* Update RELEASE.md and 5.7.0.md to match CHANGELOG.md
* Update 5.2.0.md to match CHANGELOG.md
* Copy CHANGELOG into about folder to match RELEASE
To avoid having divergence in relative links between RELEASE and CHANGELOG
Update docs with information in the AMD blog post announcing support for some RDNA3 Radeon GPUs on Linux.
Co-authored-by: Saad Rahim (AMD) <44449863+saadrahim@users.noreply.github.com>
dependabot mis-detected the repository to be a library
(instead of an application) and widened the rocm-docs-core verison
instead of increasing it. This basically disabled pinning.
Explicitly specify to increase the version instead of widening it
to hopefully prevent this in the future.
Flattened out page structure for improved navigability.
* Change Table of Contents
* Update the install guides for windows and linux
* Removed extraneous index pages
* GPU architecture pages duplicate entries removed
* spack page cleanup
---------
Co-authored-by: Sam Wu <samwu103@amd.com>
Co-authored-by: Saad Rahim (AMD) <44449863+saadrahim@users.noreply.github.com>
Since we are not installing the ROCm OpenCL packages. We are not able to
test ROCm withg this command.
Signed-off-by: Tasso Zambelakis <Tasso.Zambelakis@amd.com>
* RX 6700* doc fixes in windows_support.md
Correct RX 6700* LLVM target to gfx1031 windows_support.md
Change name from "RX 6750" to "RX 6750 XT"
* Fix RX7600 LLVM to gfx1102 in windows-support.md
---------
Co-authored-by: Saad Rahim (AMD) <44449863+saadrahim@users.noreply.github.com>
This should increase usability and prevent errors, since the most common
use case is the user using the latest version of their OS,
rather than the oldest supported one.
* update relative link to llvm asan guide
remove docs dir from path
* Minor typo and update on supported OSes
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* Update install instructions to 5.7
* RTG additions to install instructions
* update install instructions for multi version
---------
Co-authored-by: Máté Ferenc Nagy-Egri <mate@streamhpc.com>
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* Update using-gpu-sanitizer.md
Updated content
* fixes for markdown linting
use * instead of + for lists
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* Create using_gpu_sanitizer.md
* Created GPU Sanitizer File and Title
* add technical terms to wordlist and fix spelling
* spelling
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
Co-authored-by: b-sumner <brian.sumner@amd.com>
* Added deleted sections to openmp.md and other improvements
* Update CONTRIBUTING.md
* add example of snake case
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* spell out HPC acronym in explanation doc
* update toolchain docs
order in importance descending
* update Contributing guide
add discussions
update formatting and grammar
* separate contributing section for readability
* fix formatting for mdl
* fix spelling
* Update Links (#2240)
* update link to PCIe Gen 4 pdf
* fix broken links
* remove references to broken links
* fix spelling of data center
* Fixing HIP link (#2236)
* Swati develop (#2245)
* Added deleted sections to openmp.md and other improvements
* Update openmp.md
Tagged `ICV`
* Solving indiscrepencies in openmp.md
There are apparently differences in the published document and information conveyed by the Dev. Fixed it.
* add new words to wordlist
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* fix rocm_smi_lib link in toc (#2260)
* ROCm FHS Reorganization, Backward Compatibility, and Versioning - rev (#2255)
* update requirements
---------
Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
Co-authored-by: srawat <120587655+SwRaw@users.noreply.github.com>
Co-authored-by: Ehud Sharlin <112672820+Ehud-Sharlin@users.noreply.github.com>
* docs: update changelog and release notes with hipStreamGetDevice
* docs: fix typos and add version update notes
* docs: add HIP changelog
* remove What's New section from changelog
* docs: clean up SLES tab-sets
- Always use a tab-set for SLES 15.4
- In the toplevel SLES title don't say version 15
- harmonize the `:sync:` labels between documents
* docs: Misc fixes in installation
- Fix rocm repository url in the installer script installation for SLES
- Add a missing :sync: tab in installation prerequisites
* docs: add SLES 15.5 support to installation and OS support pages
* Added deleted sections to openmp.md and other improvements
* Update openmp.md
Tagged `ICV`
* Solving indiscrepencies in openmp.md
There are apparently differences in the published document and information conveyed by the Dev. Fixed it.
* add new words to wordlist
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* Update tensorflow_install.md
Fixed writing commands so that when executed by copy paste it doesn't cause an error.
