AI Hardware Software Co-Design

AI HW SW CoDesign is a Sub-Project of the OCP Server Project. Because AI innovation is often constrained by compute platforms with long design cycles, the workgroup aims to enable a scalable yet composable infrastructure that adapts dynamically to an application's runtime characteristics. Its 2023 whitepaper, Polymorphic Architecture for Future AI Applications, proposed hardware-software co-design with large-scope scalability and hierarchical transformability and composability; a newer whitepaper details the Polymorphic Architecture around continuous transformability and recursive composability, letting resources from chiplets to data centers reconfigure for diverse workloads. Current study areas include seamless scaling across hierarchies, HW/SW co-design for utilization and energy efficiency, flexible configuration, heterogeneous resources and cost savings.

AI HW SW CoDesign

Get involved

Steering Committee Representative

Gregory D Sellman

Project Leads

Subscribe to mailing list

About This Workstream

AI HW SW CoDesign is a Sub-Project under the OCP Server Project.

Rapidly evolving artificial intelligence (AI) continues to drive ever-increasing computation demands. However, innovation in AI models and the capability to train and infer from these models are often constrained by compute platform capabilities. Since system hardware design cycles are long and design goals may be simplified to reduce development cycles, optimization of these systems for performance and operational efficiency targets is hard. Thus, the AI co-design workgroup aims to innovate and enable a scalable yet composable computing infrastructure which can dynamically adapt to the application’s runtime characteristics and execute the application with high efficiency.

In 2023, a vision addressing the current system shortcomings was outlined in a whitepaper titled “Polymorphic Architecture for Future AI Applications.” It proposed using hardware-software co-design principles with some key requirements, including large-scope scalability and a polymorphic computing infrastructure with hierarchical transformability and composability.

Polymorphic Architecture

Development of these ideas continued through deeper discussions of architecture, system design considerations, and the evolution of the state of the art in AI models and performance expectations. The result of OCP’s Community effort culminates in a new whitepaper which describes, in greater detail, a Polymorphic Architecture for AI computing that unleashes unprecedented flexibility, efficiency, and scalability.

Polymorphic Architecture represents a paradigm shift in how we conceptualize and implement AI computing systems. At its core are two key principles:

  • Continuous transformability
  • Recursive composability

These principles enable computing resources — from individual chiplets to entire data centers — to dynamically reconfigure themselves, optimizing performance for diverse AI workloads in real time.

Roadmap and Participation

As outlined in the white papers above, the foundational concepts, technical specifications, and potential applications of Polymorphic Architecture will continue to be a focus to accelerate AI computation. The group also has a high-level roadmap for implementation, including simulation frameworks and APIs for standardization. As the AI landscape continues to evolve, Polymorphic Architecture offers a forward-thinking solution to the challenges of next-generation AI computing. Industry leaders, researchers, and developers are invited to explore this transformative approach and contribute to shaping the future of AI infrastructure.

Scope

Using the Polymorphic Architecture as an exploration platform, the AI co-design workgroup continues its study in the following technology areas:

Seamless scaling — across computing hierarchies, from chips and systems to large-scale clusters.
HW/SW co-design — to improve resource (e.g. compute and interconnect) utilization and energy efficiency.
Flexible configuration — configuration and adaptation of resources for future-proofing against rapidly evolving AI algorithms.
Heterogeneous resources — means to realistically utilize heterogeneous resources in systems.
Cost savings — potential for significant cost savings in AI infrastructure.

Resources

As artificial intelligence (AI) continues to evolve at a rapid pace, computing infrastructure supporting it must adapt to meet dramatically ever-changing demands. Innovation in AI models, and the capability to train and infer from these models are constrained by compute platform capabilities. As development will eventually give way to deployment, optimization of these systems for performance and operational efficiency will be imperative. System hardware design cycles are long and design goals are simplified in this phase of reducing development cycles to meet the market with actual platforms. Today’s systems do not offer high levels of performance and flexibility while meeting performance and operational efficiency targets. In 2023, a vision addressing this shortcoming was outlined in a whitepaper titled “Polymorphic Architecture for Future AI Applications.” It proposed addressing these challenges using hardware software co-design principles. Some key requirements included large-scope scalability, and polymorphic computing infrastructure with hierarchical transformability and composability. In 2024 this working group formalized the Polymorphic Architecture into a definition for design. this formed the basis for the work in 2025 to specify key elements of the AI ecosystem such as AI Fabrics, Topologies as well as algorithms for new compute models and data services. These will be verified through simulation and modeling, shared with the greater community through OCP.

In February 2022, AI Co-Design workgroup was officially kicked off with a group charter to focus on exploration of evolutional AI Acceleration architecture. The workstream graduated in late 2024 and has been formally promoted as a sub-project of OCP Server Project.

Documents

Recordings from Past Calls

2026
2025

Schedule

For more details, refer to the workstream calendar:

AI HW SW CoDesign Calendar

This project meets every other Friday.

AI Hardware Software Co-Design calendar

Related Details

Contributions

All project contributions