Red Hat Summit Red Hat SummitSupportKonsoleEntwicklerDemo startenKontakt

Dieser Inhalt ist in der von Ihnen ausgewählten Sprache nicht verfügbar.

Chapter 4. Types of certification workflow

4.1. Single Instance Type certification
Link kopieren

A Single Instance Type Certification consists of bootable, installable, and operableable collection of physical or virtual hardware features, as defined by a specification provided by a Partner.

The specification may define features as standard, or optional. The Instance Type is considered to provide all of the features from the complete collection of standard and optional components unless explicitly excluded by or from the specification.

4.2. SuperSet Instance Type certification
Link kopieren

The SuperSet Instance Type Certification covers a variety of different configurations of the same Instance Type. The SuperSet Instance Type is known by a unique name within a naming convention.

Example

compute_slow, compute_medium, compute_fast

The certification is conducted as per the basic process where all of the configurations are reviewed. The test plan will consider multiple configurations in order to increase the testing and processing efficiency without creating risks. The certification publication in the catalog will be a combined entry where the sizes covered by the certification will be displayed on the Instance Type certification entry.

4.3. Supplemental certification
Link kopieren

A supplemental certification allows a certified Instance Type to extend or alter the configuration of the Instance Type.

4.4. Pass-Through Instance Type certification
Link kopieren

A Pass-Through Instance Type Certification refers to the ability of a third party system or component to be granted the same certification as Instance Type previously certified by the Original Provider.

The Original Provider can extend a certification granted to their instance type to another Partner’s Instance Type where the original provider:

  • Has permission from the third party
  • Has the mechanics to ensure the third party does not alter the hardware in such a way that it would no longer be considered a subset of the original model certified by Red Hat
  • Extends their responsibilities of support and representative Instance Type to include situations involving the third party Instance Type

The third party however cannot extend their pass-through certification to another Partner.

Important

Both Partners are required to be members of the CCSP program; only the original provider may request Pass-Through certifications.

Note

You may also utilize the pass-through process where the same Instance Type is available with multiple names.

4.5. Red Hat Enterprise Linux AI
Link kopieren

Red Hat Enterprise Linux AI (RHEL AI) is a foundation model platform designed to seamlessly develop, test, and run Granite family large language models (LLMs) for enterprise applications. RHEL AI uses image mode, providing the AI platform as a container image, streamlining management & deployment.

A RHEL cloud instance certification is a prerequisite for a RHEL AI cloud instance certification. The RHEL cloud instance certification testing verifies the general functionality of your system.

The RHEL AI system certification testing focuses on verifying Granite and InstructLab functionality, ensuring DataCenter GPUs on your system function correctly.

You must apply for the Red Hat Enterprise Linux AI cloud instance certification after your system has received the required RHEL certification, as indicated in the table below.

Expand
Desired RHEL AI CertificationRequired RHEL Version CertificationArchitecture
 

Minimum

Maximum

 

1.3

9.0

9.4

x86_64

1.4

9.0

9.4

x86_64

Note

Certifications must be completed before the end of full support for each RHEL AI version.

After you open a RHEL AI cloud instance certification, Red Hat will provide a certification test plan based on the model specifications you provide. The test plan follows the same process as RHEL cloud instance certification:

  • If your system remains unchanged or if modifications were already addressed through a RHEL supplemental certification, only the Data Center GPUs will remain as incomplete in the test plan.
  • If new hardware components are identified in your model specifications compared with the RHEL certification, Red Hat will open a RHEL supplemental certification for these components and this supplemental certification will block your RHEL AI cloud instance certification.

To meet the test plan requirements and complete the certification, you must provide test results or leverage information for each DataCenter GPU, following the requirements in table 5.1 Accelerators.

Certification policies

RHEL AI cloud instance certifications follow the standard RHEL certification policies, except where otherwise noted. This includes the general certification policies and requirements.

  • Testing must be conducted on production hardware, using an unmodified, bootable RHEL AI container without additional layers.

  • RHEL AI certifications will remain valid for the life of major release (X), starting from the minor release (X.Y) initially certified, until the next major release (X+1).

    • Certification validity will typically be displayed with a specific starting release and symbolic end release expressed as a range (e.g.” RHEL AI 1.2 - 1.x”).
  • RHEL AI cloud instance certification supports system pass-through certification, similar to RHEL system certification.

  • Component leveraging is supported only between different sizes within the same instance type (e.g., 16xlarge and 32xlarge), but not between different instance types or different partners. However, to qualify for leverage the original test results must be:

    • From the same major release of RHEL AI.
    • From a minor release equal to or prior to the targeted RHEL AI minor release.
    • Not older than one year.

Testing requirements

The Red Hat Certification Hardware AI test suite includes the necessary tests to obtain a RHEL AI cloud instance certification. To run the tests, you must first deploy RHEL AI on a system that meets the minimum RHEL AI criteria. Instructions for running the tests are found in the Red Hat Cloud Instance Type Workflow Guide. Red Hat strongly recommends obtaining a Red Hat Certified Engineer accreditation as well as familiarity with image mode, Granite, and instruct lab before participating in RHEL AI cloud instance certification.

Nach oben
Red Hat logoGithubredditYoutubeTwitter

Lernen

Testen, kaufen und verkaufen

Communitys

Über Red Hat Dokumentation

Wir helfen Red Hat Benutzern, mit unseren Produkten und Diensten innovativ zu sein und ihre Ziele zu erreichen – mit Inhalten, denen sie vertrauen können. Entdecken Sie unsere neuesten Updates.

Mehr Inklusion in Open Source

Red Hat hat sich verpflichtet, problematische Sprache in unserem Code, unserer Dokumentation und unseren Web-Eigenschaften zu ersetzen. Weitere Einzelheiten finden Sie in Red Hat Blog.

Über Red Hat

Wir liefern gehärtete Lösungen, die es Unternehmen leichter machen, plattform- und umgebungsübergreifend zu arbeiten, vom zentralen Rechenzentrum bis zum Netzwerkrand.

Theme

© 2025 Red Hat