(Newbi) Please help activate GPU on Ubuntu (Radeon 7700)

[nada391]

Hello ,

i am new to this project and looking for ways to utilize computer responsibly;

have installed this software FAH-Client 8.4 on Ubuntu 24.04;

the GPU is recognized, but seems not supported?

have already the quite latest AMD Driver installed, and OpenCL is active (at least last time i checked it was).


this is the info getting from FAH

Code: Select all

gpu:03:00:00
Description
Navi 33 [Radeon RX 7700S/7600(S)]
Vendor
amd
Supported
false
UUID
PCI Device ID
0x7480
PCI Vendor ID
0x1002
OpenCL
unsupported
CUDA
unsupported

Thank you very much.

[Joe_H]

The guide written up here on installing drivers for AMD GPUs on Linux may be helpful - viewtopic.php?p=367242#p367242

[nada391]

thank you ,
but i have installed the latest drivers as of march 2025
from amd
amdgpu-install_6.3.60304-1_all.deb
(Radeon™ Software for Linux® version 24.30.4 for Ubuntu 24.04.2 HWE)


https://www.amd.com/en/support/download ... ivers.html

now there is another package that states
(Radeon™ Software for Linux® version 24.30.4 for Ubuntu 24.04.2 HWE with ROCm 6.3.4)

i was thinking this package is only for mayor cards like 7900 xt(x) and the likes;


sorry for the confusion, but on the page you linked, the author is stating to install amdgpu_6.2 and not 6.3 and i see no reason to do that since the older version has given problems with newer kernel and made the system broke somehow;

besides, i have followed the instructions for Ubuntu (installing the package rocm-opencl-sdk and checking for groupstatus) but the card still results as unsupported, although OpenCL is installed and should be usable.

Code: Select all

Number of platforms:				 1
  Platform Profile:				 FULL_PROFILE
  Platform Version:				 OpenCL 2.1 AMD-APP (3635.0)
  Platform Name:				 AMD Accelerated Parallel Processing
  Platform Vendor:				 Advanced Micro Devices, Inc.
  Platform Extensions:				 cl_khr_icd cl_amd_event_callback 


