Releases: intel/intel-technology-enabling-for-openshift
Intel® Technology Enabling for OpenShift* version 1.3.0
We are thrilled to announce the release of Intel Technology Enabling for OpenShift version 1.3.0 for OpenShift 4.15.15+.
What’s new in version 1.3.0
- Intel Device Plugin Operator and the related device plugins are upgraded to v0.29.0.
- Intel Data Center GPU Driver for OpenShift is upgraded to v2.2.0.
- Developer Preview of Media feature provisioning for Intel® Data Center GPU.
- Prototype of one stop online documentation publishing.
- Prototype of Red Hat Operator and Ansible based One-Click solution.
About this release
General Available features:
- Intel® Data Center GPU Flex Series provisioning
- Intel® Software Guard Extensions (Intel® SGX) provisioning
- Intel® QuickAssist Technology (Intel® QAT) provisioning
- Intel AI inference E2E solution based on
- OpenVINO™ Operator that simplifies managing optimized deep learning inference at scale in OpenShift
- Red Hat OpenShift AI
- Intel Data Center GPU Flex Series provisioning
Developer Preview feature:
- Intel Data Center GPU Max Series provisioning
- Media feature provisioning for Intel® Data Center GPU
Prototype feature:
- One Stop Online Document
- Red Hat Operator and Ansible based One Click solution.
Intel technology features availability continuously verified on OpenShift release using DCI based CI/CD pipeline.
Fixed issues
NFD SGX EPC extended resource support issue
Intel® Technology Enabling for OpenShift* version 1.2.1
We are thrilled to announce the release of Intel Technology Enabling for OpenShift version 1.2.1 for OpenShift 4.14.11+.
What’s new in version 1.2.1
- Fast GPU provisioning supported
With the prebuilt Intel Data Center GPU driver container images managed by the KMM Operator and Intel Data Center GPU bootless technology, provisioning of Intel GPUs is reduced from more than three minutes to less than one second for each node on OpenShift. This provisioning improvement enhances the Intel Data Center GPU scale-out performance and simplifies the scale-out and on-demand allocation of the GPU resources (see Fast-GPU-Provisioning-Technology). - Developer Preview Intel® Data Center GPU Max Series
- Bootless GPU provisioning, see issue 34
- Reduced driver container image size, see issue 33
- Previously released Intel features are now continuously verified with each new OpenShift release allowing for features to be always available.
- Intel Data Center GPU Driver for OpenShift is upgraded to v2.2.0.
- Includes the use of Red Hat OpenShift AI* v2.8.0
About this release
1.2.1 release includes the below General Available features:
- Intel® Data Center GPU Flex Series provisioning
- Intel® Software Guard Extensions (Intel® SGX) provisioning
- Intel® QuickAssist Technology (Intel® QAT) provisioning
- Intel AI inference E2E solution based on
- OpenVINO™ Operator that simplifies managing optimized deep learning inference at scale in OpenShift
- Red Hat OpenShift AI
- Intel Data Center GPU Flex Series provisioning
Developer Preview feature:
- Intel Data Center GPU Max Series provisioning
Fixed issues
Intel® Technology Enabling for OpenShift* version 1.2.0
We are pleased to announce the release of Intel Technology Enabling for OpenShift version 1.2.0.
What’s New in Version 1.2.0
Intel Technology Enabling for OpenShift version 1.2.0 is the GA release with the critical SeLinux regression issues fixed on OpenShift-4.14.11+. The changes include:
- Intel Data Center GPU Driver for OpenShift is upgraded to v2.1.0 released using Intel Data Center GPU driver CI/CD pipeline.
- Using Red Hat OpenShift AI v2.6.0 release.
- Using KMM Operator v2.0.2 release
- Below critical SeLinux issues got fixed
Fixed issues
Intel® Technology Enabling for OpenShift* version 1.1.0 Release Notes
We are pleased to announce the release of Intel Technology Enabling for OpenShift version 1.1.0.
What’s New in Version 1.1.0
Intel Technology Enabling for OpenShift version 1.1.0 is the technical preview release with below new features for RHOCP-4.13.10.
