AMD EPYC 2x 7551 on a SM H11DSi

[thorsam]

Is V7.5.1 ready for a dual AMD EPYC system?
It seems that the software only find one of the cpu?

Code: Select all

*********************** Log Started 2018-07-30T14:49:08Z ***********************
14:49:08:************************* Folding@home Client *************************
14:49:08:        Website: https://foldingathome.org/
14:49:08:      Copyright: (c) 2009-2018 foldingathome.org
14:49:08:         Author: Joseph Coffland <joseph@cauldrondevelopment.com>
14:49:08:           Args: --open-web-control
14:49:08:         Config: C:\Users\thorsam\AppData\Roaming\FAHClient\config.xml
14:49:08:******************************** Build ********************************
14:49:08:        Version: 7.5.1
14:49:08:           Date: May 11 2018
14:49:08:           Time: 13:06:32
14:49:08:     Repository: Git
14:49:08:       Revision: 4705bf53c635f88b8fe85af7675557e15d491ff0
14:49:08:         Branch: master
14:49:08:       Compiler: Visual C++ 2008
14:49:08:        Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
14:49:08:       Platform: win32 10
14:49:08:           Bits: 32
14:49:08:           Mode: Release
14:49:08:******************************* System ********************************
14:49:08:            CPU: AMD EPYC 7551 32-Core Processor
14:49:08:         CPU ID: AuthenticAMD Family 23 Model 1 Stepping 2
14:49:08:           CPUs: 32
14:49:08:         Memory: 127.91GiB
14:49:08:    Free Memory: 122.25GiB
14:49:08:        Threads: WINDOWS_THREADS
14:49:08:     OS Version: 6.2
14:49:08:    Has Battery: false
14:49:08:     On Battery: false
14:49:08:     UTC Offset: 2
14:49:08:            PID: 11076
14:49:08:            CWD: C:\Users\thorsam\AppData\Roaming\FAHClient
14:49:08:             OS: Windows 10 Enterprise
14:49:08:        OS Arch: AMD64
14:49:08:           GPUs: 1
14:49:08:          GPU 0: Bus:33 Slot:0 Func:0 NVIDIA:3 GK208 [GeForce GT 710 LP]
14:49:08:  CUDA Device 0: Platform:0 Device:0 Bus:33 Slot:0 Compute:3.5 Driver:9.2
14:49:08:OpenCL Device 0: Platform:0 Device:0 Bus:33 Slot:0 Compute:1.2 Driver:398.36
14:49:08:  Win32 Service: false
14:49:08:***********************************************************************
14:49:09:<config>
14:49:09:  <!-- Network -->
14:49:09:  <proxy v=':8080'/>
14:49:09:
14:49:09:  <!-- User Information -->
14:49:09:  <passkey v='********************************'/>
14:49:09:  <team v='34361'/>
14:49:09:  <user v='thorsam'/>
14:49:09:
14:49:09:  <!-- Folding Slots -->
14:49:09:  <slot id='0' type='CPU'/>
14:49:09:</config>
14:49:09:Trying to access database...
14:49:09:Successfully acquired database lock
14:49:09:Enabled folding slot 00: READY cpu:31
14:49:09:WU00:FS00:Starting
14:49:09:WARNING:WU00:FS00:AS lowered CPUs from 31 to 30
14:49:09:WU00:FS00:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:\Users\thorsam\AppData\Roaming\FAHClient\cores/cores.foldingathome.org/Win32/AMD64/AVX/Core_a7.fah/FahCore_a7.exe -dir 00 -suffix 01 -version 705 -lifeline 11076 -checkpoint 15 -np 30
14:49:09:WU00:FS00:Started FahCore on PID 15300
14:49:10:WU00:FS00:Core PID:15008
14:49:10:WU00:FS00:FahCore 0xa7 started
