With an emphasis on AI inference, Intel unveiled a new data center GPU and a rack-scale reference design for Gaudi 3.
HPE Sharpens iLO 7 and Gen12 Servers at OCP 2025
At OCP 2025, HPE upgrades iLO 7 and Gen12 servers with stronger security, modular designs, NVIDIA support, and MLPerf-leading AI results.
Broadcom Thor Ultra: UEC-Compliant 800G AI Ethernet NIC for 100K+ XPU Scale-Out
Broadcom’s Thor Ultra 800G AI Ethernet NIC is designed for clusters with more than 100,000 XPUs.
AI ◇
Enterprise
NVIDIA Offers a Preview of What’s Next for Gigawatt-Scale AI Factories at the OCP Global Summit
NVIDIA previews Vera Rubin NVL144, an MGX-generation open-architecture rack server.
Intel Unwraps Core Ultra Series 3 “Panther Lake” and Xeon 6+ “Clearwater Forest” on Intel 18A
Intel Panther Lake and Clearwater Forest processors are built on an 18A process node.
IBM Spyre Accelerator: Low-Latency Inference for Generative and Agentic AI on IBM Z, LinuxONE, and Power
IBM’s low-latency inference engine, Spyre Accelerator, will be generally available beginning October 28 for z17 and LinuxONE 5. Power 11 system users will get access shortly thereafter.
AIC SB407-VA: High-Density 4U Storage Server Designed for Heavy-Duty Data Demands
AIC SB407-VA 4U server packs up to 60+8 hot-swap bays, PCIe Gen5, DDR5, dual Intel Xeon, and IPMI/Redfish for AI, data lakes, and analytics.
Intel Expands U.S. Manufacturing With Fab 52, Reveals Panther Lake Platform
Intel debuts its next-gen Panther Lake chip and ramps up U.S. production with the launch of Fab 52 in Arizona.
AMD Launches Solarflare X4 Ethernet Adapters for Ultra-Low Latency Applications
AMD’s new Solarflare X4 Ethernet Adapters aim to cut network latency and boost performance for high-speed trading and analytics systems.
Broadcom Ships Tomahawk 6-Davisson CPO Ethernet Switch, Doubling Bandwidth to 102.4 Tb/s for AI Fabrics
Broadcom begins shipping the Tomahawk 6-Davisson CPO Ethernet switch, doubling the bandwidth to 102.4 Tb/s.
Amazon