- Intel Device Plugin Operator and the related device plugins are upgraded to v0.28.0.
- Intel Data Center GPU Driver for OpenShift is upgraded to v2.0.0 to match RHEL/RHCOS-9.2.
- Using RHODS v2.4.0 release
- Add Intel Flex 140 card into RHODS Dashboard using OpenShift AI Accelerator Profile.
- 5th Gen Intel® Xeon® Scalable processors support added
- Using Red Hat built and distributed RPM QAT lib package.
- Using ubi9 Image in the test cases.
Fixed issues
support for accelerators via OdhDashboardConfig
Known Issues and Planned Enhancements
Below OpenShift-4.13 regression issues have been fixed and will be merged into OpenShift-4.14. To make everything work properly on the 1.1.0 release for OpenShift-4.13.10, users must apply the instructions in the workaround section.
Intel SGX Device Plugin returns error "permission denied" for OpenShift 4.13
GPU workload can not access the GPU devices from the Container environment
Intel® Technology Enabling for OpenShift* version 1.0.1 Release Notes
We are pleased to announce the release of Intel Technology Enabling for OpenShift version 1.0.1.
The Intel Technology Enabling for OpenShift project provides Intel Data Center hardware features provisioning technologies with Red Hat OpenShift Container Platform (RHOCP). The technology to deploy and manage the End-to-End (E2E) solutions as well as the related reference workloads for these features are also included in the project.
The project source code is maintained in the Intel Technology Enabling for OpenShift Repo
The 1.0.1 document is published at Intel Technology Enabling for OpenShift Repo Readme
The related certified images for RHOCP are published at Red Hat Ecosystem Catalog:
- Intel Device Plugins Operator v0.26.2
- Intel Device Plugins Operator Container Images - SGX Plugin
- Intel Device Plugins Operator Container Images - SGX Initcontainer
- Intel Device Plugins Operator Container Images - GPU Plugin
- Intel Device Plugins Operator Container Images - GPU Initcontainer
- Intel® Device Plugins Operator Container Images - QAT Plugin
- Intel® Device Plugins Operator Container Images – QAT Initcontainer
- Intel Data Center GPU Drivers for OpenShift
What’s in Version 1.0.1
Intel Technology Enabling for OpenShift version 1.0.1 release provides the following solutions with RHOCP-4.12.26. The new feature in the release is highlighted in bold.
- Intel® Data Center GPU Flex Series provisioning
- Intel® Software Guard Extensions (Intel® SGX) provisioning
- Intel® QuickAssist Technology (Intel® QAT) provisioning
- Intel AI inference E2E solution based on
- Intel® OpenVINO™ technology
- Red Hat OpenShift Data Science (RHODS)
- Intel Data Center GPU Flex Series provisioning
Users can use the Intel Technology Enabling for OpenShift version 1.0.1 readme to setup these solutions with the following images:
- Intel Device Plugin Operator certified with RHOCP-4.12
- Intel SGX device plugin images certified with RHOCP-4.12
- Intel GPU device plugin images certified with RHOCP-4.12
- Intel QAT device plugin images certified with RHOCP-4.12
- Intel Data Center GPU Driver Container Image certified with OCP-4.12
New Features Description
- Intel QuickAssist Technology (Intel QAT) provisioning
Intel QuickAssist Technology can be used to accelerate data encryption and compression for applications from networking to enterprise, cloud to storage, and content delivery to database.
In this release, the QAT provisioning is verified and supported for QAT 4xxx devices on 4th Gen Intel® Xeon® Scalable processors and verified on 5th Gen Intel® Xeon® Scalable processors. The qatlib based workload has been used to verify QAT provisioning. The qatlib based workload can claim and consume the QAT resource provided from QAT provisioning on RHOCP.
Release contacts
Intel Business Contact:
[email protected]
Engineering Manager:
[email protected]
Release Support Contact:
[email protected]
[email protected]
[email protected]
Known Issues and Planned Enhancements
Intel® Technology Enabling for OpenShift* version 1.0.0 Release Notes
General
We are pleased to announce the release of the project Intel Technology Enabling for OpenShift version 1.0.0.