* Update tensorflow_install.md
Following @saadrahim's suggestion of using "\" to signify a line break in bash.
* Remove package pin from quick start quide
When installing a single-package fashion, no version pinning is needed
* Add package pinning to quick start guide
Pinning the packages is required to make apt prefer the rocm packages
instead of the system ones when both provide the same package (e.g
`rocm-smi`).
* Removing Ubuntu 20.04 change
---------
Co-authored-by: Gergely Meszaros <gergely@streamhpc.com>
Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
* Added specialized kernels to openmp.md
A few formatting changes and addition of specialized kernels section at the end.
* Added Specialized kernels in openmp.md
Some formatting changes and addition of specialized kernels instead of no loop and cross team kernels
* Added specialized kernel to openmp.md
* Added specialized kernels to openmp.md
* Replaced the usage of uncertain clauses(may/might) in openmp.md
* Attempt to align the table headings for environment variables in openmp.md
* Feedback from Dhruva
---------
Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
The hip-devel package depends on perl modules not distributed by default
on RHEL and SLES distriubutions, these can be installed from EPEL and
the `devel:languages:perl` repository respectively.
Ideally in the future these dependencies would be replaced with packages
available from default repositories, but in the meanwhile this should
be at least documented.
* Remove install instructions for unsuported RHEL 8.8 and 9.2
Current ROCm release does not support these versions of RHEL
* Centralize disclaimers and perquisites for installation
- Move the single-version to multi-version diclaimer to the install
overview page where single vs multi installs are discussed.
- Move the installation of kernel-headers and development packages
to the install preparation page. Unify it mainly from the quick start
content.
* s/Name/name/ in repository config files for RHEL
The repository name can be set as `name=><name>` instead of `Name`,
otherwise yum complains about the repo not having a name, e.g:
```output
Repository 'ROCm-5.3.3' is missing name in configuration, using id.
```
This is fixed with this commit.
* Clean up render/video group section on prerequisites
* Installation and Upgrade restructuring & fixes
- Fix the rocm package urls for RHEL in the install & upgrade guides
- RHEL8 and 9 have different URLs, add a tab-set similar to ubuntu
for them.
- Fix the package URL in the upgrade guide for SLES (previously pointed
to the amdgpu url)
- Change the apt-signing key download and conversion to the method used
in the quick start guide, which is the recommended by ubuntu maintainers
- Change the install steps from list items to rubrics with numbered entries
which is more readable and matches the style in the quick start guide
- Do not pass `--append` to `tee` in the upgrade guide, because it is
meant to overwrite.
- Split the one long tab-set to multiple tab-sets in the upgrade guide
to improve readability
Add empty cells to list tables to make them uniform (all rows have the
same number of cells), before this the tables errored out with:
> ERROR: Error parsing content block for the "list-table" directive:
> uniform two-level bullet list expected, but row 13 does not contain
> the same number of items as row 1 (3 vs 4)
and the table did not show up.
* ci: change markdown linting to use the NodeJs markdownlint
The original ruby based markdownlint has a few shortcomings not known
when it was introduced:
- no support for myst extensions
- no support for disabling specific rules for specific files or regions
These two combined make it very hard to use when used for this project
when it has false positives around myst extensions.
Luckily there's a NodeJS based version of markdownlint [1] supporting the
same ruleset that is more configurable:
- seems to support myst extensions better
- has an html comment based syntax to disable specific rules
The library seem to be better maintained too and with better tooling:
e.g. there's a vscode extension using the engine for local use:
markdownlint (DavidAnson.vscode-markdownlint).
[1]: https://github.com/DavidAnson/markdownlint
* docs: hotfix empty links
There are missing links in the docs, these should get fixed, but for now
they are just monkey patched to make CI happy.
* docs: fix links
---------
Co-authored-by: Nara Prasetya <nara@streamhpc.com>
* update the gpu-aware-mpi page
Three changes:
- add the ucx compatibility table
- add the --with-rocm=/opt/rocm option to the compilation of Open MPI
- add a section about how to compile and use UCC for collective
operations.
