mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware > GPU Computing > GpuOwl

Reply
 
Thread Tools
Old 2018-03-30, 08:00   #364
SELROC
 

3×17×191 Posts
Default

Quote:
Originally Posted by M344587487 View Post
Yes I said that. I think that it should work, and that the API just hasn't been fully updated for PRP yet. Short of scraping your assignment details page (or the API equivalent whatever that is) I'm not sure there's a solution. I could be wrong.

I have read that Mihai is working on this topic of integration with primenet.
  Reply With Quote
Old 2018-03-31, 09:02   #365
SELROC
 

215748 Posts
Default

Quote:
Originally Posted by M344587487 View Post
Yes I said that. I think that it should work, and that the API just hasn't been fully updated for PRP yet. Short of scraping your assignment details page (or the API equivalent whatever that is) I'm not sure there's a solution. I could be wrong.

I have added some diagnostics to this python script, the content of the fetch is effectively empty, so nothing is being written in the worktodo file except newlines.
  Reply With Quote
Old 2018-04-06, 06:21   #366
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

2·32·5·72 Posts
Default

Quote:
Originally Posted by kracker View Post
As requested... instructions on how to compile on windows (I use msys2.. and also there are probably better ways to do it but it's just how I did it)

1) Download, install and follow the instructions for updating MSYS2 here: https://www.msys2.org/
2) Download and install AMD APP SDK(make sure you use the 64bit version) for Windows: https://developer.amd.com/amd-accele...ssing-app-sdk/
3) Copy the contents of C:\Program Files (x86)\AMD APP SDK\3.0\lib\x86_64 to C:\msys64\mingw64\lib and C:\Program Files (x86)\AMD APP SDK\3.0\include to C:\msys64\mingw64\include
4) Install gcc (pacman -S mingw-w64-x86_64-gcc)
5) Download gpuowl sources and drop them somewhere(to /home/username/ is probably easiest)
6) Run MSYS2 from mingw64.exe and cd to the directory you extracted the source to
7) Compile by:
g++ -c gpuowl.cpp
g++ -o gpuowl.exe gpuowl.o -lOpenCL -static
strip gpuowl.exe
Very helpful.

I tried gpuowl v2 code with the above, got the following.

Code:
C:\msys64\home\ken>gpuowl -user kriesel -cpu condorella-rx550-0 -device 0
gpuOwL v2.0- GPU Mersenne primality checker
gfx804-8x1203-@28:0.0 Radeon 500 Series
Note: using short, fused carry and fused tail kernels
OpenCL compilation in 1466 ms, with " -DEXP=83871443u  -I. -cl-fast-relaxed-math -cl-kernel-arg-info "
PRP-3: FFT 5000K (625 * 4096 * 2) of 83871443 (16.38 bits/word) [2018-04-06 00:59:55 Central Daylight Time]
Starting at iteration 3058000
error -55 (fft4K)
Assertion failed!

Program: C:\msys64\home\ken\gpuowl.exe
File: clwrap.h, Line 267

Expression: check(clEnqueueNDRangeKernel(queue, kernel, 1, __null, &workSize, &groupSize, 0, __null, __null), name.c_str())

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Windows 7 x64 sp1 & fully patched displayed an appcrash with the following:

Code:
  Problem Event Name:    APPCRASH
  Application Name:    gpuowl.exe
  Application Version:    0.0.0.0
  Application Timestamp:    00000000
  Fault Module Name:    gpuowl.exe
  Fault Module Version:    0.0.0.0
  Fault Module Timestamp:    00000000
  Exception Code:    40000015
  Exception Offset:    0000000000018f36
  OS Version:    6.1.7601.2.1.0.256.48
  Locale ID:    1033
  Additional Information 1:    29e3
  Additional Information 2:    29e3d2050e41c7f5d21e1089382ae02d
  Additional Information 3:    de1c
  Additional Information 4:    de1ce85cf0adf294aec65c4dce7db4bf

Last fiddled with by kriesel on 2018-04-06 at 06:33
kriesel is online now   Reply With Quote
Old 2018-04-06, 06:52   #367
SELROC
 

3×41×73 Posts
Default

Quote:
Originally Posted by kriesel View Post
Very helpful.

I tried gpuowl v2 code with the above, got the following.

