New OpenCL application for AMD GPUs
Message boards :
News :
New OpenCL application for AMD GPUs
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 16 Nov 22 Posts: 19 Credit: 7,510,317 RAC: 0 |
|
|
Send message Joined: 14 Jun 23 Posts: 91 Credit: 5,914 RAC: 0 |
Last modified: 11 Oct 2023, 17:24:28 UTC Judging by https://asteroidsathome.net/boinc/gpu_list.php your card isn't mentioned. So either not many people are using one here, or nobody with one has managed to complete a task. I'm not sure if that lists every card used or only the popular ones. I see yours last worked on 30 Sept. Not sure if the list I linked to above shows back that far. |
|
Send message Joined: 11 Sep 12 Posts: 12 Credit: 5,812,568 RAC: 0 |
|
|
Send message Joined: 7 Dec 24 Posts: 4 Credit: 1,427,375 RAC: 0 |
I'm having a bit of trouble with Fedora 40 on a Ryzen 7 4700U APU. It will happily request GPU tasks, but they immediately error out on starting: <core_client_version>7.20.2</core_client_version> <![CDATA[ <message> process exited with code 2 (0x2, -254)</message> <stderr_txt> BOINC client version 7.20.2 BOINC GPU type 'ATI', deviceId=0, slot=6 Application: ../../projects/asteroidsathome.net_boinc/period_search_10218_x86_64-pc-linux-gnu__opencl_101_amd_linux Version: 102.18.0.0 Platform name: AMD Accelerated Parallel Processing Platform vendor: Advanced Micro Devices, Inc. OpenCL device C version: OpenCL C 2.0 | OpenCL 2.0 OpenCL device Id: 0 OpenCL device name: gfx900 27GB Device driver version: 3188.4 (HSA1.1,LC) Multiprocessors: 7 Max Samplers: 5686 Max work item dimensions: 3 Resident blocks per multiprocessor: 5686 Grid dim: 79604 = 2 * 7 * 5686 Block dim: 128 Binary build log for gfx900: Error: Failed to compile opencl source (from CL or HIP source to LLVM IR). (-2) Error creating queue: out of host memory(-6) </stderr_txt> ]]> Except that I have 64 GB RAM and have explicitly reserved 8 GB for video in the BIOS. Here's what I get from clinfo: Number of platforms 1
Platform Name AMD Accelerated Parallel Processing
Platform Vendor Advanced Micro Devices, Inc.
Platform Version OpenCL 2.0 AMD-APP (3188.4)
Platform Profile FULL_PROFILE
Platform Extensions cl_khr_icd cl_amd_event_callback
Platform Extensions function suffix AMD
Platform Name AMD Accelerated Parallel Processing
Number of devices 1
Device Name gfx900
Device Vendor Advanced Micro Devices, Inc.
Device Vendor ID 0x1002
Device Version OpenCL 2.0
Driver Version 3188.4 (HSA1.1,LC)
Device OpenCL C Version OpenCL C 2.0
Device Type GPU
Device Board Name (AMD) AMD Radeon Graphics
Device PCI-e ID (AMD) 0x1636
Device Topology (AMD) PCI-E, 0000:05:00.0
Device Profile FULL_PROFILE
Device Available Yes
Compiler Available Yes
Linker Available Yes
Max compute units 7
SIMD per compute unit (AMD) 4
SIMD width (AMD) 16
SIMD instruction width (AMD) 1
Max clock frequency 1600MHz
Graphics IP (AMD) 9.0
Device Partition (core)
Max number of sub-devices 7
Supported partition types None
Supported affinity domains (n/a)
Max work item dimensions 3
Max work item sizes 1024x1024x1024
Max work group size 256
Preferred work group size (AMD) 256
Max work group size (AMD) 1024
=== CL_PROGRAM_BUILD_LOG ===
Error: AMD HSA Code Object loading failed: HSA_STATUS_ERROR_OUT_OF_RESOURCES: The runtime failed to allocate the necessary resources. This error may also occur when the core runtime library needs to spawn threads or create internal OS-specific events.
