Limitations and recommended settings

Limitations and recommended settings#

This section provides information on software and configuration limitations.

Note
For ROCm on Instinct known issues, refer to AMD ROCm Documentation

For OpenMPI limitations, see ROCm UCX OpenMPI on Github

6.2.3 release known issues#

ONNX RT EP will fall back to CPU with Llama2-7B model
Performance drop seen when running separate TensorFlow workloads across multiGPU configurations
Performance drop observed with RetinaNet when using MIGraphX
Hang observed with Ollama or llama.cpp when loading Llama3-70BQ4 on W7900

WSL specific issues#

TDR may be triggered by some long running rocsparse kernels
TDR may be triggered by running rocsolver-test
Text to text generation causes a driver timeout when using torch-2.3.0
Command line scripts may hang when running models that exceed the memory capacity of the GPU
Failures may occur when running ONNX Runtime training scripts

Important!
Radeon™ PRO Series graphics cards are not designed nor recommended for datacenter usage. Use in a datacenter setting may adversely affect manageability, efficiency, reliability, and/or performance. GD-239.

Important!
ROCm is not officially supported on any mobile SKUs.

Multi-GPU configuration#

See mGPU known issues and limitations.

Windows Subsystem for Linux (WSL)#

WSL recommended settings and limitations.

WSL recommended settings#

Optimizing GPU utilization
WSL overhead is a noted bottleneck for GPU utilization. Increasing the batch size of operations will load the GPU more optimally, reducing time required for AI workloads. Optimal batch sizes vary by model, and macro-parameters.

ROCm support in WSL environments#

ROCm-smi support

Due to WSL architectural limitations for native Linux User Kernel Interface (UKI), rocm-smi is not supported.

Issue	Limitations
UKI does not currently support rocm-smi	No current support for: Active compute processes GPU utilization Modifiable state features

ROCm-profiler support

Not currently supported.

Debugger

Not currently supported.

Running PyTorch in virtual environments#

Running PyTorch in virtual environments requires a manual libhsa-runtime64.so update.

When using the WSL usecase and hsa-runtime-rocr4wsl-amdgpu package (installed with PyTorch wheels), users are required to update to a WSL compatible runtime lib.

Solution:

Enter the following commands:

location=`pip show torch | grep Location | awk -F ": " '{print $2}'`
cd ${location}/torch/lib/
rm libhsa-runtime64.so*
cp /opt/rocm/lib/libhsa-runtime64.so.1.2 libhsa-runtime64.so