:19] 24 CPU core/s - 48 threads available
[2019-11-12 16:56:19] Using 46 threads for CPU mining
[2019-11-12 16:56:19] CPU extension : SSE4.1 | AES
[2019-11-12 16:56:19] Gpu-watchdog started [triggers after 5 rounds: 150 sec]
[2019-11-12 16:56:19] HTTP API disabled
[2019-11-12 16:56:19] Connected to arqma.herominers.com:10641
[2019-11-12 16:56:19] Initialising datasets & caches & vms
[2019-11-12 16:56:34] miner_result: CPU result 0xa5f5a6b7 accepted [44ms]
[2019-11-12 16:56:37] miner_result: CPU result 0x5cb79005 accepted [45ms]
[2019-11-12 16:56:38] miner_result: CPU result 0x8b5fc8c2 accepted [49ms]
[2019-11-12 16:56:41] miner_result: CPU result 0xac85d39b accepted [43ms]

Notice I'm talking 46 threads, and he's 24 threads.
Thank you

Try adding :

for 48 threads

--cpu-threads 48 --cpu-affinity 0x0000FFFFFFFFFFFF

for 48 threads

--cpu-threads 46 --cpu-affinity 0x00003FFFFFFFFFFF

doktor83

hero member

Activity: 2548

Merit: 626

V0.1.6 beta

+ Fixed threads binding on systems with more than 1 numa node
+ Better auto configuration for cpu mining
+ Previous two fixes should bring higher hashrates
+ Added pool controlled algorithm switching capability (Monero Ocean)
+ Added pool config parameter "algo_min_time" and --algo-min-time to cli (Monero Ocean)
+ Added pool config parameter "keepalive" and --keepalive to cli

tallman_ay

newbie

Activity: 5

Merit: 0

R u going to update your miner with the new randomx for GPUs

Marvell2

full member

Activity: 1148

Merit: 132

also can you ad sigma for gulden cpu algo @Doctur

Marvell2

full member

Activity: 1148

Merit: 132

whata the performance on eaglesong on rx cards

EnglishGentleman

jr. member

Activity: 33

Merit: 1

Quote from: doktor83 on November 07, 2019, 12:25:37 PM

Quote from: EnglishGentleman on November 07, 2019, 12:06:41 PM

Hello Doktor

Thanks for the initiative. Single core performance on RandomXL seems good. However there are some error messages thrown up and multi-core performance is poor at the moment. Can you help? The test system is Dell R620 server with two Xeon E5-2620 V1 and 28 GBytes of RAM.

First the error messages:

Cannot bind database to node Cannot bind cache to node See https://drive.google.com/file/d/1o8AnMtIU3J9WstytBRktOa-L9f_C0BeN/view?usp=sharing

Cannot bind scratchpad to node See https://drive.google.com/file/d/18G_bTBlkV3hUXC-D9X0sMrui1dSmjkFe/view?usp=sharing

The errors appear if I use the wizard to create the BAT file or make my own BAT file.

How can these errors be fixed please?

The performance issue seems to be that the second CPU is not utilised even though SRBMiner identifies 12 cores and 24 threads.

See pics from HWInfo of the two cores

CPU 0 https://drive.google.com/file/d/1IhRbu9z1F1-DWaJpz7o3dyQvazSOe3VM/view?usp=sharing
CPU 1 https://drive.google.com/file/d/1C2s1PBm9hmzNAv91Jrsk2ZiWGlvW14Va/view?usp=sharing

Here is the command line used SRBMiner-MULTI.exe --algorithm randomxl --cpu-threads 0 --pool loki.miner.rocks:5005 --wallet L7f9SMpr43wNERBxCngFyqMoJvkMax8xSitiJY335ra12ucbiEFzgqKfHJg1Q9y58RGMi3TyGvPC94e 6w3zYKK9i9PHmDDr --cpu-threads

Finally I can't see documentation on how to use CPU affinity and --cpu-priority. Please can you add it?

