Samsung HBM output for U.S. customers slips a quarter after tool delays
High-bandwidth memory remains the hidden bottleneck in every GPU allocation meeting.
James WhitakerTechnology Editor
Earth viewed from orbit at night with city lights
AUSTIN, Texas — High-bandwidth memory remains the hidden bottleneck in every GPU allocation meeting.
Two U.S. server builders said their HBM-attached accelerator shipments slipped after Samsung pushed a process tool install, forcing a reshuffle of Nvidia and AMD board mixes.
James Whitaker
Technology Editor
Reports on semiconductors, cloud infrastructure, and the industrial politics of AI.