Page 4 of 5
Re: Intel® HD Graphics
Posted: Wed May 29, 2019 8:37 pm
by foldy
First question: I would assign AMD iGPU always as GPU slot and only Intel iGPU for CPU acceleration.
Second question AVX vs iGPU: FAHBench
Re: Intel® HD Graphics
Posted: Wed May 29, 2019 9:32 pm
by bruce
What are the ID strings for the AMD GPUs and their GFLOPS.
Has anybody tested them on FAHBench?
Re: Intel® HD Graphics
Posted: Wed May 29, 2019 11:57 pm
by Theodore
AMD IGPs are significantly slower than a GT 1030. Newer Intel IGPs are faster than older AMD IGPs.
GPU folding on an Intel IGP more than doubles the PPD score, of just CPU folding, provided it has sufficient cooling (which most aftermarket coolers provide).
Re: Intel® HD Graphics
Posted: Thu May 30, 2019 4:33 am
by bruce
Both the (i)GPU and the CPU are capable of doing 3D FP32 calculations. SSE2 calculations are about double those on the original FPU hardware and AVX increases it again. A very fundamental question is how the 3D calculations on AVX compare to the same 3D calculations on the iGPU.
If I were the Intel chip designer, I would probably use the same hardware real-estate to do AVX and iGPU calculations so the might actually produce the same GFLOPS when only one of them is busy doing FP32. Can somebody please test that for us?
Re: Intel® HD Graphics
Posted: Thu May 30, 2019 7:47 am
by foldy
Intel CPU has its own AVX units which are independent of an iGPU bundled or not.
https://software.intel.com/en-us/forums ... pic/537333
Isn't there a gromacs benchmark for that? Gromacs avx vs OpenCL intel iGPU?
Re: Intel® HD Graphics
Posted: Fri May 31, 2019 7:47 am
by MeeLee
@Bruce, not sure if I understand your question, however,
For 2019 we're expecting VEGA 8 and up GPUs built into mobile processors, that will surpass 1Tflops FP32, according to Wikipedia.
Mobile, VEGA 8 and 10:
https://en.wikipedia.org/wiki/List_of_A ... o%22_(2019)
and Desktopm VEGA 8 and 11:
https://en.wikipedia.org/wiki/List_of_A ... e%22_(2018)
Vega 8 is stated to have 512 cores, at 1100Mhz,
Vega 10, 640 cores @ 1,4Ghz.
Vega 11, 704 @ 1,25Ghz.
These graphics cards are faster than a GT1030, and could aid CPU folding by doing GPU WUs.
While Intel only has their future (non-released) IGPs at 1+ TFlops.
However, if a driver can recognize Intel IGP graphics, and is backwards compatible with older Intel IGPs, I'm sure that a lot of current Intel CPUs with integrated graphics, would benefit from the graphics core boost.
Additional sources:
https://www.notebookcheck.net/AMD-Radeo ... 178.0.html
JimboPalmer wrote:
The operative word is much. GPUs which cannot do Double Precision (FP64) are not supported, even though use of Double Precision is minimized.
I guess the question is if the new intel IGPs can do FP64, and it's a good one!
Their FP32 performance is pretty good!
But as far as Intel graphics cards go, they usually only have the things they need. And nothing more.
There's a chance they won't get optimized for FP64.
But what prevents the CPU from doing those computations, and leave the FP32 computations for the IGP?
Re: Intel® HD Graphics
Posted: Fri May 31, 2019 3:41 pm
by bruce
As I understand it, OpenCL 1.1 support is insufficient while 1.2 or greater does support FP64, so it looks like Vega 8/11 should be fine. We can cross that bridge when we get to it.
The old versions of Intel's iGPU should be excluded, of course, but I think they already are. Devices identified as Vega will be (or already are) included in GPUs.txt and recognized by FAHClient.
The only potential unsolved problem is the hybrid GPU which switches between a supported GPU and an unsupported iGPU to save power when the driver thinks the graphics activity is light enough that it doesn't need to be in gamining/folding mode.
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 8:14 am
by hman2
bruce wrote:
There are no active GPU projects with deadlines long enough for the typical iGPU so those projects are assigned to AVX or SSE based hardware. Allowing longer deadlines for slow GPUs is detrimental to FAH's overall performance.
