Install MIGraphX

MIGraphX is AMD’s graph inference engine for optimizing and executing ONNX models on AMD GPUs using ROCm. This page describes how to install MIGraphX 2.16.0.

ROCm version

10.0.0

7.14.1

7.14.0

Installation method

pip

Installation method

Package manager

pip

Prerequisites

MIGraphX is currently supported on:

• gfx950 AMD Instinct MI355X and MI350X

• gfx942 AMD Instinct MI325X and MI300X

• gfx1200, gfx1201, gfx1100, gfx1101, and gfx1102 Radeon GPUs.

See the ROCm compatibility matrix for more information.

MIGraphX is currently supported on gfx950 AMD Instinct MI355X and MI350X data center GPUs and gfx942 MI325X and MI300X GPUs. See the ROCm compatibility matrix for more information.

MIGraphX is currently supported on gfx950 AMD Instinct MI355X and MI350X data center GPUs and gfx942 MI325X and MI300X GPUs. See the ROCm compatibility matrix for more information.

• Ensure your system has Python 3.14 or 3.12 installed and accessible.

Install MIGraphX on Linux

MIGraphX requires ROCm to be installed on your system first.

Install ROCm

For instructions, see Install AMD ROCm 10.0.0. Use the selector panel on that page to view instructions appropriate for your system environment.

For instructions, see Install AMD ROCm 7.14.1. Use the selector panel on that page to view instructions appropriate for your system environment.

For instructions, see Install AMD ROCm 7.14.0. Use the selector panel on that page to view instructions appropriate for your system environment.

Install MIGraphX 2.17.0 via package manager

This method installs MIGraphX system-wide via .deb or .rpm packages.

1. Download the packages.

   Debian-based distro

   wget https://rocm.frameworks-prereleases.amd.com/deb-multi-arch-staging/amdrocm-migraphx_2.17.0%2Brocm10.0.0rc4.becdb3d_amd64.deb
   wget https://rocm.frameworks-prereleases.amd.com/deb-multi-arch-staging/amdrocm-migraphx-dev_2.17.0%2Brocm10.0.0rc4.becdb3d_amd64.deb
   

   RPM-based distro

   wget https://rocm.frameworks-prereleases.amd.com/rpm-multi-arch-staging/amdrocm-migraphx-2.17.0.rocm10.0.0rc4.becdb3d-1.el8.x86_64.rpm
   wget https://rocm.frameworks-prereleases.amd.com/rpm-multi-arch-staging/amdrocm-migraphx-devel-2.17.0.rocm10.0.0rc4.becdb3d-1.el8.x86_64.rpm
   

2. Install the packages. Supported Debian-based Linux distributions are Ubuntu and Debian. Supported RPM-based distributions are RHEL, Oracle Linux, Rocky Linux, and SLES.

   Debian-based distro

   sudo dpkg -i \
     amdrocm-migraphx_2.17.0+rocm10.0.0rc4.becdb3d_amd64.deb \
     amdrocm-migraphx-dev_2.17.0+rocm10.0.0rc4.becdb3d_amd64.deb
   

   RPM-based distro

   sudo rpm -i \
     amdrocm-migraphx-2.17.0.rocm10.0.0rc4.becdb3d-1.el8.x86_64.rpm \
     amdrocm-migraphx-devel-2.17.0.rocm10.0.0rc4.becdb3d-1.el8.x86_64.rpm
   

3. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.

Install MIGraphX 2.16.0 via package manager

This method installs MIGraphX system-wide via .deb or .rpm packages.

1. Download the packages.

   Debian-based distro

   wget https://rocm.frameworks.amd.com/deb-multi-arch/amdrocm-migraphx/pool/main/amdrocm-migraphx_2.16.0-3.py312_amd64.deb
   wget https://rocm.frameworks.amd.com/deb-multi-arch/amdrocm-migraphx/pool/main/amdrocm-migraphx-dev_2.16.0-3.py312_amd64.deb
   

   RPM-based distro

   wget https://rocm.frameworks.amd.com/rpm-multi-arch/amdrocm-migraphx/amdrocm-migraphx-2.16.0-3.x86_64.rpm
   wget https://rocm.frameworks.amd.com/rpm-multi-arch/amdrocm-migraphx/amdrocm-migraphx-devel-2.16.0-3.x86_64.rpm
   

2. Install the packages. Supported Debian-based Linux distributions are Ubuntu and Debian. Supported RPM-based distributions are RHEL, Oracle Linux, Rocky Linux, and SLES.