Hi, first try removing the --cpu-threads 0 because that sets the number of threads to 1 Grin

Code:

SRBMiner-MULTI.exe --algorithm randomxl --pool loki.miner.rocks:5005 --wallet L7f9SMpr43wNERBxCngFyqMoJvkMax8xSitiJY335ra12ucbiEFzgqKfHJg1Q9y58RGMi3TyGvPC94e6w3zYKK9i9PHmDDr

You can also try to find the best --cpu-threads by hand for your system.

--cpu-priority is just a number from 1-5, the higher the priority level of a process, the more will be its assigned processor time.
--cpu-affinity is a mask (can be used hex like 0x.... or just the decimal number). An affinity mask is a bit mask indicating what processor(s) a thread or process should be run on by the scheduler of an operating system.

Hello Doktor

I tried SRMMiner 0.1.5 on a 2 core Celeron machine. It worked first time on the RX.MINEXMR RandomX TestNet. No errors. I followed your instructions for enabling huge pages. That also worked. Performance was 7% higher than XMR-RIG. Great.

Thanks for the feedback about the Dell R620 server. I started a fresh with your guided-setup.bat and used the RandomX testnet at RX.MINEXMR.

Unfortunately I got exactly the error codes as the first time + the second CPU is not being utilised.

I was able to utilise the second CPU by adding --cpu-affinity 0x555555 . That problem is solved. . Since the miner recognises all the threads and core automatically would it be possible to automatically set the miner up to use all CPUs?

Mining performance of SRBMiner 0-1-5 on RandomX at rx.minexmr.com was initially 92% of the hashrate achieved with XMRIG 4.5.0 Beta but when I ran it overnight hashrate exceeded XMRIG by 5%. Not bad for your initial release. I tried adding additional threads but it made no difference. Whereas XMRIG actually goes slower on these CPU when more than 1 thread per core is used.

The Celeron and the Dell R620 run Windows 10 Professional.

How about the binding errors? Do they matter? Perhaps additional performance could be achieved without these errors?

See links below. Thanks in advance for your help.

Screenshots of errors: Cannot bind to XXX errors
https://drive.google.com/file/d/1Mo6NM_wMCjEEB1nyF6Z3OSuPw2XItVFi/view?usp=sharing

https://drive.google.com/file/d/1XnHKjVxDX7qkFwbjIudWVE_0CyJbtUJo/view?usp=sharing

https://drive.google.com/file/d/1pas-ckr-JiVyvtJfA0HktwfhyE0WTZgX/view?usp=sharing

FOOTNOTE

I also have an R9 Fury on the Celeron machine. When I run SRBMiner 0.1.5 in parallel with Claymore 15 mining ETH I got a small speed improvement of 0.8% on Claymore 15! Claymore reduced the performance of SRBMiner on RandomX by 15% when run in parallel.

CjMapope

legendary

Activity: 1820

Merit: 1092

~Full-Time Minter since 2016~

Quote from: doktor83 on November 07, 2019, 08:02:40 AM

V0.1.5 beta

+ Added cpu algorithms : RandomX, RandomXL, RandomWow, RandomArq
+ Added cli parameters: --cpu-priority, --disable-huge-pages, --disable-hw-aes, --disable-numa
+ Added pool config parameter 'start_block_height' and --start-block-height cli parameter
+ Fixed a bug that limited the number of threads useable for cpu mining
+ Fixed cpu affinity parameter and thread/s binding to affinity mask
+ Where possible, now using huge pages if available, this should bring a slight performance increase
+ Removed OpenCL dependency if using only cpu miner
+ Added limited Overdrive8 support (adl type 3)

oh duuude, fuk ya! going to roll this out to the CPUs and play with it right now, ill report back any issues that arise
im starting to get hyped for RandomX day by day, go'in get me a fancy 3900x, only 700 Beaverdollars, so crazy cheap
Thx for your work

Topic: SRBMiner-MULTI GPU & CPU Miner 0.9.4 - page 57. (Read 45425 times)