* Changing link to relative
* Update gpu_aware_mpi.md
---------
Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
* fix rocmcc link
* remove unused link
* remove unused linkcheck configs
* update amd smi section
add link to ami smi github
---------
Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
* update links to new docs and rename .sphinx dir to sphinx
* fix spelling and formatting
add new words to wordlist
remove empty headers
remove version number for ROCm in conf.py
fix typos
* add more formats to rtd config
* Clean up the deployment related pages
- Add an index page for the linux deployment submenu
- Remove deployment options that are not yet completed (i.e. spack,
from source installation)
- remove the general deployment index page
- various cleanups and clarifications in the rest of the pages
* Move all deploy pages to deploy folder
---------
Co-authored-by: Gergely Meszaros <gergely@streamhpc.com>
- Unify code block style (indent vs. fence)
- Mark code languages
- Increase heading level one at a time
- No extra newlines between paragraphs
- List for header reorg stages
- Shrink ascii table (mobile friendlyness)
- 80-column width
* add url to ROCgdb-docs
update reqs and gitignore
* add validation tools section for RVS and TransferBench
* stub in links for validation/mgmt tools
* populate compilers page
* add cards for ai libs and computer vision pages
* add content to math lib pages
* reorg hip and math libs
* update index
* consolidate linear algebra libs
* fix release info order in toc
* fix links and content cards for libraries
* update mdl ignored files
* update understand rocm section
* fix formatting errors
* add link to openmp
* ignore md041
* add url to ROCgdb-docs
update reqs and gitignore
* add validation tools section for RVS and TransferBench
* stub in links for validation/mgmt tools
* populate compilers page
* add cards for ai libs and computer vision pages
* add content to math lib pages
* reorg hip and math libs
* update index
* consolidate linear algebra libs
* Pulling libraries out
* add libraries listed in left sidebar to index page
* Adding all
* Updating nav tree
* fix link to rocm-examples in toc
* update TOC
---------
Co-authored-by: Sam Wu <sam.wu2@amd.com>
* Add C++ algorithm primitive lib cards
* Add PRNG section
* API Reference Manuals first
* Add Tensile and rocWMMA
* Change rocFFT and hipFFT order for consistency
* Add RCCL
* Fix PRNG links
* Add rocSOLVER and hipSOLVER
* Add general note on rocLIB vs hipLIB
* linux quick start: Mention correct package to install
* linux quick start: Rephrase prerequisites
Mention that installing the headers is usually not required by hand.
* linux quick start: Simplify command to get singing key
* linux quick start: Add instructions for RHEL and SLES
Reorganize the quick start guide for linux, adding multi level
tab selection for just the commands where it makes sense.
Currently mostly Ubuntu commands are filled out, if the structure
looks fine, then more will follow.
Add more of the sphinx generated files, so generating the docs does not
add untracked files. Ignore the folder `.venv` typically used for
virtual environments.
Also sort the ignored file list for easier maintenance.
The screenshots are from tables with text, which are not easily searchable,
are bigger in size than needed – increasing load times – and are in a
resolution, causing them to be blurry on HiDPI displays. Therefore, use a
Markdown table instead solving all the issues above, and delete the images
from the repository.
The SLES service pack version differs in the two screenshots: SP2 vs SP3.
Go for *SP3*.
Resolves: https://github.com/RadeonOpenCompute/ROCm/issues/1591
This document describes the features, fixed issues, and information about downloading and installing the AMD ROCm™ software. It also covers known issues and deprecations in the AMD ROCm v4.1 release.
- [Supported Operating Systems and Documentation Updates](#Supported-Operating-Systems-and-Documentation-Updates)
* [Compiler Generated Code Object Version 2 Deprecation ](#Compiler-Generated-Code-Object-Version-2-Deprecation)
- [Deploying ROCm](#Deploying-ROCm)
- [Hardware and Software Support](#Hardware-and-Software-Support)
- [Machine Learning and High Performance Computing Software Stack for AMD GPU](#Machine-Learning-and-High-Performance-Computing-Software-Stack-for-AMD-GPU)
* CentOS 8.3 (4.18.0-193.el8) and RHEL 8.3 (4.18.0-193.1.1.el8) (devtoolset is not required)
* SLES 15 SP2
### FRESH INSTALLATION OF AMD ROCM V4.1 RECOMMENDED
A complete uninstallation of previous ROCm versions is required before installing a new version of ROCm. An upgrade from previous releases to AMD ROCm v4.1 is not supported. For more information, refer to the AMD ROCm Installation Guide at
**Note**: AMD ROCm release v3.3 or prior releases are not fully compatible with AMD ROCm v3.5 and higher versions. You must perform a fresh ROCm installation if you want to upgrade from AMD ROCm v3.3 or older to 3.5 or higher versions and vice-versa.