  Platform Name:				 AMD Accelerated Parallel Processing
Number of devices:				 2
  Device Type:					 CL_DEVICE_TYPE_GPU
  Vendor ID:					 1002h
  Board name:					 AMD Radeon RX 7600
  Device Topology:				 PCI[ B#3, D#0, F#0 ]
  Max compute units:				 16
  Max work items dimensions:			 3
    Max work items[0]:				 1024
    Max work items[1]:				 1024
    Max work items[2]:				 1024
  Max work group size:				 256
  Preferred vector width char:			 4
  Preferred vector width short:			 2
  Preferred vector width int:			 1
  Preferred vector width long:			 1
  Preferred vector width float:			 1
  Preferred vector width double:		 1
  Native vector width char:			 4
  Native vector width short:			 2
  Native vector width int:			 1
  Native vector width long:			 1
  Native vector width float:			 1
  Native vector width double:			 1
  Max clock frequency:				 2356Mhz
  Address bits:					 64
  Max memory allocation:			 7287183768
  Image support:				 Yes
  Max number of images read arguments:		 128
  Max number of images write arguments:		 8
  Max image 2D width:				 16384
  Max image 2D height:				 16384
  Max image 3D width:				 16384
  Max image 3D height:				 16384
  Max image 3D depth:				 8192
  Max samplers within kernel:			 16
  Max size of kernel argument:			 1024
  Alignment (bits) of base address:		 2048
  Minimum alignment (bytes) for any datatype:	 128
  Single precision floating point capability
    Denorms:					 Yes
    Quiet NaNs:					 Yes
    Round to nearest even:			 Yes
    Round to zero:				 Yes
    Round to +ve and infinity:			 Yes
    IEEE754-2008 fused multiply-add:		 Yes
  Cache type:					 Read/Write
  Cache line size:				 64
  Cache size:					 32768
  Global memory size:				 8573157376
  Constant buffer size:				 7287183768
  Max number of constant args:			 8
  Local memory type:				 Local
  Local memory size:				 65536
  Max pipe arguments:				 16
  Max pipe active reservations:			 16
  Max pipe packet size:				 2992216472
  Max global variable size:			 7287183768
  Max global variable preferred total size:	 8573157376
  Max read/write image args:			 64
  Max on device events:				 1024
  Queue on device max size:			 8388608
  Max on device queues:				 1
  Queue on device preferred size:		 262144
  SVM capabilities:				 
    Coarse grain buffer:			 Yes
    Fine grain buffer:				 Yes
    Fine grain system:				 No
    Atomics:					 No
  Preferred platform atomic alignment:		 0
  Preferred global atomic alignment:		 0
  Preferred local atomic alignment:		 0
  Kernel Preferred work group size multiple:	 32
  Error correction support:			 0
  Unified memory for Host and Device:		 0
  Profiling timer resolution:			 1
  Device endianess:				 Little
  Available:					 Yes
  Compiler available:				 Yes
  Execution capabilities:				 
    Execute OpenCL kernels:			 Yes
    Execute native function:			 No
  Queue on Host properties:				 
    Out-of-Order:				 No
    Profiling :					 Yes
  Queue on Device properties:				 
    Out-of-Order:				 Yes
    Profiling :					 Yes
  Platform ID:					 0x7989e89f0ff0
  Name:						 gfx1102
  Vendor:					 Advanced Micro Devices, Inc.
  Device OpenCL C version:			 OpenCL C 2.0 
  Driver version:				 3635.0 (HSA1.1,LC)
  Profile:					 FULL_PROFILE
  Version:					 OpenCL 2.0 
  Extensions:					 cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_media_ops cl_amd_media_ops2 cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_depth_images cl_amd_copy_buffer_p2p cl_amd_assembly_program 


  Device Type:					 CL_DEVICE_TYPE_GPU
  Vendor ID:					 1002h
  Board name:					 AMD Radeon Graphics
  Device Topology:				 PCI[ B#19, D#0, F#0 ]
  Max compute units:				 1
  Max work items dimensions:			 3
    Max work items[0]:				 1024
    Max work items[1]:				 1024
    Max work items[2]:				 1024
  Max work group size:				 256
  Preferred vector width char:			 4
  Preferred vector width short:			 2
  Preferred vector width int:			 1
  Preferred vector width long:			 1
  Preferred vector width float:			 1
  Preferred vector width double:		 1
  Native vector width char:			 4
  Native vector width short:			 2
  Native vector width int:			 1
  Native vector width long:			 1
  Native vector width float:			 1
  Native vector width double:			 1
  Max clock frequency:				 2200Mhz
  Address bits:					 64
  Max memory allocation:			 13920624024
  Image support:				 Yes
  Max number of images read arguments:		 128
  Max number of images write arguments:		 8
  Max image 2D width:				 16384
  Max image 2D height:				 16384
  Max image 3D width:				 16384
  Max image 3D height:				 16384
  Max image 3D depth:				 8192
  Max samplers within kernel:			 16
  Max size of kernel argument:			 1024
  Alignment (bits) of base address:		 2048
  Minimum alignment (bytes) for any datatype:	 128
  Single precision floating point capability
    Denorms:					 Yes
    Quiet NaNs:					 Yes
    Round to nearest even:			 Yes
    Round to zero:				 Yes
    Round to +ve and infinity:			 Yes
    IEEE754-2008 fused multiply-add:		 Yes
  Cache type:					 Read/Write
  Cache line size:				 128
  Cache size:					 16384
  Global memory size:				 16377204736
  Constant buffer size:				 13920624024
  Max number of constant args:			 8
  Local memory type:				 Local
  Local memory size:				 65536
  Max pipe arguments:				 16
  Max pipe active reservations:			 16
  Max pipe packet size:				 1035722136
  Max global variable size:			 13920624024
  Max global variable preferred total size:	 16377204736
  Max read/write image args:			 64
  Max on device events:				 1024
  Queue on device max size:			 8388608
  Max on device queues:				 1
  Queue on device preferred size:		 262144
  SVM capabilities:				 
    Coarse grain buffer:			 Yes
    Fine grain buffer:				 Yes
    Fine grain system:				 No
    Atomics:					 No
  Preferred platform atomic alignment:		 0
  Preferred global atomic alignment:		 0
  Preferred local atomic alignment:		 0
  Kernel Preferred work group size multiple:	 32
  Error correction support:			 0
  Unified memory for Host and Device:		 1
  Profiling timer resolution:			 1
  Device endianess:				 Little
  Available:					 Yes
  Compiler available:				 Yes
  Execution capabilities:				 
    Execute OpenCL kernels:			 Yes
    Execute native function:			 No
  Queue on Host properties:				 
    Out-of-Order:				 No
    Profiling :					 Yes
  Queue on Device properties:				 
    Out-of-Order:				 Yes
    Profiling :					 Yes
  Platform ID:					 0x7989e89f0ff0
  Name:						 gfx1036
  Vendor:					 Advanced Micro Devices, Inc.
  Device OpenCL C version:			 OpenCL C 2.0 
  Driver version:				 3635.0 (HSA1.1,LC)
  Profile:					 FULL_PROFILE
  Version:					 OpenCL 2.0 
  Extensions:					 cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_media_ops cl_amd_media_ops2 cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_depth_images cl_amd_copy_buffer_p2p cl_amd_assembly_program 

