![]() |
![]() |
#1 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24·3·157 Posts |
![]()
This thread is intended as a place to post for sharing, successful builds of Gpuowl for Windows. There's no plan to duplicate previous build posts.
To make your own builds, see directions at https://www.mersenneforum.org/showpo...4&postcount=21 For the current version source (and previous too) https://github.com/preda/gpuowl The most current build I have for posting as of now is v6.11-318-g310999, attached. There's the mersenne software mirror at https://download.mersenne.ca/gpuowl There's a large list of links to prior version gpuowl Windows build posts, at https://www.mersenneforum.org/showpo...39&postcount=4 It goes all the way back to gpuowl v0.1! Some of the very early versions might be useful in rare special cases, such as if 4M fft LL with offset on AMD OpenCL is useful, or zero-offset 4M fft LL with Jacobi check. Also the v1.9 multiple transform types can be interesting; I was able to run V1.9 PRP -fft M61 on an Intel igp, according to a log file I found recently. Performance was not great, but that's a slow igp competing for TDP with prime95 on the CPU. Note that all of this software development, testing, documentation, mirror hosting, forum hosting, etc. is done by volunteers for free. Respect them, and show gratitude. Have fun. |
![]() |
![]() |
![]() |
#2 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24·3·157 Posts |
![]()
This is at the moment the current commit.
Note that as of somewhere between -295 and -318, maximum fft length increased from 96M to 120M, nominally supporting exponent up to a little over 231. Note also that since I do not currently have the permissions to modify my posts in ordinary threads in most of the forum past the 1-hour-after-first-posted deadline for ordinary forum users, I am currently unable to update post oneof this thread with the most current build, or to directly correct any posting errors later. Last fiddled with by kriesel on 2020-06-21 at 18:52 |
![]() |
![]() |
![]() |
#3 |
"Dylan"
Mar 2017
3·199 Posts |
![]()
Attached below is a build of commit g43cdf1c, which at the time of posting is the latest commit. This was built on a system with an Nvidia card (GTX 1050ti Max-Q) using the instructions as given in the first post, with the following addition: I copied C:\msys64\mingw64\lib\x86_64\libOpenCL.a to C:\msys64\mingw64\lib\gcc\x86_64-w64-mingw32\10.1.0\libOpenCL.a, as the linker could not find the OpenCL library, even when the path to the library was specified in the Makefile. Also attached is the build log and help file output.
|
![]() |
![]() |
![]() |
#4 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24×3×157 Posts |
![]()
Attached is a Windows x64 build of Gpuowl v6.11-330-ge5a8f2c, which is the current commit at the moment.
|
![]() |
![]() |
![]() |
#5 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24×3×157 Posts |
![]()
This was built from the commit current minutes before posting, in the usual way.
The build-log.txt contains the console commands and console output of the build process, including the usual shower of warnings from the msys2 environment. The .7z file contains, as usual, the Win7 x64 built executable, help output from the program on the build machine, the license file, the readme file, and the text file use-flags.txt containing the top (comment) portion of gpuowl.cl. |
![]() |
![]() |
![]() |
#6 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24×3×157 Posts |
![]()
As usual, a build-log.txt, and a .7z holding the executable an a few other files
|
![]() |
![]() |
![]() |
#7 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
1D7016 Posts |
![]()
Here's the latest for the moment.
Thanks Mihai for a couple of commits reestablishing Windows msys2 build compatibility. And something seems to have reduced the compile warnings count considerably, so the build log is ~ half the previous size. Last fiddled with by kriesel on 2020-07-21 at 22:50 |
![]() |
![]() |
![]() |
#8 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24×3×157 Posts |
![]()
Here's the Windows build for the latest commit as of minutes ago.
As usual, build-log.txt, and a .7z file containing the rest. |
![]() |
![]() |
![]() |
#9 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
1D7016 Posts |
![]()
Here's the Windows build for the latest commit as of this morning.
As usual, build-log.txt, and a .7z file containing the rest. |
![]() |
![]() |
![]() |
#10 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24·3·157 Posts |
![]()
As usual, a .7z file containing executable, help output, license, use-flags, and readme, and a separate build log. This has the P-1 at low differential cost approach as part of a PRP run, and omits LLDC capability. See https://www.mersenneforum.org/showthread.php?t=26007 for more info.
|
![]() |
![]() |
![]() |
#11 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
24·3·157 Posts |
![]()
As usual, a .7z file containing executable, help output, license, use-flags, and readme, and a separate build log. This has the Windows worktodo file rename fix.
|
![]() |
![]() |
![]() |
Thread Tools | |
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
gpuOWL for Wagstaff | GP2 | GpuOwl | 22 | 2020-06-13 16:57 |
gpuOwl Windows setup for Radeon VII | Prime95 | GpuOwl | 91 | 2019-12-30 08:30 |
GPUOWL AMD Windows OpenCL issues | xx005fs | GpuOwl | 0 | 2019-07-26 21:37 |
ARM builds and SIMD-assembler prospects | ewmayer | Mlucas | 183 | 2019-02-25 08:17 |
Running 32-bit builds on a Win7 system | ewmayer | Programming | 34 | 2010-10-18 22:36 |