14:49:11:WU00:FS00:0xa7:*********************** Log Started 2018-07-30T14:49:11Z ***********************
14:49:11:WU00:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
14:49:11:WU00:FS00:0xa7:       Type: 0xa7
14:49:11:WU00:FS00:0xa7:       Core: Gromacs
14:49:11:WU00:FS00:0xa7:    Website: https://foldingathome.org/
14:49:11:WU00:FS00:0xa7:  Copyright: (c) 2009-2018 foldingathome.org
14:49:11:WU00:FS00:0xa7:     Author: Joseph Coffland <joseph@cauldrondevelopment.com>
14:49:11:WU00:FS00:0xa7:       Args: -dir 00 -suffix 01 -version 705 -lifeline 15300 -checkpoint 15 -np
14:49:11:WU00:FS00:0xa7:             30
14:49:11:WU00:FS00:0xa7:     Config: <none>
14:49:11:WU00:FS00:0xa7:************************************ Build *************************************
14:49:11:WU00:FS00:0xa7:    Version: 0.0.17
14:49:11:WU00:FS00:0xa7:       Date: Apr 27 2018
14:49:11:WU00:FS00:0xa7:       Time: 16:19:36
14:49:11:WU00:FS00:0xa7: Repository: Git
14:49:11:WU00:FS00:0xa7:   Revision: 21359963583d09ec2063ef946399441c4df4ccd7
14:49:11:WU00:FS00:0xa7:     Branch: master
14:49:11:WU00:FS00:0xa7:   Compiler: Visual C++ 2008
14:49:11:WU00:FS00:0xa7:    Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
14:49:11:WU00:FS00:0xa7:   Platform: win32 10
14:49:11:WU00:FS00:0xa7:       Bits: 64
14:49:11:WU00:FS00:0xa7:       Mode: Release
14:49:11:WU00:FS00:0xa7:       SIMD: avx_256
14:49:11:WU00:FS00:0xa7:************************************ System ************************************
14:49:11:WU00:FS00:0xa7:        CPU: Unknown
14:49:11:WU00:FS00:0xa7:     CPU ID: 
14:49:12:WU00:FS00:0xa7:       CPUs: 64
14:49:12:WU00:FS00:0xa7:     Memory: 127.91GiB
14:49:12:WU00:FS00:0xa7:Free Memory: 122.24GiB
14:49:12:WU00:FS00:0xa7:    Threads: WINDOWS_THREADS
14:49:12:WU00:FS00:0xa7: OS Version: 6.2
14:49:12:WU00:FS00:0xa7:Has Battery: false
14:49:12:WU00:FS00:0xa7: On Battery: false
14:49:12:WU00:FS00:0xa7: UTC Offset: 2
14:49:12:WU00:FS00:0xa7:        PID: 15008
14:49:12:WU00:FS00:0xa7:        CWD: C:\Users\thorsam\AppData\Roaming\FAHClient\work
14:49:12:WU00:FS00:0xa7:         OS: Windows 10 Pro
14:49:12:WU00:FS00:0xa7:    OS Arch: AMD64
14:49:12:WU00:FS00:0xa7:********************************************************************************
14:49:12:WU00:FS00:0xa7:Project: 14050 (Run 0, Clone 743, Gen 48)
14:49:12:WU00:FS00:0xa7:Unit: 0x000000370002894c5ad5f6e7599cd346
14:49:12:WU00:FS00:0xa7:Digital signatures verified
14:49:12:WU00:FS00:0xa7:Calling: mdrun -s frame48.tpr -o frame48.trr -cpi state.cpt -cpt 15 -nt 30
14:49:12:8:127.0.0.1:New Web connection
14:49:13:WU00:FS00:0xa7:Steps: first=12000000 total=250000
14:49:17:WU00:FS00:0xa7:Completed 3252 out of 250000 steps (1%)
14:50:17:Adding folding slot 01: READY cpu:31
14:50:17:Saving configuration to config.xml
14:50:17:<config>
14:50:17:  <!-- Network -->
14:50:17:  <proxy v=':8080'/>
14:50:17:
14:50:17:  <!-- User Information -->
14:50:17:  <passkey v='********************************'/>
14:50:17:  <team v='34361'/>
14:50:17:  <user v='thorsam'/>
14:50:17:
14:50:17:  <!-- Folding Slots -->
14:50:17:  <slot id='0' type='CPU'/>
14:50:17:  <slot id='1' type='CPU'/>
