AMD Helion: CPU Rack Cuts AI Agent Token Costs 43%
AMD unveiled Helion at the OCP Global Summit — an open rack blueprint with an integrated model router that routes AI agent calls ...
Intel GPU D270: 480GB Memory for AI Inference
Intel named Crescent Island the GPU D270 at OCP Summit. It packs up to 480GB ...
Ledger Spy Chip: $93M Stolen Before Wallets Were Opened
Ledger confirms a hardware spy chip inside CryptoBilis-sold wallets drained $93M. The Genuine Check passed ...
Surface RTX Spark Dev Box: The Case Against Cloud AI
Microsoft's $5,999 Surface RTX Spark Dev Box runs 120B parameter AI models locally. Here is ...
Googlebook OS Developer Guide: Linux Gaps and What Works
Googlebook OS launched October 4. Here's the developer reality: Linux GUI apps are terminal-only, Slack ...
NVIDIA DGX Station for Windows: 748GB, 20 PFLOPs
NVIDIA put a trillion-parameter AI supercomputer in a box that runs Windows. The DGX Station for Windows — announced October 7 at Microsoft’s ...
openTPU: AI Built the FPGA Accelerator That Runs Itself
An AI agent designed openTPU: a full-stack FPGA inference accelerator with custom ISA, RTL, and ...
RTX Spark Ships: Run 120B AI Models on Windows Locally
Surface Laptop Ultra with NVIDIA RTX Spark launches today — 128GB unified memory, CUDA support, ...
Vinci Raises $250M to Simulate Chips 1,000x Faster
Vinci closed a $250M Series B at a $1.5B valuation. The startup's AI physics model ...
Qualcomm Licenses Huawei’s LogicFolding: AI Inference Impact
Qualcomm just licensed Huawei's LogicFolding chip patents — a 3D stacking technique that delivered 140% ...