News
IndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value ChainIndustryTSMC’s grip on advanced chipmaking will face competitionIndustryCadence joins European Chip Design Platform to support semiconductor startups and SMEsMaking chipsMarc Hijink on ASML’s Culture, Supply Chain, and GeopoliticsProcessorsIntel's post-Nova Lake CPUs take shape as Razor Lake desktop and laptop variants surfaceAI chipsHow Google’s New AI Chip Software Helps It Battle NvidiaIndiaMore IC Packaging Plants Emerge In IndiaIndustryUS Tightens China AI Chip Access as Nvidia, TSMC and Semiconductor Stocks SlideDesignAxelera AI moves into HPC with TitaniaMaking chipsChinese envoy says chip equipment export curbs threaten EU-China trade thawIndiaSEMICON India 2026 Day 2 Sees 25 MoUs, Strategic Collaborations Across Semiconductor Value Chain
EveryChip

High Bandwidth Memory

Stacks of DRAM dies joined by through-silicon vias and placed beside a processor on an interposer, giving huge bandwidth for GPUs and AI chips.

A diagram drawn by EveryChip.

Also: HBM, HBM2, HBM3, HBM3E, HBM4

HBM stacks eight to sixteen thin DRAM dies on a base die, linked vertically by thousands of through-silicon vias. The stack sits right next to the processor on an interposer and connects through a very wide interface, 1,024 bits per stack up to HBM3E and 2,048 in HBM4, rather than a few fast wires. JEDEC adopted the first HBM standard in 2013, and AMD's Radeon R9 Fury X of 2015 was the first GPU to use it, with memory from SK hynix. AI accelerators carry six or eight stacks. SK hynix, Samsung and Micron make it.

See also

Find chips that mention High Bandwidth Memory

The weekly digest

Every Monday: the chip of the day, new pages on EveryChip, and the conferences and anniversaries of the week ahead.