**Note**: *render* group is required only for Ubuntu v20.04. For all other ROCm supported operating systems, continue to use video group.
* For ROCm v3.5 and releases thereafter, the clinfo path is changed to /opt/rocm/opencl/bin/clinfo.
* For ROCm v3.3 and older releases, the clinfo path remains /opt/rocm/opencl/bin/x86_64/clinfo.
### ROCM MULTI-VERSION INSTALLATION UPDATE
With the AMD ROCm v4.1 release, the following ROCm multi-version installation changes apply:
The meta packages rocm-dkms<version> are now deprecated for multi-version ROCm installs. For example, rocm-dkms3.7.0, rocm-dkms3.8.0.
* Multi-version installation of ROCm should be performed by installing rocm-dev<version> using each of the desired ROCm versions. For example, rocm-dev3.7.0, rocm-dev3.8.0, rocm-dev3.9.0.
* Version files must be created for each multi-version rocm <= 4.1.0
* Command: echo <version> | sudo tee /opt/rocm-<version>/.info/version
* Example: echo 4.1.0 | sudo tee /opt/rocm-4.1.0/.info/version
* The rock-dkms loadable kernel modules should be installed using a single rock-dkms package.
* ROCm v3.9 and above will not set any ldconfig entries for ROCm libraries for multi-version installation. Users must set LD_LIBRARY_PATH to load the ROCm library version of choice.
**NOTE**: The single version installation of the ROCm stack remains the same. The rocm-dkms package can be used for single version installs and is not deprecated at this time.
# Driver Compatibility Issue in This Release
In certain scenarios, the ROCm 4.1 run-time and userspace environment are not compatible with ROCm v4.0 and older driver implementations for 7nm-based (Vega 20) hardware (MI50 and MI60).
To mitigate issues, the ROCm v4.1 or newer userspace prevents running older drivers for these GPUs.
Users are notified in the following scenarios:
* Bare Metal
* Containers
## Bare Metal
In the bare-metal environment, the following error message displays in the console:
To test the compatibility, run the ROCm v4.1 version of rocminfo using the following instruction:
*/opt/rocm-4.1.0/bin/rocminfo 2>&1 | less*
## Containers
A container (built with error detection for this issue) using a ROCm v4.1 or newer run-time is initiated to execute on an older kernel. The container fails to start and the following warning appears:
*Error: Incompatible ROCm environment. The Docker container requires the latest kernel driver to operate correctly.
Upgrade the ROCm kernel to v4.1 or newer, or use a container tagged for v4.0.1 or older.*
To inspect the version of the installed kernel driver, run either:
* dpkg --status rock-dkms [Debian-based]
or
* rpm -ql rock-dkms [RHEL, SUSE, and others]
To install or update the driver, follow the installation instructions at:
The new TargetID functionality allows compilations to specify various configurations of the supported hardware.
Previously, ROCm supported only a single configuration per target.
With the TargetID enhancement, ROCm supports configurations for Linux, PAL and associated configurations such as XNACK. This feature addresses configurations for the same target in different modes and allows applications to build executables that specify the supported configurations, including the option to be agnostic for the desired setting.
### New Code Object Format Version for TargetID
* A new clang option -mcode-object-version can be used to request the legacy code object version 3 or code object version 2. For more information, refer to
* A new clang --offload-arch= option is introduced to specify the offload target architecture(s) for the HIP language.
* The clang --offload-arch= and -mcpu options accept a new Target ID syntax. This allows both the processor and target feature settings to be specified. For more details, refer to
* ROCm v4.1 corrects code object loading to enforce target feature settings of the code object to match the setting of the agent. It also corrects the recording of target feature settings in the code object. As a consequence, the legacy code objects may no longer load due to mismatches.
* gfx802, gfx803, and gfx805 do not support the XNACK target feature in the ROCm v4.1 release.
### New Code Object Tools
AMD ROCm v4.1 provides new code object tools *roc-obj-ls* and *roc-obj-extract*. These tools allow for the listing and extraction of AMD GPU ROCm code objects that are embedded in HIP executables and shared objects. Each tool supports a --help option that provides more information.
Refer to the HIP Programming Guide v4.1 for additional information and examples.
The ROCm Data Center (RDC) Tool is enhanced with the Grafana plugin. Grafana is a common monitoring stack used for storing and visualizing time series data. Prometheus acts as the storage backend, and Grafana is used as the interface for analysis and visualization. Grafana has a plethora of visualization options and can be integrated with Prometheus for the ROCm Data Center (RDC) dashboard.