The Intel Technology Enabling for OpenShift project provides Intel Data Center hardware features provisioning technologies with Red Hat OpenShift Container Platform (RHOCP). The technology to deploy and manage the End-to-End (E2E) solutions as well as the related reference workloads for these features are also included in the project.
The project source code is maintained in the Intel Technology Enabling for OpenShift Repo.
The 1.0.0 document is published at Intel Technology Enabling for OpenShift Repo Readme.
The related certified images for RHOCP are published in Red Hat Ecosystem Catalog.
What’s New in Version 1.0.0
Intel Technology Enabling for OpenShift version 1.0.0 release provides the following solutions with RHOCP-4.12.
- Intel® Data Center GPU Flex Series provisioning
- Intel® Software Guard Extensions (Intel® SGX) provisioning
- Intel AI inference E2E solution based on
- Intel® OpenVINO™ technology
- Red Hat OpenShift Data Science (RHODS)
- Intel Data Center GPU Flex Series provisioning
You can follow the readme and use below images to setup these solutions:
- Intel Device Plugins Operator for OpenShift 0.26.1
- Intel Device Plugins Operator Container Images - SGX Plugin
- Intel Device Plugins Operator Container Images - SGX Initcontainer
- Intel Device Plugins Operator Container Images - GPU Plugin
- Intel Device Plugins Operator Container Images - GPU Initcontainer
- Intel Data Center GPU Drivers for OpenShift
Features Description
1. Intel Data Center GPU Flex Series provisioning
The Intel Data Center GPU Flex Series is flexible, robust, and the industry's most open GPU solution for the intelligent visual cloud. The GPU supports a diverse range of workloads in the industry starting with media streaming and cloud gaming, followed by support for AI visual inference and virtual desktop Infrastructure workloads. It supports an open, standards-based software stack optimized for density and quality with critical server capabilities for high reliability, availability, and scalability. This helps reduce the need for data centers to use disparate solutions and manage heterogenous or proprietary environments.
To provision Intel Data Center GPU cards on RHOCP, use the Kernel Module Management (KMM) Operator to manage the Intel Data Center GPU driver container image. Intel Device Plugins Operator is used to manage the GPU device plugin. The device plugin advertises the GPU card resources to the cluster and allows workloads to use the resource.
2. Intel Software Guard Extensions (Intel SGX) provisioning
Intel SGX offers hardware-based memory encryption that isolates specific application code and data in memory. Intel SGX allows user-level code to allocate private regions of memory, called enclaves, which are designed to be protected from processes running at higher privilege levels.
Intel Device Plugins Operator is used to manage the Intel SGX device plugin. The device plugin advertises the Intel SGX resources to the cluster and allows workloads to use the Intel SGX resources.
3. Intel AI Inference end-to-end solution
Intel AI inference end-to-end solution for RHOCP is built upon Intel Date Center GPU Flex Series provisioning and Intel® Xeon® processors. The two following AI inference modes can be used to verify the provisioning solution:
- Interactive Mode
Open Data Hub (ODH) and Red Hat OpenShift Data Science (RHODS) provide OpenVINO toolkit- based Jupiter Notebooks to help users interactively debug the inferencing applications or optimize the models with OCP using Intel Data Center GPU cards and Intel Xeon processors. - Deployment Mode
OpenVINO Toolkit Operator for OpenShift and Kubernetes provides the OpenVINO Model Server (OVMS) for users to deploy their inferencing workload using Intel Data Center GPU cards and Intel Xeon processors on a RHOCP cluster in a cloud or edge environment.
Release contacts
Intel Business Contact:
[email protected]
Engineering Manager:
[email protected]
Release Support Contact:
[email protected]
[email protected]
[email protected]
Known Issues
- Claiming i915 resources in a single container has 2 cards accessible for Intel GPU Flex 140 intel-device-plugins-for-kubernetes#1377
- Improve Driver Container Image Size intel-data-center-gpu-driver-for-openshift#33
- Support for accelerators (include Intel Data Center GPU) via OdhDashboardConfig in ODH and OCP