OCP Global Summit
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 8
- Goldman lifts WNC target price as it pushes into AI data-center networking
Citing a new Goldman Sachs research note, TechNews reports that Wistron NeWeb (WNC) is extending beyond its low-Earth-orbit satellite, Wi-Fi 7 and automotive connectivity businesses into optical circuit switches (OCS), 800G optical transceivers and 1.6TbE high-speed switches, products the company showed at the 2026 OCP Global Summit. According to the note, the new 1.6TbE switch uses 64 1.6TbE NPO ports for up to 102.4Tbps of switching capacity and emphasizes lower power, while the OCS is pitched on cross-generation Ethernet compatibility. The report also cites September revenue of NT$16.079 billion, up 64% year on year, and third-quarter revenue of NT$48.696 billion, up 86%, and raises the 12-month target price to NT$380 from NT$371 while keeping a Buy rating.
科技新报 · 🔥 0 - Intel calls silicon photonics next-gen 'Golden Hammer', claims sole 300mm InP volume fab
At the 2026 OCP Global Summit, Intel data center systems vice president Nic Dubé described silicon photonics as the "Golden Hammer" of next-generation system architecture, saying it will accelerate high-performance segments including HPC and AI. Intel claims it is the only high-volume fab able to supply 300mm (12-inch) indium phosphide silicon photonics at scale, citing a unique track record in integrated laser reliability, and says it is pushing community standards for external light sources. Intel estimates photonics systems can reach transmission energy below roughly 5 picojoules per bit (IT之家 reported under 5.5 pJ/bit), and says it will offer advanced interconnect topologies with runtime orchestration of CPUs, GPUs and memory pools as AI models grow. The exclusivity and efficiency figures are Intel's own claims and were not independently verified in the reports.
Wccftech · 🔥 0 - Lenovo starts global delivery of Nvidia Vera Rubin NVL72 AI racks
Lenovo announced at the 2026 OCP Global Summit that it has begun delivering Nvidia's next-generation Vera Rubin NVL72 rack-scale AI systems worldwide, according to 36Kr Express. Lenovo also said it has added the full Vera Rubin platform to its AI Cloud Gigafactory solution. The company framed the move as strengthening AI infrastructure buildout and delivery for cloud service providers globally; the report gave no delivery volumes, named customers or timeline.
36氪快讯 · 🔥 0 - AMD shows EPYC Verano: first EPYC with LPDDR memory, due 2027
AMD publicly showed its next-generation EPYC Verano data-center CPU for the first time at the OCP Global Summit that opened October 12, 2026, with CEO Lisa Su calling it "the next step in AMD's CPU roadmap for the AI era"; the company plans to make it available in 2027. Verano is built on Zen 6 and is the first EPYC product with an LPDDR memory interface, which reports describe as LPDDR5X paired with replaceable SOCAMM2 modules to improve AI data-center power efficiency and rack space usage. AMD positions it as a host-node CPU for AI accelerators intended to pair with future Instinct GPUs, and reports cite up to 112 Gbps xGMI connectivity for linking Instinct MI500. The core count remains disputed: Wccftech and cnBeta, citing ComputerBase, report a configuration of up to 192 cores, while IT之家 notes AMD has not confirmed that figure and Tom's Hardware previously reported up to 72 cores; AMD has not disclosed final core counts, memory specs, power, pricing or benchmarks. Su also said AMD expects the server CPU market to exceed $200 billion by 2030.
Wccftech · 🔥 0 - NVIDIA and Cisco unveil USFP optical module spec for 400G per lane
NVIDIA and Cisco announced at the OCP Global Summit plans for Ultra-Dense Small Form Factor Pluggable (USFP), a new optical module form factor they describe as the first pluggable module designed for 400Gb/s-per-lane electrical signals, with capacity up to 12.8Tb/s. The companies say initial USFP designs fit 16 full-duplex lanes in an OSFP-like size, expandable to 32 lanes, and claim 4x the bandwidth density and about 1.3x the power efficiency of current 200Gb/s-per-lane pluggable modules; these figures are vendor-stated and have not been independently verified. NVIDIA and Cisco are working with TE Connectivity and Amphenol on cages, connectors and modules, and plan to create an open industry specification through an upcoming community-led multi-source project.
NVIDIA Newsroom · 🔥 0 - Frore unveils diamond-infused coldplates for 2,300W Nvidia Rubin, claims up to 22C drop
Frore Systems introduced LiquidJet Diamond, a bespoke liquid-cooling coldplate series that adds diamond-wafer heat spreaders at the bottom of the coldplate to target hotspots on next-generation AI accelerators such as Nvidia's Rubin and AMD's Instinct MI400 and MI500. The company claims the design handles hotspot power densities up to 770 W/cm², lowers GPU temperatures by 16–22C versus conventional coldplates and improves AI token throughput per watt by 30%–35%, though those figures come from its own models and it has not disclosed how the diamond wafers are manufactured or integrated. Frore plans to demonstrate the technology at the OCP Global Summit on October 12–16, showing cooling for Nvidia's 2,300W Rubin, but has not said whether the demo uses production GPUs or a thermal test vehicle.
Tom’s Hardware · 🔥 0 - AI Data Pipeline Forum on Oct. 13 to spotlight infrastructure bottlenecks
SiliconANGLE reports that the AI Data Pipeline Forum will be held Oct. 13 in San Jose, California, alongside the OCP Global Summit, with discussions shifting from individual chip performance to system-level bottlenecks in data supply, high-speed networking, power delivery and cooling. TheCUBE will livestream the event, hosted by John Furrier, with guests from ScaleFlux, Cerebras, Cisco, Dell and others. The piece notes ScaleFlux introduced PCIe Gen6 SSD and CXL memory controllers, FC6116 and MC600, in July, while a Dell engineer stressed modular rack-scale infrastructure that adapts to varied power and cooling setups.
SiliconANGLE AI · 🔥 0 - AcBel and OmniOn to show 100kW power shelf and liquid cooling at OCP
AcBel Polytech will showcase new high-power supply designs with subsidiary OmniOn Power at OCP Global Summit 2026 in San Jose from October 13–15. Its 100kW Power Shelf fits 100kW of output into 1OU and, using a patented power distribution unit, can be configured for either ±400V or 800V DC, while a 19-inch 72kW Power Shelf for 50V architectures uses 12kW modules at up to 102 W/in³ and has earned 80 PLUS Ruby certification. AcBel said it is also developing next-generation liquid-cooled power supplies to handle rising power density in AI servers.
科技新报 · 🔥 0
Experience and discussion from the community
Share my OCP Global Summit experienceAsk about OCP Global Summit
Nobody has shared their experience with OCP Global Summit yet.