For more information about Grafana integration and installation, refer to the ROCm Data Center Tool User guide at:
* generic API including axpby, gather, scatter, rot, spvv, spmv, spgemm, sparsetodense, densetosparse
* mixed indexing types in matrix formats
For more information, see
https://rocsparse.readthedocs.io/en/latest/
### rocSOLVER
rocSOLVER extends support for:
* Eigensolver routines for symmetric/hermitian matrices:
- STERF, STEQR
* Linear solvers for general non-square systems:
- GELS (API added with batched and strided_batched versions. Only the overdetermined non-transpose case is implemented in
this release. Other cases will return rocblas_status_not_implemented status for now.)
* Extended test coverage for functions returning information
* Changelog file
* Tridiagonalization routines for symmetric and hermitian matrices:
- LATRD
- SYTD2, SYTRD (with batched and strided_batched versions)
- HETD2, HETRD (with batched and strided_batched versions)
* Sample code and unit test for unified memory model/Heterogeneous Memory Management (HMM)
For more information, see
https://rocsolver.readthedocs.io/en/latest/
### hipCUB
The new iterator DiscardOutputIterator in hipCUB represents a special kind of pointer that ignores values written to it upon dereference. It is useful for ignoring the output of certain algorithms without wasting memory capacity or bandwidth. DiscardOutputIterator may also be used to count the size of an algorithm's output, which was not known previously.
For more information, see
https://hipcub.readthedocs.io/en/latest/
## HIP Enhancements
### Support for hipEventDisableTiming Flag
HIP now supports the hipEventDisableTiming flag for hipEventCreateWithFlags. Note, events created with this flag do not record profiling data and provide optimal performance when used for synchronization.
### Cooperative Group Functions
Cooperative Groups defines, synchronizes, and communicates between groups of threads and blocks for efficiency and ease of management. HIP now supports the following kernel language Cooperative Groups types and functions:
Previously, it was required to declare dynamic shared memory using the HIP_DYNAMIC_SHARED macro for accuracy as using static shared memory in the same kernel could result in overlapping memory ranges and data-races.
Now, the HIP-Clang compiler provides support for extern shared declarations, and the HIP_DYNAMIC_SHARED option is no longer required.
You may use the standard extern definition:
extern __shared__ type var[];
## OpenMP Enhancements and Fixes
This release includes the following OpenMP changes:
* Usability Enhancements
* Fixes to Internal Clang Math Headers
* OpenMP Defect Fixes
### Usability Enhancements
* OMPD updates for flang
* To support OpenMP debugging, the selected OpenMP runtime sources are included in lib-debug/src/openmp. The ROCgdb debugger
will find these automatically.
* Threadsafe hsa plugin for libomptarget
* Support multiple devices with malloc and hostrpc
* Improve hostrpc version check
* Add max reduction offload feature to flang
* Integration of changes to support HPC Toolkit
* Support for fprintf
* Initial support for GPU malloc and Free. The internal (device rtl) is required for GPU malloc and Free for nested parallelism.
GPU malloc and Free are now replaced, which improves the device memory footprint.
* Increase detail of debug printing controlled by LIBOMPTARGET_KERNEL_TRACE environment variable
* Add support for -gpubnames in Flang Driver
* Increase detail of debug printing controlled by LIBOMPTARGET_KERNEL_TRACE environment variable
* Add support for -gpubnames in Flang Driver
### Fixes to Internal Clang Math Headers
This release includes a set of changes applied to Clang internal headers to support OpenMP C, C++, FORTRAN, and HIP C. This establishes consistency between NVPTX and AMDGCN offloading, and OpenMP, HIP, and CUDA. OpenMP uses function variants and header overlays to define device versions of functions. This causes Clang LLVM IR codegen to mangle names of variants in both the definition and callsites of functions defined in the internal Clang headers. The changes apply to headers found in the installation subdirectory lib/clang/11.0.0/include.
The changes also temporarily eliminate the use of the libm bitcode libraries for C and C++. Although math functions are now defined with internal clang headers, a bitcode library of the C functions defined in the headers is still built for the FORTRAN toolchain linking. This is because FORTRAN cannot use C math headers. This bitcode library is installed in lib/libdevice/libm-.bc. The source build of the bitcode library is implemented with the aomp-extras repository and the component-built script build_extras.sh.