14:50:17:</config>
14:50:17:FS00:Shutting core down
14:50:17:WU00:FS00:0xa7:WARNING:Console control signal 1 on PID 15008
14:50:17:WU00:FS00:0xa7:Exiting, please wait. . .
14:50:17:WU01:FS01:Connecting to 65.254.110.245:8080
14:50:18:WU01:FS01:Assigned to work server 155.247.166.220
14:50:18:WU01:FS01:Requesting new work unit for slot 01: READY cpu:15 from 155.247.166.220
14:50:18:WU01:FS01:Connecting to 155.247.166.220:8080
14:50:19:WU01:FS01:Downloading 801.61KiB
14:50:22:WU01:FS01:Download complete
14:50:22:WU01:FS01:Received Unit: id:01 state:DOWNLOAD error:NO_ERROR project:14049 run:0 clone:309 gen:44 core:0xa4 unit:0x000000320002894c5aa42842628cbe0d
14:50:22:WU01:FS01:Downloading core from http://cores.foldingathome.org/Win32/AMD64/Core_a4.fah
14:50:22:WU01:FS01:Connecting to cores.foldingathome.org:80
14:50:22:WU01:FS01:FahCore a4: Downloading 2.89MiB
14:50:24:WU00:FS00:0xa7:Folding@home Core Shutdown: INTERRUPTED
14:50:25:WU00:FS00:FahCore returned: INTERRUPTED (102 = 0x66)
14:50:25:WU00:FS00:Starting
14:50:25:WARNING:WU00:FS00:Changed SMP threads from 31 to 15 this can cause some work units to fail
14:50:25:WU00:FS00:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:\Users\thorsam\AppData\Roaming\FAHClient\cores/cores.foldingathome.org/Win32/AMD64/AVX/Core_a7.fah/FahCore_a7.exe -dir 00 -suffix 01 -version 705 -lifeline 11076 -checkpoint 15 -np 15
14:50:25:WU00:FS00:Started FahCore on PID 16084
14:50:25:WU00:FS00:Core PID:16108
14:50:25:WU00:FS00:FahCore 0xa7 started
14:50:25:WU01:FS01:FahCore a4: Download complete
14:50:26:WU00:FS00:0xa7:*********************** Log Started 2018-07-30T14:50:25Z ***********************
14:50:26:WU00:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
14:50:26:WU00:FS00:0xa7:       Type: 0xa7
14:50:26:WU00:FS00:0xa7:       Core: Gromacs
14:50:26:WU00:FS00:0xa7:    Website: https://foldingathome.org/
14:50:26:WU00:FS00:0xa7:  Copyright: (c) 2009-2018 foldingathome.org
14:50:26:WU00:FS00:0xa7:     Author: Joseph Coffland <joseph@cauldrondevelopment.com>
14:50:26:WU00:FS00:0xa7:       Args: -dir 00 -suffix 01 -version 705 -lifeline 16084 -checkpoint 15 -np
14:50:26:WU00:FS00:0xa7:             15
14:50:26:WU00:FS00:0xa7:     Config: <none>
14:50:26:WU00:FS00:0xa7:************************************ Build *************************************
14:50:26:WU00:FS00:0xa7:    Version: 0.0.17
14:50:26:WU00:FS00:0xa7:       Date: Apr 27 2018
14:50:26:WU00:FS00:0xa7:       Time: 16:19:36
14:50:26:WU00:FS00:0xa7: Repository: Git
14:50:26:WU00:FS00:0xa7:   Revision: 21359963583d09ec2063ef946399441c4df4ccd7
14:50:26:WU00:FS00:0xa7:     Branch: master
14:50:26:WU00:FS00:0xa7:   Compiler: Visual C++ 2008
14:50:26:WU00:FS00:0xa7:    Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
14:50:26:WU00:FS00:0xa7:   Platform: win32 10
14:50:26:WU00:FS00:0xa7:       Bits: 64
14:50:26:WU00:FS00:0xa7:       Mode: Release
14:50:26:WU00:FS00:0xa7:       SIMD: avx_256
14:50:26:WU00:FS00:0xa7:************************************ System ************************************
14:50:26:WU00:FS00:0xa7:        CPU: Unknown
14:50:26:WU00:FS00:0xa7:     CPU ID: 
