Hardware1 hOctober 30, 2026, 09:37

HBM4 cycle tightens as capacity and specs become critical

Rising memory performance requirements from AI chips are accelerating the shift from HBM4 to HBM4E, according to an analysis of the South Korean supply chain. Capacity constraints and spec compliance are emerging as key bottlenecks.

Digitimes

Open source

HBM4 cycle tightens as capacity and specs become critical

AI chip demand accelerates the move from HBM4 to HBM4E
Capacity and spec limits constrain the HBM4 cycle
Analysis draws on South Korean supply chain data

#HBM4#HBM4E

NVIDIA lowers listed HBM4 and NVLink bandwidth in Vera Rubin specs

Hardware

China SiC capacity tightens, orders spill over to Taiwanese foundries

Hardware

Samsung weighs 4nm expansion as HBM4 demand strains capacity

Science

HBM4 cycle tightens as capacity and specs become critical

NVIDIA lowers listed HBM4 and NVLink bandwidth in Vera Rubin specs

China SiC capacity tightens, orders spill over to Taiwanese foundries

Samsung weighs 4nm expansion as HBM4 demand strains capacity

Silicon-rich cells retain 83% capacity after 1,000 cycles, Fraunhofer says