### OpenMP Defect Fixes
The following OpenMP defects are fixed in this release:
* Openmpi configuration issue with real16.
* [flang] The AOMP 11.7-1 Fortran compiler claims to support the -isystem flag, but ignores it.
* [flang] producing internal compiler error when the character is used with KIND.
* [flang] openmp map clause on complex allocatable expressions !$omp target data map( chunk%tiles(1)%field%density0).
* Add a fatal error if missing -Xopenmp-target or -march options when -fopenmp-targets is specified. However, this requirement is not
applicable for offloading to the host when there is only a single target and that target is the host.
* Openmp error message output for no_rocm_device_lib was asserting.
* Linkage on constant per-kernel symbols from external to weaklinkageonly to prevent duplicate symbols when building kokkos.
* Add environment variables ROCM_LLD_ARGS ROCM_LINK_ARGS ROCM_SELECT_ARGS to test driver options without compiler rebuild.
* Fix problems with device math functions being ambiguous, especially the pow function.ix aompcc to accept file type cxx.
* Fix a latent race between host runtime and devicertl.
## MIOPEN TENSILE INTEGRATION
MIOpenTensile provides host-callable interfaces to the Tensile library and supports the HIP programming model. You may use the Tensile feature in the HIP backend by setting the building environment variable value to ON.
MIOPEN_USE_MIOPENTENSILE=ON
MIOpenTensile is an open-source collaboration tool where external entities can submit source pull requests (PRs) for updates. MIOpenTensile maintainers review and approve the PRs using standard open-source practices.
For more information about the sources and the build system, see
The following are the known issues in this release.
## Upgrade to AMD ROCm v4.1 Not Supported
An upgrade from previous releases to AMD ROCm v4.1 is not supported. A complete uninstallation of previous ROCm versions is required before installing a new version of ROCm.
## Performance Impact for Kernel Launch Bound Attribute
Kernels without the *__launch_bounds__* attribute assume the default maximum threads per block value. In the previous ROCm release, this value was 256. In the ROCm v4.1 release, it is changed to 1024. The objective of this change ensures the actual threads per block value used to launch a kernel, by default, are always within the launch bounds, thus, establishing the correctness of HIP programs.
**NOTE**: Using the above-mentioned approach may incur performance degradation in certain cases. Users must add a minimum launch bound to each kernel, which covers all possible threads per block values used to launch that kernel for correctness and performance.
The recommended workaround to recover the performance is to add *--gpu-max-threads-per-block=256* to the compilation options for HIP programs.
## Issue with Passing a Subset of GPUs in a Multi-GPU System
ROCm support for passing individual GPUs via the docker --device flag in a Docker run command has a known issue when passing a subset of GPUs in a multi-GPU system. The command runs without any warning or error notification. However, all GPU executable run outputs are randomly corrupted.
Using GPU targeting via the Docker command is not recommended for users of ROCm 4.1. There is no workaround for this issue currently.
## Performance Impact for LDS-Bound Kernels
The compiler in ROCm v4.1 generates LDS load and stores instructions that incorrectly assume equal performance between aligned and misaligned accesses. While this does not impact code correctness, it may result in sub-optimal performance.
This issue is under investigation, and there is no known workaround at this time.
# Deprecations
This section describes deprecations and removals in AMD ROCm.
## Compiler Generated Code Object Version 2 Deprecation
Compiler-generated code object version 2 is no longer supported and has been completely removed. Support for loading code object version 2 is also deprecated with no announced removal release.
# Deploying ROCm
AMD hosts both Debian and RPM repositories for the ROCm packages.
For more information on ROCM installation on all platforms, see
ROCm is focused on using AMD GPUs to accelerate computational tasks such as machine learning, engineering workloads, and scientific computing.
In order to focus our development efforts on these domains of interest, ROCm supports a targeted set of hardware configurations which are detailed further in this section.
**Note:** The AMD ROCm™ open software platform is a compute stack for headless system deployments. GUI-based software applications are currently not supported.
#### Supported GPUs
Because the ROCm Platform has a focus on particular computational domains, we offer official support for a selection of AMD GPUs that are designed to offer good performance and price in these domains.