14:50:26:WU00:FS00:0xa7:       CPUs: 64
14:50:26:WU00:FS00:0xa7:     Memory: 127.91GiB
14:50:26:WU00:FS00:0xa7:Free Memory: 122.52GiB
14:50:26:WU00:FS00:0xa7:    Threads: WINDOWS_THREADS
14:50:26:WU00:FS00:0xa7: OS Version: 6.2
14:50:26:WU00:FS00:0xa7:Has Battery: false
14:50:26:WU00:FS00:0xa7: On Battery: false
14:50:26:WU00:FS00:0xa7: UTC Offset: 2
14:50:26:WU00:FS00:0xa7:        PID: 16108
14:50:26:WU00:FS00:0xa7:        CWD: C:\Users\thorsam\AppData\Roaming\FAHClient\work
14:50:26:WU00:FS00:0xa7:         OS: Windows 10 Pro
14:50:26:WU00:FS00:0xa7:    OS Arch: AMD64
14:50:26:WU00:FS00:0xa7:********************************************************************************
14:50:26:WU00:FS00:0xa7:Project: 14050 (Run 0, Clone 743, Gen 48)
14:50:26:WU00:FS00:0xa7:Unit: 0x000000370002894c5ad5f6e7599cd346
14:50:26:WU00:FS00:0xa7:Digital signatures verified
14:50:26:WU00:FS00:0xa7:Calling: mdrun -s frame48.tpr -o frame48.trr -cpi state.cpt -cpt 15 -nt 15
14:50:26:WU01:FS01:Valid core signature
14:50:26:WU01:FS01:Unpacked 9.59MiB to cores/cores.foldingathome.org/Win32/AMD64/Core_a4.fah/FahCore_a4.exe
14:50:26:WU01:FS01:Starting
14:50:26:WU01:FS01:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:\Users\thorsam\AppData\Roaming\FAHClient\cores/cores.foldingathome.org/Win32/AMD64/Core_a4.fah/FahCore_a4.exe -dir 01 -suffix 01 -version 705 -lifeline 11076 -checkpoint 15 -np 15
14:50:27:WU01:FS01:Started FahCore on PID 16132
14:50:27:WU01:FS01:Core PID:16160
14:50:27:WU01:FS01:FahCore 0xa4 started
14:50:28:WU01:FS01:0xa4:
14:50:28:WU01:FS01:0xa4:*------------------------------*
14:50:28:WU01:FS01:0xa4:Folding@Home Gromacs GB Core
14:50:28:WU01:FS01:0xa4:Version 2.27 (Dec. 15, 2010)
14:50:28:WU01:FS01:0xa4:
14:50:28:WU01:FS01:0xa4:Preparing to commence simulation
14:50:28:WU01:FS01:0xa4:- Looking at optimizations...
14:50:28:WU01:FS01:0xa4:- Created dyn
14:50:28:WU01:FS01:0xa4:- Files status OK
14:50:28:WU01:FS01:0xa4:- Expanded 820337 -> 1946552 (decompressed 237.2 percent)
14:50:28:WU01:FS01:0xa4:Called DecompressByteArray: compressed_data_size=820337 data_size=1946552, decompressed_data_size=1946552 diff=0
14:50:28:WU01:FS01:0xa4:- Digital signature verified
14:50:28:WU01:FS01:0xa4:
14:50:28:WU01:FS01:0xa4:Project: 14049 (Run 0, Clone 309, Gen 44)
14:50:28:WU01:FS01:0xa4:
14:50:28:WU01:FS01:0xa4:Assembly optimizations on if available.
14:50:28:WU01:FS01:0xa4:Entering M.D.
14:50:30:WU00:FS00:0xa7:Steps: first=12000000 total=250000
14:50:33:WU00:FS00:0xa7:Completed 3792 out of 250000 steps (1%)
14:50:34:WU01:FS01:0xa4:Mapping NT from 15 to 15 
14:50:34:WU01:FS01:0xa4:Completed 0 out of 2500000 steps  (0%)
14:51:11:Saving configuration to config.xml
14:51:11:<config>
14:51:11:  <!-- Network -->
14:51:11:  <proxy v=':8080'/>
14:51:11:
14:51:11:  <!-- User Information -->
14:51:11:  <passkey v='********************************'/>
14:51:11:  <team v='34361'/>
14:51:11:  <user v='thorsam'/>
14:51:11:
14:51:11:  <!-- Folding Slots -->
14:51:11:  <slot id='0' type='CPU'/>
14:51:11:  <slot id='1' type='CPU'/>
14:51:11:</config>
14:53:15:WU00:FS00:0xa7:Completed 5000 out of 250000 steps (2%)

