NVIDIA has run a 104-qubit simulation of a fault-tolerant quantum processor on JUPITER, Europe’s first exascale supercomputer at the Jülich Supercomputing Centre, using 8,192 GH200 GPUs. Announced ahead of GTC Berlin, the run models magic state cultivation, a method of producing the non-Clifford states a fault-tolerant computer consumes during logical operations, under realistic hardware noise.
HPE announced HPE Spaceborne Computer-4, a compute module that will run AI inference and what HPE says will be the first HPC benchmarks on the Moon. Loaded with KIOXIA SSDs, the module will fly as a payload on Astrolab’s FLIP rover, which Voyager Technologies’ Griffin-1 lander will deliver near the lunar south pole. Launching in
Neo4j announced the Neo4j Context Engine today, a product built to give AI agents a shared, reusable knowledge layer for reasoning over enterprise data. The problem it targets is one most agent deployments hit early: an agent does not know what “customer,” “account,” “supplier,” or “risk” means inside a particular company, because that meaning is
Meta published a detailed look inside Prometheus today, the gigawatt-scale AI supercluster it is building in New Albany, Ohio. The first phase is complete and has trained the Muse family of foundation models, including the Muse Wonder models, and Meta says the site is being scaled toward roughly 1 million GPUs. Two years ago, the
Leostream announced Leostream Cloud, a cloud-based remote access service, available now, that connects authorized users directly to physical and cloud-hosted workstations and other protected resources. The service combines identity-based authentication with policy-driven resource assignment, so users see and reach only the systems they are cleared to use. Leostream says that model can replace or avoid
XCENA showed two AI inference demonstrations built on its MX1 CXL computational memory at the OCP Global Summit 2026 in San Jose: a KV cache shared across inference servers through a CXL memory pool, and vector search for retrieval-augmented generation (RAG) run inside the memory device. SK hynix also unveiled a CXL Hybrid Memory product
SP Industrial introduced six enterprise SSDs, spanning E1.S, U.2, M.2 2280, and 2.5-inch SATA form factors, aimed at AI and data center storage as well as enterprise boot drives. Every model in the lineup pairs 3D eTLC flash with DRAM and power failure protection (PFP), and the drives that offer a higher endurance rating reach
Veeam made Intelligent ResOps (iResOps) for Microsoft 365 generally available today, the first iResOps product to reach general availability since Veeam introduced Intelligent ResOps in May. It is an add-on for Veeam Data Cloud for Microsoft 365 that connects to the existing service without data migration or re-architecture, and it maps the tenant’s AI agents,
Samsung is showing a CXL memory pool validated with AMD at the OCP Global Summit in San Jose this week, built on its CMM-D CXL memory modules and aimed at agentic AI inference. The companies worked together to validate CXL memory as a way to scale memory capacity and bandwidth alongside CPU core counts, on
Celestica introduced the DS6005 at the OCP Global Summit, a co-packaged optics (CPO) data center switch that puts 102.4Tbps of switching capacity and 64 optically enabled 1.6TbE ports in a 1.5OU, fully liquid-cooled chassis for Open Rack v3. It is built on Broadcom’s Tomahawk 6 Davisson, the CPO version of the 102.4T Tomahawk 6, which
Amazon