[muziqaz]

thank you ,
but i have installed the latest drivers as of march 2025
from amd
amdgpu-install_6.3.60304-1_all.deb
(Radeon™ Software for Linux® version 24.30.4 for Ubuntu 24.04.2 HWE)


https://www.amd.com/en/support/download ... ivers.html

now there is another package that states
(Radeon™ Software for Linux® version 24.30.4 for Ubuntu 24.04.2 HWE with ROCm 6.3.4)

i was thinking this package is only for mayor cards like 7900 xt(x) and the likes;
<snip>
Version: OpenCL 2.0
Extensions: cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_media_ops cl_amd_media_ops2 cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_depth_images cl_amd_copy_buffer_p2p cl_amd_assembly_program

[/code]

Please read the guide linked by Joe, and follow it, instead of thinking that AMD uses logical thinking
6.3.x branch is for HPC products (Instinct, or supercomputer stuff). I need to talk to AMD again about this nonsense.

6.2.4 works fine with all the latest kernels, I have a system with 6.8 as well as with 6.13.6. Works like a charm

[nada391]

alright, thank you very much both of you, this has now functioned,
had so much trouble with these driver, but for now everything seem settled, very pleased.
6.2.(6) and your installation routine did the trick, it seem stable.
please mark solved.

[muziqaz]

alright, thank you very much both of you, this has now functioned,
had so much trouble with these driver, but for now everything seem settled, very pleased.
6.2.(6) and your installation routine did the trick, it seem stable.
please mark solved.

I'm testing 6.3.x now, and it is not looking good at all. So the guide still stands

[nada391]

had installed according to the AMDGPU docs,

but it bit confusing after all;

https://amdgpu-install.readthedocs.io/e ... lling.html


if i would choose

amdgpu-install --usecase=workstation -y --vulkan=pro --opencl=rocr --accept-eula


(mind the 7600 is no workstation card)

it would function, but Telegram Desktop or even Thunderbird would crash sometime ;


if would revert

to

amdgpu-install --usecase=graphics -y --vulkan=amdvlk, pro --opencl=rocr --accept-eula

the system would be more stable, but FAHclient show not supported;

For getting OpenCL recognized, first had to

sudo usermod -a -G render $LOGNAME
sudo usermod -a -G video $LOGNAME

as per AMD instructions and then reboot; afterwards, clinfo would show card;

now with your 6.2.(x), the above would not function but give amdgpu-dkms error, but this simple one-liner

sudo apt install rocm-opencl-sdk


seemed to take care of it all.

