TLDR: Leaders from AMD, UALink, and UEC are slated to discuss advancements in AI infrastructure at the upcoming OCP Summit. This comes as the Open Compute Project Foundation (OCP) and the Ultra Accelerator Link (UALink) Consortium have partnered to accelerate the development of high-performance interconnects for AI and High-Performance Computing (HPC) workloads, aiming to establish an open alternative to existing proprietary technologies like Nvidia’s NVLink.
Industry leaders from Advanced Micro Devices (AMD), the Ultra Accelerator Link (UALink) Consortium, and UEC are set to convene and discuss the latest advancements in AI infrastructure, with their insights anticipated to be a key highlight leading up to the Open Compute Project (OCP) Summit. This discussion underscores a significant push towards open standards in the rapidly evolving field of artificial intelligence.
The Open Compute Project Foundation (OCP) and the UALink Consortium have forged a strategic partnership aimed at accelerating the development and integration of high-performance interconnects crucial for AI and High-Performance Computing (HPC) workloads. This collaboration is designed to foster an open alternative to proprietary technologies, such as Nvidia’s NVLink, which currently dominate the market. The UALink Consortium, established in mid-2024 and including major players like AMD, Intel, and Meta, has already ratified its first specification, UALink 1.0, earlier this year.
George Tchaparian, CEO of the OCP Foundation, emphasized the importance of this collaboration, stating, “By collaborating, the UALink Consortium and the OCP Community can shape system specifications to address critical challenges in interconnect bandwidth and scalability posed by advanced AI models.” This partnership is expected to align UALink’s interconnect technology with OCP’s ‘Open Systems for AI’ initiative and its ‘Future Technologies Initiative’s short-reach optical interconnect workstream’.
The goal is to accelerate the adoption of open interconnects in hyperscale data centers, particularly in systems built for training and deploying cutting-edge AI models. Peter Onufryk, an Intel Fellow and president of the UALink Consortium, highlighted the necessity for such advancements: “AI and HPC workloads require ultra-low latency and massive bandwidth to handle the scale and complexity of accelerated compute data processing to meet LLM requirements. Partnering with the OCP Community will accelerate the adoption of UALink’s innovations into complete systems, delivering transformative performance for AI markets.”
The upcoming 2025 OCP APAC Summit, scheduled for August 5-6 in Taipei, Taiwan, will serve as a significant platform for these discussions. The event will feature keynote presentations from industry experts, including AMD, and will host over 100 AI-focused breakout sessions. The OCP community has made substantial progress towards Open Systems for AI this year, with plans for over 100 AI-focused breakout sessions across its 2025 in-person Summits and Tech Day events, along with virtual technical workshops that have already seen significant attendance.
Also Read:
- AMD Elevates MI350 AI GPU Pricing by 70% Amidst Strong Demand
- Fireworks AI Nears $4 Billion Valuation Amidst Surging Demand for Open-Source AI Infrastructure
This initiative is particularly relevant as the Asia-Pacific region is expected to account for 36% of OCP-recognized equipment shipments in 2025, underscoring its critical role in driving AI innovation from semiconductor development to comprehensive server, cooling, and power systems infrastructure. The collaboration between OCP and UALink aims to integrate UALink’s scale-up AI interconnect technology into OCP Community-delivered AI clusters, providing the high-bandwidth, low-latency, and low-power connectivity essential for high-performance AI training and inference.
