Dr-Noob
e0b0a6913c
Build pciutils only if neccesary
2025-10-16 07:52:58 +02:00
Dr-Noob
8794cd322d
[v0.30] Add support for building on AMD where rocm-cmake is not installed
2025-10-16 07:24:45 +02:00
Dr-Noob
5df85aea2c
[v0.30] Add uarch detection to AMD GPUs
...
Similarly to NVIDIA and Intel GPUs, we now detect microarchitecture,
also with manufacturing process and specific chip name. We infer all
of this from the gfx name (in the code we use the term llvm_target),
altough it's not clear yet that this method is completely reliable (see
comments for more details). In the future we might want to replace that
with a better way. Once we have the gfx name, we *should* be able to
infer the specific chip, and from the chip we can easily infer the
microarchitecture.
This commit also includes some refactorings and code improvements on
the HSA backend.
2025-10-15 08:23:28 +02:00
Dr-Noob
b29b17d14f
[v0.30] Add support for AMD GPUs
...
Adds very basic support for AMD (experimental). The only install
requirement is ROCm. Unlike NVIDIA, we don't need the CUDA equivalent
(HIP) to make gpufetch work, which reduces the installation
requirements quite significantly.
Major changes:
* CMakeLists:
- Make CUDA not compiled by default (since we now may want to target
AMD only)
- Set build flags on gpufetch cmake target instead of doing
"set(CMAKE_CXX_FLAGS". This fixes a warning coming from ROCm.
- Assumes that the ROCm CMake files are installed (should be fixed
later)
* hsa folder: AMD support is implemented via HSA (Heterogeneous System
Architecture) calls. Therefore, HSA is added as a new backend to
gpufetch. We only print basic stuff for now, so we may need more
things in the future to give full support for AMD GPUs.
NOTE: This commit will probably break AUR packages since we used to
build CUDA by default, which is no longer the case. The AUR package
should be updated and use -DENABLE_CUDA_BACKEND or -DENABLE_HSA_BACKEND
as appropriate.
2025-10-12 12:34:56 +02:00
Dr-Noob
57caadf530
[v0.25] Add Intel Whiskey Lake SoC ( #42 )
2023-10-20 07:59:07 +01:00
Dr-Noob
ed35cb872b
[v0.25] Leave cuda/intel backend to decide how to report PCI vendor failure
2023-03-31 16:16:46 +02:00
Dr-Noob
3d36852f9d
[v0.25] Fix for PCI class 0302 can also be responsible for GPUs (like in AWS)
2023-03-31 16:12:22 +02:00
Dr-Noob
fb0109d327
[v0.25] PCI class 0302 can also be responsible for GPUs
2023-03-31 16:08:59 +02:00
Dr-Noob
68619aa03e
[v0.25] Avoid segfault when the pci vendor is not found
2023-03-31 15:50:37 +02:00
Dr-Noob
a4006db616
[v0.25] Remove warning notice
2022-12-03 18:06:36 +01:00
Dr-Noob
774550307c
[v0.25] Add option to print all GPUs as requested in #33
2022-12-03 18:04:50 +01:00
Dr-Noob
06dc50b6a5
[v0.25] Updated cuda_helper to support latest GPUs
2022-12-03 16:39:18 +00:00
Dr-Noob
9837236c7e
[v0.25] Fixed some details in README and build.sh
2022-12-03 14:46:48 +00:00
Dr-Noob
a6f0c18fcb
[v0.25] Add missing Ampere GPU chips and new uarchs: ada and hopper
2022-10-25 20:13:29 +02:00
Dr-Noob
94490b3f38
[v0.24] Fix typo in error message (thanks #22 and #28 )
2022-10-25 19:41:46 +02:00
Dr-Noob
5faac7a756
[v0.24] Update PCI ids to pciutils/pciids@06c4c9a
2022-10-25 19:30:24 +02:00
Dr-Noob
8c62e9ebaf
[v0.24] Added generic KBL UHD Graphics. Should fix #19
2022-07-13 13:27:22 +02:00
Dr-Noob
4d948eb80a
[v0.24] Remove CUDA driver initialization message before printing any other message
2022-05-21 23:19:03 +02:00
Dr-Noob
cf96628385
[v0.24] Fix topology for currently supported ALD iGPUs
2022-05-14 20:25:08 +02:00