[muziqaz]

I don't play around with use cases. I just go for amdgpu-install and that's all. Figuring out what AMD means by one or the other use case is same as figuring out if there is life in other galaxies

[nada391]

some app like svp4 or handbrake need vulkan pro( for amd-vce) or opencl, and since use consumer cards, found it difficult to understand if usecase=workstation and vulkan=pro would even function on consumer card; it is so many monikers rocr,rocm.vulkan etc bit confusing but it seems the installation routine on li9nux getting better slowely but surely.
as a sidenote, i tested the arm64 version, which would function on raspberry pi3, but not on another sbc (pine a64). there i had to revert to fah client 8.2 , it would then show function, but according to htop not being utilized nevertheless;
installing latest.deb or fah client 8.4 would give consistent error; maybe it is just the sbc pine a64, since it is already an older model, but was markedted as the first 64bit sbc , so should be possible to get it run.

[nada391]

Thank you,

was being a bit too optimistic,
had a lot of crashes (especially with Firefox-Tabs and Thunderbird) (could also be-snap-related, but it pointed to amdgpu), so had to reinstall Ubuntu several times;

now i am back stable on latest 24.04 running HWE Kernel 6.11 (which also supports better recent Processor in my System).

maybe this software you proposing "rocm-opencl-sdk" is not compatible with consumer graphic cards?
on another side of their AMD Driver page it says that


Radeon™ Software for Linux® version 24.30.4 for Ubuntu 24.04.2 HWE with ROCm 6.3.4

or


Radeon™ Software for Linux® version 24.30.4 for Ubuntu 22.04.5 HWE with ROCm 6.3.4


is only for AMD Radeon™ RX 7900 XTX
AMD Radeon™ RX 7900 XT
AMD Radeon™ RX 7900 GRE
AMD Radeon™ PRO W7900
AMD Radeon™ PRO W7900 Dual-Slot
AMD Radeon™ PRO W7800
AMD Radeon™ PRO W7800 48GB

sadly, the FaH-Client does not like it enough;
do i still need to add "rocm-opencl-sdk" into the mix? or it this a lost case with 6.3.4?

Thanks

EDIT:

have found this informative post which explain alittle bit what this rocm-sdk is
https://www.phoronix.com/news/ROCm-SDK-Builder-6.1.2

[muziqaz]

6.2.4 not 6.3

[nada391]

on a fresh installation of Ubuntu 24.04/10
would automatically install HWE enablement stack kernel 6.11.x
and hence Amdgpu 6.2.4/6.2.3 would not compile for kernel;

maybe i am just doing it wrong, but that was my experience;

amdgpu 6.3 would now install the required, drivers, and found out why previously it wouldnt'

since on my computer it must have been "intel xmp" or "amd expo" enabled the culprit since it led to many crashes during compile; after disabling , the install went smooth;

now, i am "back2o" on amdgpu 6.3.x but the rest remains confusing;


why would i need additional package "rocm" for consumer graphic card if opencl is already installed?
is "rocm" not supposed only for high end graphics card?

if "rocm-opencl-sdk" and "rocm-opencl-icd" (as available on Ubuntu 24.10/25.04) is additionally installed, it still would not enable gpu support in fah-client, but now have 3 layers of graphics driver or is system being messed up?

1. amdgpu-install --usecase=graphics,opencl --vulkan
2. rocm-opencl-sdk?
3 rocm-opencl-icd?

should not n.1 suffice?
thanks for the help.

[muziqaz]

viewtopic.php?p=368095#p368095
from step 9.

By the way if you are trying to run HD 7700 on rocm 6.3, AMD does not support that anymore. Latest distros are also a lottery on such old hardware

[nada391]

sorry, it is radeon rx 7600 /rdna3, will take alook.