Question 1

What is HBM and why is it critical for AI chips?

Accepted Answer

HBM (High Bandwidth Memory) is a 3D-stacked DRAM technology that delivers 5–10× the memory bandwidth of standard GDDR. AI training and inference workloads are memory-bandwidth-bound, making HBM essential for GPUs like NVIDIA H100/B200 and AMD MI300X. Each HBM stack uses through-silicon vias (TSVs) to vertically connect multiple DRAM dies.

Question 2

What is the difference between HBM3, HBM3E, and HBM4?

Accepted Answer

HBM3 offers up to 819 GB/s per stack (used in H100/MI300X). HBM3E increases bandwidth to 1.18 TB/s per stack with higher density (used in H200/B200). HBM4, expected in 2025–2026, will use a new base-logic die architecture for 1.5+ TB/s per stack with up to 48GB capacity, targeting next-gen AI accelerators.

Question 3

Who are the major HBM suppliers and what is their market share?

Accepted Answer

SK Hynix leads with approximately 50–55% market share, followed by Samsung at 35–40% and Micron at 5–10%. SK Hynix was first to mass-produce HBM3E and has secured the majority of NVIDIA supply contracts. Samsung is ramping its 12-high HBM3E, while Micron supplies select designs.

HBM Market Analysis - Pricing, Supply Chain & Forecast Dashboard