**Note:** The integrated GPUs of Ryzen are not officially supported targets for ROCm.
ROCm officially supports AMD GPUs that use following chips:
* GFX9 GPUs
- "Vega 10" chips, such as on the AMD Radeon RX Vega 64 and Radeon Instinct MI25
- "Vega 7nm" chips, such as on the Radeon Instinct MI50, Radeon Instinct MI60 or AMD Radeon VII, Radeon Pro VII
* CDNA GPUs
- MI100 chips such as on the AMD Instinct™ MI100
ROCm is a collection of software ranging from drivers and runtimes to libraries and developer tools.
Some of this software may work with more GPUs than the "officially supported" list above, though AMD does not make any official claims of support for these devices on the ROCm software platform.
The following list of GPUs are enabled in the ROCm software, though full support is not guaranteed:
* GFX8 GPUs
* "Polaris 11" chips, such as on the AMD Radeon RX 570 and Radeon Pro WX 4100
* "Polaris 12" chips, such as on the AMD Radeon RX 550 and Radeon RX 540
* GFX7 GPUs
* "Hawaii" chips, such as the AMD Radeon R9 390X and FirePro W9100
As described in the next section, GFX8 GPUs require PCI Express 3.0 (PCIe 3.0) with support for PCIe atomics. This requires both CPU and motherboard support. GFX9 GPUs require PCIe 3.0 with support for PCIe atomics by default, but they can operate in most cases without this capability.
The integrated GPUs in AMD APUs are not officially supported targets for ROCm.
As described [below](#limited-support), "Carrizo", "Bristol Ridge", and "Raven Ridge" APUs are enabled in our upstream drivers and the ROCm OpenCL runtime.
However, they are not enabled in the HIP runtime, and may not work due to motherboard or OEM hardware limitations.
As such, they are not yet officially supported targets for ROCm.
For a more detailed list of hardware support, please see [the following documentation](https://en.wikipedia.org/wiki/List_of_AMD_graphics_processing_units).
#### Supported CPUs
As described above, GFX8 GPUs require PCIe 3.0 with PCIe atomics in order to run ROCm.
In particular, the CPU and every active PCIe point between the CPU and GPU require support for PCIe 3.0 and PCIe atomics.
The CPU root must indicate PCIe AtomicOp Completion capabilities and any intermediate switch must indicate PCIe AtomicOp Routing capabilities.
Current CPUs which support PCIe Gen3 + PCIe Atomics are:
* AMD Ryzen CPUs
* The CPUs in AMD Ryzen APUs
* AMD Ryzen Threadripper CPUs
* AMD EPYC CPUs
* Intel Xeon E7 v3 or newer CPUs
* Intel Xeon E5 v3 or newer CPUs
* Intel Xeon E3 v3 or newer CPUs
* Intel Core i7 v4, Core i5 v4, Core i3 v4 or newer CPUs (i.e. Haswell family or newer)
* Some Ivy Bridge-E systems
Beginning with ROCm 1.8, GFX9 GPUs (such as Vega 10) no longer require PCIe atomics.
We have similarly opened up more options for number of PCIe lanes.
GFX9 GPUs can now be run on CPUs without PCIe atomics and on older PCIe generations, such as PCIe 2.0.
This is not supported on GPUs below GFX9, e.g. GFX8 cards in the Fiji and Polaris families.
If you are using any PCIe switches in your system, please note that PCIe Atomics are only supported on some switches, such as Broadcom PLX.
When you install your GPUs, make sure you install them in a PCIe 3.1.0 x16, x8, x4, or x1 slot attached either directly to the CPU's Root I/O controller or via a PCIe switch directly attached to the CPU's Root I/O controller.
In our experience, many issues stem from trying to use consumer motherboards which provide physical x16 connectors that are electrically connected as e.g. PCIe 2.0 x4, PCIe slots connected via the Southbridge PCIe I/O controller, or PCIe slots connected through a PCIe switch that does
not support PCIe atomics.
If you attempt to run ROCm on a system without proper PCIe atomic support, you may see an error in the kernel log (`dmesg`):
Experimental support for our Hawaii (GFX7) GPUs (Radeon R9 290, R9 390, FirePro W9100, S9150, S9170)
does not require or take advantage of PCIe Atomics. However, we still recommend that you use a CPU
from the list provided above for compatibility purposes.
#### Not supported or limited support under ROCm
##### Limited support
* ROCm 4.x should support PCIe 2.0 enabled CPUs such as the AMD Opteron, Phenom, Phenom II, Athlon, Athlon X2, Athlon II and older Intel Xeon and Intel Core Architecture and Pentium CPUs. However, we have done very limited testing on these configurations, since our test farm has been catering to CPUs listed above. This is where we need community support. _If you find problems on such setups, please report these issues_.