   Debian-based distro

   sudo dpkg -i \
     amdrocm-migraphx_2.16.0-3.py312_amd64.deb \
     amdrocm-migraphx-dev_2.16.0-3.py312_amd64.deb
   

   RPM-based distro

   sudo rpm -i \
     amdrocm-migraphx-2.16.0-3.x86_64.rpm \
     amdrocm-migraphx-devel-2.16.0-3.x86_64.rpm
   

3. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.

Install MIGraphX 2.16.0 via package manager

This method installs MIGraphX system-wide via .deb or .rpm packages.

1. Download the packages.

   Debian-based distro

   wget https://rocm.frameworks.amd.com/deb-multi-arch/amdrocm-migraphx/pool/main/amdrocm-migraphx_2.16.0-3.py312_amd64.deb
   wget https://rocm.frameworks.amd.com/deb-multi-arch/amdrocm-migraphx/pool/main/amdrocm-migraphx-dev_2.16.0-3.py312_amd64.deb
   

   RPM-based distro

   wget https://rocm.frameworks.amd.com/rpm-multi-arch/amdrocm-migraphx/amdrocm-migraphx-2.16.0-3.x86_64.rpm
   wget https://rocm.frameworks.amd.com/rpm-multi-arch/amdrocm-migraphx/amdrocm-migraphx-devel-2.16.0-3.x86_64.rpm
   

2. Install the packages. Supported Debian-based Linux distributions are Ubuntu and Debian. Supported RPM-based distributions are RHEL, Oracle Linux, Rocky Linux, and SLES.

   Debian-based distro

   sudo dpkg -i \
     amdrocm-migraphx_2.16.0-3.py312_amd64.deb \
     amdrocm-migraphx-dev_2.16.0-3.py312_amd64.deb
   

   RPM-based distro

   sudo rpm -i \
     amdrocm-migraphx-2.16.0-3.x86_64.rpm \
     amdrocm-migraphx-devel-2.16.0-3.x86_64.rpm
   

3. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.

Install MIGraphX 2.17.0 using pip

After installing ROCm, install MIGraphX. This method installs MIGraphX into a Python virtual environment.

1. Create and activate a virtual environment or activate an existing ROCm 10.0.0 environment.

   Python 3.14

   python3.14 -m venv .venv
   source .venv/bin/activate
   

   Python 3.12

   python3.12 -m venv .venv
   source .venv/bin/activate
   

2. Install the MIGraphX and migraphx-libs wheels.

   python -m pip install --index-url https://stable.repo.amd.com/rocm/migraphx/whl-next/ \
       "migraphx==2.17.0+rocm10.0.0"
   

3. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.

Install MIGraphX 2.16.0 using pip

After installing ROCm, install MIGraphX. This method installs MIGraphX into a Python virtual environment.

1. Create and activate a virtual environment or activate an existing ROCm 10.0.0 environment.

   python3.12 -m venv .venv
   source .venv/bin/activate
   

2. Install the MIGraphX and migraphx-libs wheels.

   python -m pip install --index-url https://rocm.frameworks.amd.com/whl-multi-arch/ \
       "migraphx==2.16.0+rocm7.14.1" \
       https://rocm.frameworks.amd.com/whl-multi-arch/migraphx/migraphx_libs-2.16.0%2Brocm7.14.1-py3-none-manylinux_2_28_x86_64.whl
   

3. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.

Install MIGraphX 2.16.0 using pip

After installing ROCm, install MIGraphX. This method installs MIGraphX into a Python virtual environment.

1. Create and activate a virtual environment or activate an existing ROCm 7.14.0 environment. To create a new Python 3.12 virtual environment:

   python3.12 -m venv .venv
   source .venv/bin/activate
   

2. Download the wheel.

   wget https://rocm.frameworks.amd.com/whl-multi-arch/migraphx/migraphx-2.16.0%2Brocm7.14.0-cp312-none-manylinux_2_28_x86_64.whl
   

3. Install the wheels and required packages.

   python -m pip install migraphx-2.16.0+rocm7.14.0-cp312-none-manylinux_2_28_x86_64.whl
   

4. ONNX Runtime accelerates machine learning inference using the MIGraphX execution provider on ROCm-supported GPUs. See the installation guidance.