Preferred work group size multiple (kernel) <getWGsizes:1980: create kernel : error -46>
Wavefront width (AMD) 64
Preferred / native vector sizes
char 4 / 4
short 2 / 2
int 1 / 1
long 1 / 1
half 1 / 1 (cl_khr_fp16)
float 1 / 1
double 1 / 1 (cl_khr_fp64)
Half-precision Floating-point support (cl_khr_fp16)
Denormals No
Infinity and NANs No
Round to nearest No
Round to zero No
Round to infinity No
IEEE754-2008 fused multiply-add No
Support is emulated in software No
Single-precision Floating-point support (core)
Denormals Yes
Infinity and NANs Yes
Round to nearest Yes
Round to zero Yes
Round to infinity Yes
IEEE754-2008 fused multiply-add Yes
Support is emulated in software No
Correctly-rounded divide and sqrt operations Yes
Double-precision Floating-point support (cl_khr_fp64)
Denormals Yes
Infinity and NANs Yes
Round to nearest Yes
Round to zero Yes
Round to infinity Yes
IEEE754-2008 fused multiply-add Yes
Support is emulated in software No
Address bits 64, Little-Endian
Global memory size 29437693952 (27.42GiB)
Global free memory (AMD) 28747748 (27.42GiB) 28747748 (27.42GiB)
Global memory channels (AMD) 4
Global memory banks per channel (AMD) 4
Global memory bank width (AMD) 256 bytes
Error Correction support No
Max memory allocation 25022039859 (23.3GiB)
Unified memory for Host and Device No
Shared Virtual Memory (SVM) capabilities (core)
Coarse-grained buffer sharing Yes
Fine-grained buffer sharing Yes
Fine-grained system sharing No
Atomics No
Minimum alignment for any data type 128 bytes
Alignment of base address 1024 bits (128 bytes)
Preferred alignment for atomics
SVM 0 bytes
Global 0 bytes
Local 0 bytes
Max size for global variable 25022039859 (23.3GiB)
Preferred total size of global vars 29437693952 (27.42GiB)
Global Memory cache type Read/Write
Global Memory cache size 16384 (16KiB)
Global Memory cache line size 64 bytes
Image support Yes
Max number of samplers per kernel 5686
Max size for 1D images from buffer 65536 pixels
Max 1D or 2D image array size 2048 images
Base address alignment for 2D image buffers 256 bytes
Pitch alignment for 2D image buffers 256 pixels
Max 2D image size 16384x16384 pixels
Max 3D image size 2048x2048x2048 pixels
Max number of read image args 128
Max number of write image args 8
Max number of read/write image args 64
Max number of pipe args 16
Max active pipe reservations 16
Max pipe packet size 3547203379 (3.304GiB)
Local memory type Local
Local memory size 65536 (64KiB)
Local memory size per CU (AMD) 65536 (64KiB)
Local memory banks (AMD) 32
Max number of constant args 8
Max constant buffer size 25022039859 (23.3GiB)
Preferred constant buffer size (AMD) 16384 (16KiB)
Max size of kernel argument 1024
Queue properties (on host)
Out-of-order execution No
Profiling Yes
Queue properties (on device)
Out-of-order execution Yes
Profiling Yes
Preferred size 262144 (256KiB)
Max size 8388608 (8MiB)
Max queues on device 1
Max events on device 1024
Prefer user sync for interop Yes
Number of P2P devices (AMD) 0
Profiling timer resolution 1ns
Profiling timer offset since Epoch (AMD) 0ns (Wed Dec 31 18:00:00 1969)
Execution capabilities
Run OpenCL kernels Yes
Run native kernels No
Thread trace supported (AMD) No
Number of async queues (AMD) 8
Max real-time compute queues (AMD) 8
Max real-time compute units (AMD) 7
printf() buffer size 4194304 (4MiB)
Built-in kernels (n/a)
Device 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
NULL platform behavior
clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...) AMD Accelerated Parallel Processing
clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...) Success [AMD]
clCreateContext(NULL, ...) [default] Success [AMD]
clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT) Success (1)
Platform Name AMD Accelerated Parallel Processing
Device Name gfx900
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU) No devices found in platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU) Success (1)
Platform Name AMD Accelerated Parallel Processing
Device Name gfx900
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR) No devices found in platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM) No devices found in platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL) Success (1)
Platform Name AMD Accelerated Parallel Processing
Device Name gfx900
ICD loader properties
ICD loader Name OpenCL ICD Loader
ICD loader Vendor OCL Icd free software
ICD loader Version 2.3.2
ICD loader Profile OpenCL 3.0Could this be some sort of permission issue? |
Keith Myers
Send message Joined: 16 Nov 22 Posts: 179 Credit: 190,786,841 RAC: 63,448 |
|
|
Send message Joined: 1 Jan 13 Posts: 152 Credit: 14,972,857 RAC: 7,658 |
1/ HSA_STATUS_ERROR_OUT_OF_RESOURCES is likely due to outdated drivers https://askubuntu.com/questions/1435708/hsa-status-error-out-of-resources-rocm5-1-1-ubuntu-22-04 could you try more recent drivers if there are any for Vegas? 2/ "out of host memory" means the selected intensity is too high your driver reports CL_DEVICE_MAX_SAMPLERS = 5686, but we expect 16, 32 etc. Grid dim: 79604 = 2 * 7 * 5686 in comparison with 7900 xtx Grid dim: 1536 = 2 * 48 * 16 The driver refuses to compile the application since the required intensity is insanely high. This issue can be resolved in the app by modifying how we calculate the optimal intensity for the target GPU. We should likely avoid using CL_DEVICE_MAX_SAMPLERS in the formula. |
Message boards :
News :
New OpenCL application for AMD GPUs