AI inference
Hardware
Intel Details 'Crescent Island' AI Inference GPU at Hot Chips 2026
Intel has unveiled architectural details of its upcoming Crescent Island data center and workstation GPU, a 350W, air-cooled AI inference accelerator built on the new Xe3P architecture with up to 480 GB of LPDDR5X memory.
2 outletsHardwareSanDisk and SK hynix Release High Bandwidth Flash Specification for AI Inference
SanDisk and SK hynix have published the first technical specification for High Bandwidth Flash (HBF) through the Open Compute Project, positioning the new NAND-based memory tier as a bridge between HBM and enterprise SSDs for AI inference systems.
3 outlets