Code:
C:\msys64\home\ken>gpuowl -user kriesel -cpu condorella-rx550-0 -device 0
gpuOwL v2.0- GPU Mersenne primality checker
gfx804-8x1203-@28:0.0 Radeon 500 Series
Note: using short, fused carry and fused tail kernels
OpenCL compilation in 1466 ms, with " -DEXP=83871443u  -I. -cl-fast-relaxed-math -cl-kernel-arg-info "
PRP-3: FFT 5000K (625 * 4096 * 2) of 83871443 (16.38 bits/word) [2018-04-06 00:59:55 Central Daylight Time]
Starting at iteration 3058000
error -55 (fft4K)
Assertion failed!

Program: C:\msys64\home\ken\gpuowl.exe
File: clwrap.h, Line 267

Expression: check(clEnqueueNDRangeKernel(queue, kernel, 1, __null, &workSize, &groupSize, 0, __null, __null), name.c_str())

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Windows 7 x64 sp1 & fully patched displayed an appcrash with the following:

Code:
  Problem Event Name:    APPCRASH
  Application Name:    gpuowl.exe
  Application Version:    0.0.0.0
  Application Timestamp:    00000000
  Fault Module Name:    gpuowl.exe
  Fault Module Version:    0.0.0.0
  Fault Module Timestamp:    00000000
  Exception Code:    40000015
  Exception Offset:    0000000000018f36
  OS Version:    6.1.7601.2.1.0.256.48
  Locale ID:    1033
  Additional Information 1:    29e3
  Additional Information 2:    29e3d2050e41c7f5d21e1089382ae02d
  Additional Information 3:    de1c
  Additional Information 4:    de1ce85cf0adf294aec65c4dce7db4bf


The OpenCL error -55 is "invalid work item size". I have hit the same error once, resolved by changing the gpu driver.
Probably you have installed amdgpu, but you need to install amdgpu-pro.
  Reply With Quote
Old 2018-04-06, 13:55   #368
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

104728 Posts
Default

Quote:
Originally Posted by SELROC View Post
The OpenCL error -55 is "invalid work item size". I have hit the same error once, resolved by changing the gpu driver.
Probably you have installed amdgpu, but you need to install amdgpu-pro.
Thanks. I'll look into whether updating the Windows 7 driver makes any difference.
(Amdgpu and amdgpu-pro appear to be linux-specific.)
I've seen issues with downloaded updates for the AMD RX550 Windows driver in the past, under gpuowl 1.9, so had reverted to the version shipped with the MSI branded hardware. Maybe enough time has passed that it's worth another try with the current latest.

Last fiddled with by kriesel on 2018-04-06 at 13:59
kriesel is online now   Reply With Quote
Old 2018-04-07, 10:33   #369
SELROC
 

22×1,123 Posts
Default

Quote:
Originally Posted by kriesel View Post
Thanks. I'll look into whether updating the Windows 7 driver makes any difference.
(Amdgpu and amdgpu-pro appear to be linux-specific.)
I've seen issues with downloaded updates for the AMD RX550 Windows driver in the past, under gpuowl 1.9, so had reverted to the version shipped with the MSI branded hardware. Maybe enough time has passed that it's worth another try with the current latest.

I have installed amdgpu-pro on Debian, which requires some trickery to install.
Basically you need to edit the installation procedure and substitute "ubuntu" with "debian" in the first case statement. Then the installation will work on Debian too. Do not install firmware-linux-nonfree, it has poor performance.
  Reply With Quote
Old 2018-04-07, 15:53   #370
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

2×32×5×72 Posts
Default gpuOwL 2 on Windows 7 sp1 with current updates

It failed with an MSI-sourced driver dated April 2017 on an MSI RX550.

It worked with the latest MSI driver 18.2.3 dated Feb 26 2018.
https://us.msi.com/Graphics-card/sup...iver&Win7%2064
There's a March 23 2018 driver available from AMD, v18.3.4 that I have not tried.
https://support.amd.com/en-us/download

Performance of the 5000K fft on the RX550 with the 18.2.3 driver in a quick test (~40,000 iterations each) was:
short carry 17.3 ms/iter,
medium 17.6,
long 17.4,
compared to V1.9 gpuOwL on the same gpu, same pcie physical connection, April 2017 driver,
10.9 ms/iter for -fft DP -legacy -size 4M;
18.9 ms/iter -fft M61 -size 4M;
21.4 ms/iter -fft DP -legacy -size 8M.