Both AVX and SSE are features of the CPU instruction set (extensions), and both exist for a decade or so. My Core i5 (3470), Ivy Bridge, incorporates them. But FAH does not mention making use of these instruction set extensions anywhere. Do they require a specific kernel (currently my Mint (Ubuntu) runs on 4.15.0-99)? Thanks
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 9:13 am
by PantherX
Welcome to the F@H Forum hman2,
Can you please post the log file? Ensure you include the first 100 lines which will inform us of what the system configuration is and what the client settings are. If you require guidance, please view this topic: viewtopic.php?f=24&t=26036
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 10:08 am
by JimboPalmer
hman2 wrote:Both AVX and SSE are features of the CPU instruction set (extensions), and both exist for a decade or so. My Core i5 (3470), Ivy Bridge, incorporates them. But FAH does not mention making use of these instruction set extensions anywhere. Do they require a specific kernel (currently my Mint (Ubuntu) runs on 4.15.0-99)? Thanks
If your CPU is capable of avx_256, then that will be used and mentioned somewhere in the log during start up of you Core_a7 job. The older, slower choice would have an sse2 in that line. The Core_a7 always uses one or the other, and has been good at choosing correcly. ( I have 3 PCs old enough to use sse2 and 4 new enough to use avx_256)
I will visit two PCs and and log snippets as examples.
newer PC, scroll to bottom:
Code: Select all
07:48:17:WU01:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
07:48:17:WU01:FS00:0xa7: Type: 0xa7
07:48:17:WU01:FS00:0xa7: Core: Gromacs
07:48:17:WU01:FS00:0xa7: Args: -dir 01 -suffix 01 -version 706 -lifeline 6156 -checkpoint 15 -np 3
07:48:17:WU01:FS00:0xa7:************************************ CBang *************************************
07:48:17:WU01:FS00:0xa7: Date: Oct 26 2019
07:48:17:WU01:FS00:0xa7: Time: 01:38:25
07:48:17:WU01:FS00:0xa7: Revision: c46a1a011a24143739ac7218c5a435f66777f62f
07:48:17:WU01:FS00:0xa7: Branch: master
07:48:17:WU01:FS00:0xa7: Compiler: Visual C++ 2008
07:48:17:WU01:FS00:0xa7: Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
07:48:17:WU01:FS00:0xa7: Platform: win32 10
07:48:17:WU01:FS00:0xa7: Bits: 64
07:48:17:WU01:FS00:0xa7: Mode: Release
07:48:17:WU01:FS00:0xa7:************************************ System ************************************
07:48:17:WU01:FS00:0xa7: CPU: Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz
07:48:17:WU01:FS00:0xa7: CPU ID: GenuineIntel Family 6 Model 42 Stepping 7
07:48:17:WU01:FS00:0xa7: CPUs: 4
07:48:17:WU01:FS00:0xa7: Memory: 3.96GiB
07:48:17:WU01:FS00:0xa7:Free Memory: 1.82GiB
07:48:17:WU01:FS00:0xa7: Threads: WINDOWS_THREADS
07:48:17:WU01:FS00:0xa7: OS Version: 6.2
07:48:17:WU01:FS00:0xa7:Has Battery: false
07:48:17:WU01:FS00:0xa7: On Battery: false
07:48:17:WU01:FS00:0xa7: UTC Offset: -5
07:48:17:WU01:FS00:0xa7: PID: 8336
07:48:17:WU01:FS00:0xa7: CWD: C:\Users\JimboPalmer\AppData\Roaming\FAHClient\work
07:48:17:WU01:FS00:0xa7:******************************** Build - libFAH ********************************
07:48:17:WU01:FS00:0xa7: Version: 0.0.18
07:48:17:WU01:FS00:0xa7: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
07:48:17:WU01:FS00:0xa7: Copyright: 2019 foldingathome.org
07:48:17:WU01:FS00:0xa7: Homepage: https://foldingathome.org/
07:48:17:WU01:FS00:0xa7: Date: Oct 26 2019
07:48:17:WU01:FS00:0xa7: Time: 01:52:30
07:48:17:WU01:FS00:0xa7: Revision: c1e3513b1bc0c16013668f2173ee969e5995b38e
07:48:17:WU01:FS00:0xa7: Branch: master
07:48:17:WU01:FS00:0xa7: Compiler: Visual C++ 2008
