Open Compute Project

The Open Compute Project continues to drive hardware commoditization and design efficiency, extending its influence from hyperscale to the broader enterprise. Its evolution supports advanced, distributed architectures, shaping modern data center infrastructure. Recent efforts are intensely focused on artificial intelligence workloads and the associated power constraints, advancing specifications for distributed data centers and high-voltage DC power distribution, including collaborations on solid-state transformer technology.

OCP's trajectory highlights a dual focus on enabling cutting-edge distributed architectures for AI while tackling practical operational realities. This includes ensuring the safe and efficient implementation of high-voltage power systems, such as 800-volt distribution, and addressing connectivity needs across a computing continuum. The project's work is crucial for supporting the complex processing and infrastructure requirements of emerging technologies and agentic AI.

New developments emphasize agentic AI driving the data center's chip-to-grid transformation, with open architectures and gigawatt-scale construction. Companies are unveiling new rack designs and GPU details specifically for AI workloads, focusing on memory, power, and deployment readiness. This includes CPU-only and mixed configurations, supporting rapid growth and increased rack densities in AI-centric environments, underscoring a significant shift towards AI-specific infrastructure solutions.

Last updated October 18, 2026

Coverage

AMD has introduced Helion, an open rack design for agentic artificial intelligence workloads, offering CPU and GPU options with planned availability for IT teams.
Intel introduced new Xeon rack designs and D-series GPUs specifically engineered for agentic artificial intelligence workloads.
AMD unveiled Helion racks for AI agents at the OCP summit, with CPU-only and mixed configurations available from partners in November.
The Open Compute Project is advancing fleet-scale hardware management for AI data centers amidst rapid growth, higher densities, and power constraints in the APAC region.
ABB offers solutions for the Open Compute Project's disaggregated power rack, addressing grounding and protection considerations to enable safe and scalable 800-volt artificial intelligence power distribution.
The Open Compute Project is collaborating with the Innovative Optical and Wireless Network stack to develop specifications for distributed data centers that can support the demands of artificial intelligence workloads requiring high-speed, low-latency connectivity.