* Thunderbolt 1, 2, and 3 enabled breakout boxes should now be able to work with ROCm. Thunderbolt 1 and 2 are PCIe 2.0 based, and thus are only supported with GPUs that do not require PCIe 3.1.0 atomics (e.g. Vega 10). However, we have done no testing on this configuration and would need community support due to limited access to this type of equipment.
* AMD "Carrizo" and "Bristol Ridge" APUs are enabled to run OpenCL, but do not yet support HIP or our libraries built on top of these compilers and runtimes.
* As of ROCm 2.1, "Carrizo" and "Bristol Ridge" require the use of upstream kernel drivers.
* In addition, various "Carrizo" and "Bristol Ridge" platforms may not work due to OEM and ODM choices when it comes to key configurations parameters such as inclusion of the required CRAT tables and IOMMU configuration parameters in the system BIOS.
* Before purchasing such a system for ROCm, please verify that the BIOS provides an option for enabling IOMMUv2 and that the system BIOS properly exposes the correct CRAT table. Inquire with your vendor about the latter.
* AMD "Raven Ridge" APUs are enabled to run OpenCL, but do not yet support HIP or our libraries built on top of these compilers and runtimes.
* As of ROCm 2.1, "Raven Ridge" requires the use of upstream kernel drivers.
* In addition, various "Raven Ridge" platforms may not work due to OEM and ODM choices when it comes to key configurations parameters such as inclusion of the required CRAT tables and IOMMU configuration parameters in the system BIOS.
* Before purchasing such a system for ROCm, please verify that the BIOS provides an option for enabling IOMMUv2 and that the system BIOS properly exposes the correct CRAT table. Inquire with your vendor about the latter.
##### Not supported
* "Tonga", "Iceland", "Vega M", and "Vega 12" GPUs are not supported.
* We do not support GFX8-class GPUs (Fiji, Polaris, etc.) on CPUs that do not have PCIe 3.0 with PCIe atomics.
* As such, we do not support AMD Carrizo and Kaveri APUs as hosts for such GPUs.
* Thunderbolt 1 and 2 enabled GPUs are not supported by GFX8 GPUs on ROCm. Thunderbolt 1 & 2 are based on PCIe 2.0.
In the default ROCm configuration, GFX8 and GFX9 GPUs require PCI Express 3.0 with PCIe atomics. The ROCm platform leverages these advanced capabilities to allow features such as user-level submission of work from the host to the GPU. This includes PCIe atomic Fetch and Add, Compare and Swap, Unconditional Swap, and AtomicOp Completion.
#### ROCm support in upstream Linux kernels
As of ROCm 1.9.0, the ROCm user-level software is compatible with the AMD drivers in certain upstream Linux kernels.
As such, users have the option of either using the ROCK kernel driver that are part of AMD's ROCm repositories or using the upstream driver and only installing ROCm user-level utilities from AMD's ROCm repositories.
These releases of the upstream Linux kernel support the following GPUs in ROCm:
| Pros | More GPU features, and they are enabled earlier | Includes the latest Linux kernel features |
| | Tested by AMD on supported distributions | May work on other distributions and with custom kernels |
| | Supported GPUs enabled regardless of kernel version | |
| | Includes the latest GPU firmware | |
| Cons | May not work on all Linux distributions or versions | Features and hardware support varies depending on kernel version |
| | Not currently supported on kernels newer than 5.4 | Limits GPU's usage of system memory to 3/8 of system memory (before 5.6). For 5.6 and beyond, both DKMS and upstream kernels allow use of 15/16 of system memory. |
| | | IPC and RDMA capabilities are not yet enabled |
| | | Not tested by AMD to the same level as `rock-dkms` package |
| | | Does not include most up-to-date firmware |
# Disclaimer
AMD®, the AMD Arrow logo, AMD Instinct™, Radeon™, ROCm® and combinations thereof are trademarks of Advanced Micro Devices, Inc.
Linux® is the registered trademark of Linus Torvalds in the U.S. and other countries.
PCIe® is a registered trademark of PCI-SIG Corporation. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.
Google® is a registered trademark of Google LLC.
Ubuntu and the Ubuntu logo are registered trademarks of Canonical Ltd.
Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.
# rocm-docs-redirects
Redirects ReadtheDocs Community documentation sites to Business
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.