07:48:17:WU01:FS00:0xa7: Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
07:48:17:WU01:FS00:0xa7: Platform: win32 10
07:48:17:WU01:FS00:0xa7: Bits: 64
07:48:17:WU01:FS00:0xa7: Mode: Release
07:48:17:WU01:FS00:0xa7:************************************ Build *************************************
07:48:17:WU01:FS00:0xa7: SIMD: avx_256
07:48:17:WU01:FS00:0xa7:********************************************************************************
older PC, scroll to bottom:
Code: Select all
22:14:32:WU02:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
22:14:32:WU02:FS00:0xa7: Type: 0xa7
22:14:32:WU02:FS00:0xa7: Core: Gromacs
22:14:32:WU02:FS00:0xa7: Args: -dir 02 -suffix 01 -version 706 -lifeline 10012 -checkpoint 15 -np
22:14:32:WU02:FS00:0xa7: 3
22:14:32:WU02:FS00:0xa7:************************************ CBang *************************************
22:14:32:WU02:FS00:0xa7: Date: Oct 26 2019
22:14:32:WU02:FS00:0xa7: Time: 01:38:35
22:14:32:WU02:FS00:0xa7: Revision: c46a1a011a24143739ac7218c5a435f66777f62f
22:14:32:WU02:FS00:0xa7: Branch: master
22:14:32:WU02:FS00:0xa7: Compiler: Visual C++ 2008
22:14:32:WU02:FS00:0xa7: Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
22:14:32:WU02:FS00:0xa7: Platform: win32 10
22:14:32:WU02:FS00:0xa7: Bits: 64
22:14:32:WU02:FS00:0xa7: Mode: Release
22:14:32:WU02:FS00:0xa7:************************************ System ************************************
22:14:32:WU02:FS00:0xa7: CPU: Intel(R) Core(TM) i3 CPU 530 @ 2.93GHz
22:14:32:WU02:FS00:0xa7: CPU ID: GenuineIntel Family 6 Model 37 Stepping 2
22:14:32:WU02:FS00:0xa7: CPUs: 4
22:14:32:WU02:FS00:0xa7: Memory: 11.87GiB
22:14:32:WU02:FS00:0xa7:Free Memory: 9.39GiB
22:14:32:WU02:FS00:0xa7: Threads: WINDOWS_THREADS
22:14:32:WU02:FS00:0xa7: OS Version: 6.2
22:14:32:WU02:FS00:0xa7:Has Battery: false
22:14:32:WU02:FS00:0xa7: On Battery: false
22:14:32:WU02:FS00:0xa7: UTC Offset: -5
22:14:32:WU02:FS00:0xa7: PID: 7468
22:14:32:WU02:FS00:0xa7: CWD: C:\Users\Windows 10 PRO\AppData\Roaming\FAHClient\work
22:14:32:WU02:FS00:0xa7:******************************** Build - libFAH ********************************
22:14:32:WU02:FS00:0xa7: Version: 0.0.18
22:14:32:WU02:FS00:0xa7: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
22:14:32:WU02:FS00:0xa7: Copyright: 2019 foldingathome.org
22:14:32:WU02:FS00:0xa7: Homepage: https://foldingathome.org/
22:14:32:WU02:FS00:0xa7: Date: Oct 26 2019
22:14:32:WU02:FS00:0xa7: Time: 01:52:44
22:14:32:WU02:FS00:0xa7: Revision: c1e3513b1bc0c16013668f2173ee969e5995b38e
22:14:32:WU02:FS00:0xa7: Branch: master
22:14:32:WU02:FS00:0xa7: Compiler: Visual C++ 2008
22:14:32:WU02:FS00:0xa7: Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
22:14:32:WU02:FS00:0xa7: Platform: win32 10
22:14:32:WU02:FS00:0xa7: Bits: 64
22:14:32:WU02:FS00:0xa7: Mode: Release
22:14:32:WU02:FS00:0xa7:************************************ Build *************************************
22:14:32:WU02:FS00:0xa7: SIMD: sse2
22:14:32:WU02:FS00:0xa7:********************************************************************************
AVX does require OS support, but any OS in the last decade has it
https://en.wikipedia.org/wiki/Advanced_ ... em_support
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 10:37 am
by hman2
It's a Core i5-3470, which reports as Genuine Intel Family 6 Model 58 stepping 9, that is third generation iCore (a.k.a. Ivy Lake), and according to Intels datasheet incorporates AVX and SSE, four cores, but no hyperthreads. I run FAH A7 (7.6.9) under Linux kernel 4.15.0-99. I have to pair this with the older viewer version 7.5.1, because the viewer 7.6.9 demands a newer glibc than a LTS Mint/Ubuntu offers - but the two do harmonize well.