Dr-Noob
5bf35ee6d7
[v0.24] Make sure we have valid data before reporting peakperf in Intel
2022-05-14 13:12:19 +02:00
Dr-Noob
fea985d08c
[v0.24] Add first support for Alder Lake iGPUs. Needs more work to check data properly
2022-05-14 13:01:34 +02:00
Dr-Noob
24f20d0901
[v0.24] Small fixes; improve PCI report when no GPU is found, speedup invalid GPU idx detection
2022-05-14 12:00:23 +02:00
Dr-Noob
c4ad2bd4f8
[v0.24] Merge bugfix branch
2022-04-17 14:04:19 +02:00
Dr-Noob
af52d2850c
[v0.24] Remove cuda-samples dependency
2022-04-17 13:55:05 +02:00
Dr-Noob
6f196c1797
[v0.23] Fix FreeBSD compilation issues as reported by #13
2022-04-10 16:52:42 +01:00
Dr-Noob
312d78b7f1
[v0.23] Fix dummy warning in intel uarch
2022-04-10 16:11:59 +01:00
Dr-Noob
ebad29e044
[v0.23] Fix CMake to find CUDA Samples in CUDA >= 11.6
2022-03-12 11:04:09 +01:00
Dr-Noob
59df3e53ec
[v0.23] Fix README text. It is written following a C style, but actually written in C++ because of CUDA
2022-01-23 10:57:02 +01:00
Dr-Noob
d120f9a1cd
[v0.23] Add --logo-short/long. Closes #11
2022-01-23 10:55:26 +01:00
Dr-Noob
bd1158c139
[v0.23] Sort PCI devices; this makes the devices list to match CUDA driver ordering, which fixes a bug when there was more than one NVIDIA GPU
2022-01-22 13:25:22 +01:00
Dr-Noob
23586a18e9
[v0.22] Fix for previous commit (dont show tensor cores in TU116)
2022-01-20 22:57:19 +01:00
Dr-Noob
d3aaf7cfe5
[v0.22] Do not show tensor cores in TU116
2022-01-12 19:34:11 +01:00
Dr-Noob
49119ae7eb
[v0.22] Disable pciutils hwdb compilation (useless for gpufetch) to avoid linking against udev
2022-01-12 19:14:56 +01:00
Dr-Noob
4cba0a7194
[v0.22] Round memory size to make output prettier
2022-01-12 18:29:49 +01:00
Dr-Noob
6d9985e5f7
[v0.22] Link against udev, which should fix the error reported by #9
2022-01-11 18:33:49 +01:00
Dr-Noob
0faa7caeee
[v0.22] Add check to properly detect TigerLake GT2 80/96 EUs
2021-12-29 21:56:19 +01:00
Dr-Noob
7f7e70bc5d
[v0.22] Add Gen11 and Gen12 Intel iGPUs (needs more work)
2021-12-28 18:34:56 +01:00
Dr-Noob
6f555f1b47
[v0.22] Small various fixes
2021-12-28 16:43:11 +01:00
Dr-Noob
98a70d5c9e
[v0.21] Print only one error message when the GPU chip is not found in the LUT
2021-12-28 16:21:04 +01:00
Dr-Noob
7ed0e4a63d
[v0.21] Small improvement to argument error reporting
2021-12-28 16:09:39 +01:00
Dr-Noob
9d2a07146a
[v0.21] Check that topology is valid in Intel backend. Print informative message if no valid topology is found
2021-12-28 15:56:44 +01:00
Dr-Noob
8d2f50b398
[v0.21] Print GPU list even when no valid GPU is detected, to improve user understanding
2021-12-28 15:40:29 +01:00
Dr-Noob
8bfe88f9f6
[v0.21] Use MiB to show memory size and do not truncate (may cause problems, as reported in #8 )
2021-12-28 13:44:53 +01:00
Dr-Noob
8fbf97c47a
[v0.21] Add verbose option. Fix CUDA driver initialization message when verbose output is used
2021-12-27 22:37:51 +01:00
Dr-Noob
59f2715149
[v0.21] Print id in hex format for consistency
2021-12-27 22:33:21 +01:00
Dr-Noob
118d9c0b67
[v0.21] Add unamed HD graphics (thanks #7 for reporting)
2021-12-27 18:48:24 +01:00
Dr-Noob
e73f301eef
[v0.20] New license
2021-12-25 11:07:07 +01:00
Dr-Noob
4883bf1ab3
[v0.20] Update README to show Intel iGPU support. Add instructions
2021-12-21 18:32:11 +01:00
Dr-Noob
4921660c24
[v0.20] Properly check GPU index range
2021-12-21 17:09:32 +01:00
Dr-Noob
a20e93f4db
[v0.20] Print help message when no GPU is detected to help people understand whats going on
2021-12-21 17:03:39 +01:00
