Skip to content

Adding initial text #31

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Adding initial text #31

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2aec952
added text
jcleung5549 Sep 26, 2025
9268680
Update System_GPU_Specification_v0.1.md
jcleung5549 Sep 26, 2025
f9f2348
Update System_GPU_Specification_v0.1.md
jcleung5549 Sep 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
49 changes: 43 additions & 6 deletions docs/System_GPU_Specification_v0.1.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,32 +32,69 @@ License](https://creativecommons.org/licenses/by-sa/4.0/).

This document specifies the requirements for managing a GPU in a system.

# Requirements
## Scoping a GPU System

## Manageability Interface
Within the document, system shall be an entity which is:

The system shall expose a Redfish Service (v??).
* Bounded by a physical chassis
* Exposes an out-of-band remote manageability interface is Redfish over Ethernet

*
* Shall contain one or more GPUs - which may be on a subsystem (UBB) or discrete
* Use Redfish, PLDM and SPDM for local manageability
*

The system:

* Shall contain a power subsystem
* Shall contain compute complexes
* Shall contain a cooling and thermal subsystem
* May contain a management controller hierarchy

The GPU system is the foundation of a datacenter manageament hierarchy, which might include the rack manager, row manager, and datacenter manager.

** TBD - Are disaggregated GPU Systems included? - with GPU shelves, JBODs, etc**

# Redfish Requirements

## Redfish Interface

The Redfish interface should be Redfish Specification v1.22, [DSP0266](https://www.dmtf.org/dsp/DSP0266)

## Redfish data model

The Redfish Service shall conform to the following OCP Profiles
* [UBB_BaselineManagement profile v1.x](https://github.com/opencomputeproject/HWMgmt-OCP-Profiles/blob/master/gpu/OCP_UBB_BaselineManagement.v1.0.0.json)
* [OCP Server Profile v1.1](https://github.com/opencomputeproject/HWMgmt-OCP-Profiles/blob/master/Server/OCPServerHardwareManagement.v1_1_0.json)

### Firmware Management
### Redfish Component Integrity
## Firmware Management

The Redfish Service should support the MultipartHttpPushUri. If MultipartHttpPushUri property is not supported, then the HttpPushUri property shall be supported.

*Note*: The HttpPushUri property is 'deprecated' in favor of the MultiPartHttpPushUri property.

## Redfish Component Integrity

## Redfish Aggregation Behavior

The system should follow the guidance of the *(insert document name*) (GPU specific version)
The system is an assemblage of subsystems. In order to improve interoperablity, the Redfish data model of the the assemblage should be consistent . Hence, the guidance of the "Redfish Aggregation Guidance for Compute Expansion Modules", [DSP2090](https://www.dmtf.org/sites/default/files/standards/documents/DSP2090_1.0.0.pdf) should be followed.

## Redfish Message Registry

# PLDM Requirements

## Firmware Management

The GPU shall support the requirements specified in the "GPU Firmware Update Requirements v1.1"

The CPU shall support the requirements specified in ...

## Sensors and Effecters

# Security Requirements (SPDM)

# Physical Interfaces and connections

## USB Requirements

# Performance requirements (e.g. KPIs)
Expand Down