And in science.log I found this:
Calling: mdrun -s frame64.tpr -o frame64.trr -x frame64.xtc -cpt 15 -nt 1
Steps: first=16000000 total=250000
Reading file frame64.tpr, VERSION 5.0.4 (single precision)
Can not increase nstlist because verlet-buffer-tolerance is not set or used
Using 1 MPI thread
Completed 1 out of 250000 steps (0%)
Completed 2500 out of 250000 steps (1%)
Completed 5000 out of 250000 steps (2%)
The current CPU can measure timings more accurately than the code in
GROMACS was configured to use. This might affect your simulation
speed as accurate timings are needed for load-balancing.
Please consider rebuilding GROMACS with the GMX_USE_RDTSCP=OFF CMake option.
Setting the maximum number of constraint warnings to -1
WARNING: Changing nstcomm from 5 to 10
starting mdrun 'Protein in water'
log.txt, first 100 or so lines are as follows:
Code: Select all
*********************** Log Started 2020-05-07T13:18:16Z ***********************
13:18:16:****************************** FAHClient ******************************
13:18:16: Version: 7.6.9
13:18:16: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
13:18:16: Copyright: 2020 foldingathome.org
13:18:16: Homepage: https://foldingathome.org/
13:18:16: Date: Apr 17 2020
13:18:16: Time: 18:11:26
13:18:16: Revision: 398c2b17fa535e0cc6c9d10856b2154c32771646
13:18:16: Branch: master
13:18:16: Compiler: GNU 8.3.0
13:18:16: Options: -std=c++11 -ffunction-sections -fdata-sections -O3 -funroll-loops
13:18:16: -fno-pie
13:18:16: Platform: linux2 4.19.0-5-amd64
13:18:16: Bits: 64
13:18:16: Mode: Release
13:18:16: Args: --child /etc/fahclient/config.xml --run-as fahclient
13:18:16: --pid-file=/var/run/fahclient.pid --daemon
13:18:16: Config: /etc/fahclient/config.xml
13:18:16:******************************** CBang ********************************
13:18:16: Date: Apr 17 2020
13:18:16: Time: 18:10:13
13:18:16: Revision: 2fb0be7809c5e45287a122ca5fbc15b5ae859a3b
13:18:16: Branch: master
13:18:16: Compiler: GNU 8.3.0
13:18:16: Options: -std=c++11 -ffunction-sections -fdata-sections -O3 -funroll-loops
13:18:16: -fno-pie -fPIC
13:18:16: Platform: linux2 4.19.0-5-amd64
13:18:16: Bits: 64
13:18:16: Mode: Release
13:18:16:******************************* System ********************************
13:18:16: CPU: Intel(R) Core(TM) i5-3470 CPU @ 3.20GHz
13:18:16: CPU ID: GenuineIntel Family 6 Model 58 Stepping 9
13:18:16: CPUs: 4
13:18:16: Memory: 15.55GiB
13:18:16:Free Memory: 15.06GiB
13:18:16: Threads: POSIX_THREADS
13:18:16: OS Version: 4.15
13:18:16:Has Battery: false
13:18:16: On Battery: false
13:18:16: UTC Offset: 2
13:18:16: PID: 1166
13:18:16: CWD: /var/lib/fahclient
13:18:16: OS: Linux 4.15.0-99-generic x86_64
13:18:16: OS Arch: AMD64
13:18:16: GPUs: 0
13:18:16: CUDA: Not detected: Failed to open dynamic library 'libcuda.so':
13:18:16: libcuda.so: cannot open shared object file: No such file or
13:18:16: directory
13:18:16: OpenCL: Not detected: Failed to open dynamic library 'libOpenCL.so':
13:18:16: libOpenCL.so: cannot open shared object file: No such file or
13:18:16: directory
13:18:16:******************************* libFAH ********************************
13:18:16: Date: Apr 15 2020
13:18:16: Time: 21:43:24
13:18:16: Revision: 216968bc7025029c841ed6e36e81a03a316890d3
13:18:16: Branch: master
13:18:16: Compiler: GNU 8.3.0
13:18:16: Options: -std=c++11 -ffunction-sections -fdata-sections -O3 -funroll-loops
