Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Open Compute Project

AI overview

Sign in and the AI will write an overview from our coverage.

Headlines · 8

  1. Morgan Stanley: OCP summit spotlights optics, 800VDC, hardware security

    Morgan Stanley issued an industry report on the Open Compute Project Global Summit identifying three core technology shifts: upgraded optical interconnect architectures, a move to 800VDC high-voltage DC power with liquid cooling, and chip-level telemetry with hardware root-of-trust upgrades. The report cites Accton and Celestica showing 1.6T switches based on Broadcom's Tomahawk 6, and an Nvidia-Delta power blueprint in which a single 660kW 800VDC power rack supports two Vera Rubin NVL72 compute racks. It also names Taiwan suppliers including BizLink, Chroma, Delta, ASPEED and eMemory as poised to benefit from the transition.

    科技新报 · 🔥 8
  2. Samsung lays out AI memory roadmap: speed, scalability, persistence

    Samsung Electronics unveiled its next-generation memory strategy at the Open Compute Project (OCP) Global Summit in California, naming speed, scalability and persistence as its three pillars. The company said that as AI models keep growing, memory capacity expansion and data storage technology have become as critical as memory bandwidth. Its plans include HBM5, custom HBM, zHBM and processing-in-memory (PIM) for throughput, CMM-D and CXL-based products for capacity expansion, and next-generation SSDs, QLC and flexible data placement for storage. Samsung also showed ecosystem work: Groq chips manufactured on its 4nm process plus the Groq 3 LPX system with Nvidia, and a CXL-based memory expansion solution demonstrated with AMD.

    科技新报 · 🔥 0
  3. AMD's Lisa Su: 2026 is the real 'year of AI inference'

    In an Oct. 12 keynote at the 2026 Open Compute Project Global Summit, AMD CEO Lisa Su called this year the real "year of AI inference," saying inference should account for about 60% of the market and that agentic workloads are exploding. She said AMD has raised its CPU market forecast three times in the past nine months and expects roughly 50% compound annual growth over the next four to five years, with the CPU total addressable market exceeding $200 billion by 2030, while the data center AI accelerator market approaches $1.5 trillion. She also said AMD began shipping Helios production systems in September — the report's related links refer to Helion — with first deployments at OpenAI, Meta, Microsoft, Oracle and Anthropic data centers, and output expanding from Q4 into the first half of next year.

    IT之家 AI · 🔥 0
  4. AMD unveils Helion agentic AI rack design; Helios racks enter production

    At the OCP Global Summit 2026, AMD chair and CEO Lisa Su introduced Helion, an open rack-scale reference design for agentic AI workloads built to OCP's Open Rack v3 (ORv3) 44OU specification, with configurations planned for availability in November 2026 and specs to be submitted to the OCP community. AMD says Helion comes in CPU-only and CPU-GPU configurations supporting up to 32,768 Zen 6 CPU cores, 16 Instinct MI350P accelerators, 74 petaflops of FP4 compute and 102 Tbps of networking, with Venice CPU and power options split across personal-assistant, enterprise and technical-computing agent workloads, plus an integrated token router that directs agent model requests across frontier and open-source models. AMD also said its Helios AI rack, first shown at OCP last year, has entered production, with the first systems installed and running at OpenAI, Meta, Microsoft and Oracle; Helios combines Instinct GPUs, EPYC processors, Pensando networking and the ROCm software stack.

    科技新报 · 🔥 0
  5. Pegatron shows Vera Rubin NVL72 rack-scale AI platforms at OCP Global Summit 2026

    Pegatron said it will showcase next-generation AI infrastructure at OCP Global Summit 2026 under the theme "Scaling Innovation for the AI Era," stressing the path from system design and liquid cooling to volume manufacturing and rack-level validation. Its centerpiece is the RA4803-72N3 rack-scale platform, described as an NVIDIA Vera Rubin NVL72 system integrating 72 NVIDIA Rubin GPUs, 36 NVIDIA Vera CPUs and sixth-generation NVIDIA NVLink in a fully liquid-cooled rack, alongside a 2U liquid-cooled AS210-2T1-8H3 (NVIDIA HGX Rubin NVL8) and 48U rack configurations for up to 64 or 72 GPUs. Pegatron also showed an NVIDIA Vera CPU liquid-cooling system and a solution it calls NVIDIA Groq 3 LPX, plus AMD EPYC 9006-based AI-ready compute and storage servers, with workflows built on the NVIDIA DSX AI factory platform linking digital-twin simulation to factory deployment, according to the company.

    科技新报 · 🔥 0
  6. Intel unveils GPU D-series: up to 480GB LPDDR5X for trillion-parameter inference

    At the OCP 2026 Global Summit, Intel named its Crescent Island inference accelerator the Intel GPU D-series, with the first product listed as the GPU D270. Intel says the Xe3P-based card pairs 32 Xe cores and 256 XMX engines with 1.3 PFLOPS of FP8 throughput, 1.5 TB/s memory bandwidth, a 350W TDP and an air-cooled PCIe Gen5 form factor, and offers up to 480GB of LPDDR5X memory (a 160GB Intel reference card and a 480GB ODM ceiling, per reports). Intel claims one card can run models up to 300 billion parameters and an eight-card system can handle more than one trillion parameters; sampling has begun, but pricing and general availability have not been announced.

    Tom’s Hardware · 🔥 48
  7. NVIDIA opens Vera Rubin rack designs and AI factory standards at OCP

    At the OCP Global Summit in San Jose, NVIDIA contributed Vera Rubin rack and compute tray designs to the Open Compute Project through its open MGX architecture, and detailed an updated MGX 1.2 specification with rack-level blind-mate liquid cooling, a 5,000A liquid-cooled busbar and support for up to 250kW racks, as characterized by NVIDIA. It also released open DSX AI Factory software components and joined AMD, Google, Meta, Microsoft, Cisco and Visionbay on open data-center, grid-resilience, USFP optics and Falcon Transport networking efforts. Chinese appliance maker Midea appeared on NVIDIA's MGX thermal-management partner list, with reports citing unnamed sources that the cooperation centers on liquid cooling.

    NVIDIA Blog · 🔥 0
  8. AMD unveils Helion agentic AI rack blueprint with up to 32,768 Zen 6 CPU cores

    AMD announced Helion at the OCP summit, a rack-scale reference design built on the Open Compute Project's ORv4 spec and the open-platform approach of its Helios rack, pitched as an open blueprint for agentic AI rather than a system it will sell itself. A CPU-only configuration can hold up to 128 of AMD's 256-core sixth-generation EPYC "Venice" (Zen 6) processors, reaching 32,768 cores for hyperscalers, while a mixed configuration supports up to 16 Instinct MI350P GPUs and 102 Tbps of networking for enterprise deployments. AMD says it will contribute the specification to OCP, with Dell, HPE, Lenovo, Supermicro and Foxconn among partners developing systems planned for November 2026, though one report points to first-half 2027 availability and no pricing was disclosed.

    Tom’s Hardware · 🔥 0

Experience and discussion from the community

Share my Open Compute Project experienceAsk about Open Compute Project

Nobody has shared their experience with Open Compute Project yet.