The driver change coincided with an increase by about 5% of iteration time, on the same gpu, in V1.9 gpuOwL:
Code:
gpuOwL v1.9- GPU Mersenne primality checker
Radeon 500 Series 8 @f:0.0, gfx804 1203MHz

OpenCL compilation in 2964 ms, with "-I. -cl-fast-relaxed-math -cl-std=CL2.0  -DEXP=149448041u -DWIDTH=2048u -DHEIGHT=2048u -DLOG_NWORDS=23u -DFP_DP=1 "
PRP-3: FFT 8M (2048 * 2048 * 2) of 149448041 (17.82 bits/word) [2018-04-06 13:34:54 Central Daylight Time]
Starting at iteration 69030500
OK 69030500 / 149448041 [46.19%], 0.00 ms/it; ETA 0d 00:00; f402d0898a94a560 [13:35:10]
OK 69031000 / 149448041 [46.19%], 21.34 ms/it; ETA 19d 20:37; 75f8954003f7699a [13:35:35]
OK 69032000 / 149448041 [46.19%], 21.30 ms/it [21.27, 21.33] CV 0.2%, check 13.87s; ETA 19d 19:48; a828e92c37962e50 [13:36:10]
OK 69035000 / 149448041 [46.19%], 21.29 ms/it [21.27, 21.33] CV 0.1%, check 13.83s; ETA 19d 19:31; 6fa8384867b99be1 [13:37:28]
OK 69040000 / 149448041 [46.20%], 21.40 ms/it [21.27, 22.24] CV 1.4%, check 13.93s; ETA 19d 21:54; 673e4eae09f49243 [13:39:29]
OK 69050000 / 149448041 [46.20%], 21.36 ms/it [21.27, 22.24] CV 1.1%, check 13.82s; ETA 19d 21:06; ff34f1d3d762e0c7 [13:43:16]
OK 69060000 / 149448041 [46.21%], 21.36 ms/it [21.25, 22.24] CV 1.1%, check 13.62s; ETA 19d 20:59; 796e6798379ef1d1 [13:47:03]
OK 69080000 / 149448041 [46.22%], 21.34 ms/it [21.25, 22.78] CV 1.3%, check 13.87s; ETA 19d 20:30; 2f4417db86c153db [13:54:24]
OK 69100000 / 149448041 [46.24%], 21.38 ms/it [21.25, 22.28] CV 1.3%, check 13.84s; ETA 19d 21:12; 1cc95987fefb3e6d [14:01:45]

Stopping, please wait..
OK 69117500 / 149448041 [46.25%], 21.36 ms/it [21.27, 22.25] CV 1.1%, check 13.57s; ETA 19d 20:38; f9d2cf95c539e93e [14:08:13]

Bye
gpuOwL v1.9- GPU Mersenne primality checker
Radeon 550 Series 8 @f:0.0, gfx804 1203MHz

OpenCL compilation in 2901 ms, with "-I. -cl-fast-relaxed-math -cl-std=CL2.0  -DEXP=149448041u -DWIDTH=2048u -DHEIGHT=2048u -DLOG_NWORDS=23u -DFP_DP=1 "
PRP-3: FFT 8M (2048 * 2048 * 2) of 149448041 (17.82 bits/word) [2018-04-06 16:18:38 Central Daylight Time]
Starting at iteration 69117500
OK 69117500 / 149448041 [46.25%], 0.00 ms/it; ETA 0d 00:00; f9d2cf95c539e93e [16:18:55]
OK 69118000 / 149448041 [46.25%], 22.59 ms/it; ETA 21d 00:02; ff48fd98e0c117e6 [16:19:21]
OK 69120000 / 149448041 [46.25%], 22.55 ms/it [22.52, 22.59] CV 0.1%, check 14.83s; ETA 20d 23:08; 32ec21c245f8704f [16:20:21]

Last fiddled with by kriesel on 2018-04-07 at 15:59
kriesel is online now   Reply With Quote
Old 2018-04-16, 05:24   #371
Lexicographer
 
Mar 2018
Shenzhen, China

2·32 Posts
Default No nVidia support yet?