13:18:16: -fno-pie
13:18:16: Platform: linux2 4.19.0-5-amd64
13:18:16: Bits: 64
13:18:16: Mode: Release
13:18:16:***********************************************************************
13:18:16:<config>
13:18:16: <!-- Client Control -->
13:18:16: <fold-anon v='true'/>
13:18:16:
13:18:16: <!-- Folding Slot Configuration -->
13:18:16: <cause v='COVID_19'/>
13:18:16: <gpu v='false'/>
13:18:16:
13:18:16: <!-- Network -->
13:18:16: <proxy v=':8080'/>
13:18:16:
13:18:16: <!-- Slot Control -->
13:18:16: <power v='light'/>
13:18:16:
13:18:16: <!-- User Information -->
13:18:16: <team v='264242'/>
13:18:16: <user v='hman2'/>
13:18:16:
13:18:16: <!-- Folding Slots -->
13:18:16: <slot id='0' type='CPU'>
13:18:16: <cpus v='2'/>
13:18:16: <paused v='true'/>
13:18:16: </slot>
13:18:16: <slot id='1' type='CPU'>
13:18:16: <cpus v='1'/>
13:18:16: </slot>
13:18:16:</config>
13:18:16:Trying to access database...
13:18:16:Successfully acquired database lock
13:18:16:Enabled folding slot 00: PAUSED cpu:2 (by user)
13:18:16:Enabled folding slot 01: READY cpu:1
13:18:16:WU01:FS01:Starting
13:18:16:WU01:FS01:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/cores.foldingathome.org/v7/lin/64bit/avx/Core_a7.fah/FahCore_a7 -dir 01 -suffix 01 -version 706 -lifeline 1166 -checkpoint 15 -np 1
13:18:16:WU01:FS01:Started FahCore on PID 1175
13:18:16:WU01:FS01:Core PID:1179
13:18:16:WU01:FS01:FahCore 0xa7 started
[93m13:18:16:WARNING:WU01:FS01:FahCore returned: BAD_FRAME_CHECKSUM (112 = 0x70)[0m
[93m13:18:16:WARNING:WU01:FS01:Fatal error, dumping[0m
13:18:16:WU01:FS01:Sending unit results: id:01 state:SEND error:DUMPED project:14190 run:3 clone:90 gen:23 core:0xa7 unit:0x0000001e0002894b5d5d2f2e005f40a9
13:18:16:WU01:FS01:Connecting to 155.247.166.219:8080
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 10:40 am
by hman2
Oh, while reading, I stumbled upon the fact that FAH does look at battery status. Thanks for that. Since this is a desktop PC and not a laptop, has_battery is false. But on_battery can nonetheless become true! If mains power goes out, my UPS will kick in, and since I have a NUT server/client, this status will get reported and it would be very nice if in such incidents the power consumption could drop
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 10:54 am
by hman2
Ah, now I see where that gets written. Looks live avx-256 got activated, thanks. The info was just buries under loads of other information, although in a separated paragraph aptly labelled SIMD
Re: Intel® HD Graphics
Posted: Mon May 11, 2020 1:30 pm
by JimboPalmer
hman2 wrote:Ah, now I see where that gets written. Looks live avx-256 got activated, thanks. The info was just buries under loads of other information, although in a separated paragraph aptly labelled SIMD
Excellent, that is the best Core_a7 is able to do at this time.
Re: Intel® HD Graphics
Posted: Fri May 29, 2020 7:20 pm
by cristipurdel
I read two pages of this forum and still not found out if Intel integrated GPU is supported or not.
I am running the portable app from portableapps, and initially I had 3 slots (1 cpu + 1 intel gpu + 1 nvidia gpu)
The fah client shows that the intel gpu is doing smth, but gpu-z/task manager says it does not do anything.
FahCore_a7 is using 2 out of 8 threads, but the Intel UHD Graphics 620 is not showing any load.
Also on an older Intel GPU I saw that it was showing as extra slot, but now it is gone.