Digitimes October 15, 2026 By DIGITIMES negative

Nvidia's Rubin expected to use mostly 12-layer HBM4 despite reported 8-layer shift

SamsungAI / LLMSemiconductorsNVIDIA / GPUMemoryManufacturingSupply ChainRegulation

Nvidia's next-generation Rubin AI accelerators are expected to rely predominantly on 12-layer HBM4 high-bandwidth memory, despite reports that the chipmaker has asked Samsung Electronics and SK hynix to increase supplies of 8-layer products. The differing assessments suggest that changes in memory configurations do not necessarily signal weaker HBM demand, as manufacturing constraints and supply shortages are expected to support memory prices in 2027.

Read original article ↗

Premium PC demand fuels tandem OLED race between Samsung Display, LG Display

Competition in the global PC market is shifting from a focus on shipment volumes of low-cost models towards premium, hig

Digitimes · October 15, 2026

Samsung, SK Hynix stand to benefit as humanoid robot DRAM demand is forecast to jump 20-fold by 2030

Global DRAM demand for humanoid robots is projected to increase more than 20-fold by 2030 as robot deployments expand an

Digitimes · October 15, 2026

UMC declines expedited Samsung order for Galaxy Z Fold8; GF eyed as backup DDI foundry

Following stronger-than-expected demand for its Galaxy Z Fold8, Samsung Electronics submitted an expedited wafer product

Digitimes · October 15, 2026

[PHOTO] Next-generation memory architecture

Visitors take photos of mockups of zHBM and zNAND-O, 3D architecture for advanced DRAM and NAND memory chips, at Samsung

Korea Times · October 14, 2026

Nvidia's Rubin expected to use mostly 12-layer HBM4 despite reported 8-layer shift

Related Articles