No luck using it with nVidia GPU:
Quote:
$ ./gpuowl
gpuOwL v2.0-dbc5a01-mod GPU Mersenne primality checker
GeForce GTX 1080 Ti-28x1607-
Note: using medium, fused carry and fused tail kernels


OpenCL compilation in 5 ms, with " -DEXP=48484841u -DCARRY_MEDIUM=1u -I. -cl-fast-relaxed-math -cl-kernel-arg-info "
PRP-3: FFT 5000K (625 * 4096 * 2) of 48484841 (9.47 bits/word) [2018-04-16 12:57:50 CST]
Starting at iteration 0
error -5 (carryFused)
gpuowl: clwrap.h:267: void run(cl_queue, cl_kernel, size_t, size_t, const string&): Assertion `check(clEnqueueNDRangeKernel(queue, kernel, 1, __null, &workSize, &groupSize, 0, __null, __null), name.c_str())' failed.
Aborted (core dumped)
Lexicographer is offline   Reply With Quote
Old 2018-05-04, 17:52   #372
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

113A16 Posts
Default

Quote:
Originally Posted by Lexicographer View Post
No luck using it with nVidia GPU:
Ditto, on Windows 7 64-bit, NVIDIA GTX1070.

It looked promising for several seconds after launch, then boom, appcrash dialog box:
Code:
Problem signature:
  Problem Event Name:    APPCRASH
  Application Name:    gpuowl.exe
  Application Version:    0.0.0.0
  Application Timestamp:    00000000
  Fault Module Name:    gpuowl.exe
  Fault Module Version:    0.0.0.0
  Fault Module Timestamp:    00000000
  Exception Code:    40000015
  Exception Offset:    0000000000018f36
  OS Version:    6.1.7601.2.1.0.256.48
  Locale ID:    1033
  Additional Information 1:    dd12
  Additional Information 2:    dd12a83cefd863936d23d0206825e069
  Additional Information 3:    f88f
  Additional Information 4:    f88f59fe4ce85bde4365a197b0e610e4

Read our privacy statement online:
  http://go.microsoft.com/fwlink/?linkid=104288&clcid=0x0409

If the online privacy statement is not available, please read our privacy statement offline:
  C:\Windows\system32\en-US\erofflps.txt
Console text:
Code:
...> gp

...> gpuowl -device 0 -user kriesel -cpu condorette-gtx1070 -carry long
gpuOwL v2.0- GPU Mersenne primality checker
GeForce GTX 1070-15x1708-
Note: using long carry and fused tail kernels


OpenCL compilation in 2740 ms, with " -DEXP=83871443u  -I. -cl-fast-relaxed-math -cl-kernel-arg-info "
PRP-3: FFT 5000K (625 * 4096 * 2) of 83871443 (16.38 bits/word) [2018-05-04 12:25:03 Central Daylight Time]
Starting at iteration 3391500
error -9999 (fft625)
Assertion failed!

Program: c:\Users\Ken\Documents\gpuowl\v2 test\gpuowl.exe
File: clwrap.h, Line 267

Expression: check(clEnqueueNDRangeKernel(queue, kernel, 1, __null, &workSize, &groupSize, 0, __null, __null), name.c_str())

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Attached Thumbnails
Click image for larger version

Name:	gtx1070 crash on launch.png
Views:	31
Size:	50.4 KB
ID:	18189  
kriesel is online now   Reply With Quote
Old 2018-05-21, 11:03   #373
SELROC
 

76628 Posts
Default gpuowl RAM consumption ?

How much CPU RAM does gpuowl consumes?
  Reply With Quote
Old 2018-05-21, 19:44   #374
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

2·32·5·72 Posts
Default

Quote:
Originally Posted by SELROC View Post
How much CPU RAM does gpuowl consumes?
Per instance, gpuOwL v1.9 on 8M fft length running exponents ~150M, ~115MB private working set, 145MB working set, 382MB peak working set on Windows 7 64-bit. Meanwhile GPU occupancy ~475-490MB each.
kriesel is online now   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
mfakto: an OpenCL program for Mersenne prefactoring Bdot GPU Computing 1634 2020-09-10 21:40
GPUOWL AMD Windows OpenCL issues xx005fs GpuOwl 0 2019-07-26 21:37
Testing an expression for primality 1260 Software 17 2015-08-28 01:35
Testing Mersenne cofactors for primality? CRGreathouse Computer Science & Computational Number Theory 18 2013-06-08 19:12
Primality-testing program with multiple types of moduli (PFGW-related) Unregistered Information & Answers 4 2006-10-04 22:38

All times are UTC. The time now is 15:07.

Mon Sep 21 15:07:40 UTC 2020 up 11 days, 12:18, 1 user, load averages: 1.57, 1.65, 1.72

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.