[bruce]

The CPU that FAH has identified has 32 threads. One CPU thread is allocated to process data for you GPU. Large prime numbers of threads (like 31) are prohibited, so you see
14:49:09:WARNING:WU00:FS00:AS lowered CPUs from 31 to 30
In this case, Project: 14050 is initially allocated 30 threads.

When you reconfigured to have two CPU slots, it looks like the system reallocated it to two slots using 15 threads + 15 threads. A new WU was downloaded from Project: 14049 to use 15 threads.

The atom-count of most active FAH projects is somewhat limited so there may be times when you have trouble finding individual projects that can successfully keep almost all of the CPU threads busy. Most donors with large CPU counts tend to split up their system into multiple CPU slots. I recommend you change the number of allocated threads from -1 (system choice) to other values that seem to work well for you.

Your GeForce GT 710 LP GPU is really very weak and won't earn a lot of points. Yes, it will work but it may not be the best choice. I had one for a while and replaced it with a GTX1050 LP because I had room for the dual-slot GPU, though that added a lot of heat to that system. I do still have a system with a GT710 because the PCIe slot configuration doesn't have enough room for the double-wide GPU.

[JimboPalmer]

Are you setting up two CPU slots of 16 CPUs each? (I almost think you set up two but left them at -1 and F@H is confused)

If two at 16 CPUS is still ugly, try 2 at 12 CPUs and 1 at 8 CPUs.

I

[toTOW]

His motherboard has two AMD EPYC 7551 CPUs ... so the client should see 128 CPUs (2*32 cores with HT) and not 32 ...

Just to be sure, does Windows see all the threads ?

[bruce]

I have not been able to find it on-line, but in the days of Windows 7, Windows Server was required to support multi-sockets hardware. That might or might not include Win10 Enterprise.
I notice that the OS version detected is 6.2, not 10. That's strange.

What does the Performance tab of Task_Manager report or CPU-Z?

(MS Doesn't make it easy to find their multi-socket support limitations.)

[thorsam]

First, sorry for my delay and thank you for your response.
I'm fighting with a pipe burst in my flat, and with Supermicro, because it's seems that the BIOS has a bug.

The entry in the tab 'Slots' is, was the software has even done.
Task_Manager and CPU-Z reportet correct 2 CPUs each with 32 cores respectively 64 threads.
The sense of my Geforce GT 710 is only, to have a better solution then 1024x768.

At the moment i'm install Linux Mint 19. I guess, this will be much better.
With my Windows_Installation the boinc_software use 128 threads.

I give you response, quick as i could.

[JimboPalmer]

I suspect that F@H will not use all 128 threads in one slot.

I think that you might try 4 slots of 32 'CPUs', and if that does not find work, 8 slots of 16 'CPUs'

Your GT710 is a very low end Kepler GPU; Kepler is still supported. I would not worry about folding on it.

https://en.wikipedia.org/wiki/Kepler_(m ... hitecture)

[bruce]

I've opened a bug report on this issue: https://github.com/FoldingAtHome/fah-issues/issues/1251

By the way, I suggest you remove the GPU slot in FAH's configuration. Inasmuch as you don't plan to fold on your GPU, this should reconfigure your system for 32 threads rather than 30. It's not much, but why not?

When you examine which CPUs are busy in Task_Manager, does it show all of the processing being done by one CPU at almost 100% or does it show the work distributed across all of the CPUs at about 50% each?

[Nathan_P]

F@H used to scale to 128 threads, i believe they have tested recently upto 96 threads. Mint will be a better bet than windows, if you get it working please post some results, i'm very interested.