Specifications

Series

SK hynix HBM4E

Stack Height

12-Hi (12-high)

Capacity

48 GB per stack

Pin Speed

Up to 16 Gbps per pin

Power Efficiency

More than 20% better than HBM4

Heat Resistance

17% improvement over HBM4

Packaging

Advanced MR-MUF (Mass Reflow Molded Underfill)

Sample Shipment

18 June 2026 to major customers

Mass Production

2027

Public Preview

Computex 2026

Base Die

1c DRAM with TSMC 3nm logic die

Package Integrity

MR-MUF moulding material injected between stacked die

Prior Generations

HBM3, HBM3E, HBM4 manufacturing and packaging heritage

Target Platforms

NVIDIA Rubin era accelerators and hyperscaler AI training systems

Overview

SK hynix HBM4E is the 12-high, 48 GB per stack follow-on to HBM4. Samples shipped to major customers on 18 June 2026, on schedule against the company's published AI memory roadmap.

The design leans on the Advanced MR-MUF packaging process SK hynix has used across HBM3, HBM3E and HBM4. MR-MUF injects a liquid moulding material between the stacked die and then cures it, which protects the internal circuitry and reinforces the stack mechanically. On HBM4E this is what allows a 48 GB 12-high stack to hold package integrity while per-pin signalling rises to 16 Gbps.

The generational gains are quoted as more than 20% better power efficiency and 17% better heat resistance than HBM4 — both meaningful for accelerator packages where the memory stack is one of the hardest components to cool.

Key Benefits

16 Gbps per pin — the highest published HBM4E pin speed. MR-MUF packaging — proven stack reinforcement from three prior HBM generations. Better thermals — 17% improved heat resistance versus HBM4.

Applications

AI training and inference accelerators requiring 48 GB memory stacks, hyperscale AI clusters, next-generation GPU packages, and high-performance computing systems with aggressive per-package memory bandwidth targets.

Request a Quote — SK HYNIX HBM4E 12-HI 48GB — 16 GBPS MR-MUF AI MEMORY

QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.

Get Your Quote →

Related Products

SK hynix HBM4 16-Hi 48GB — 16-Layer Stack Samsung HBM4E 12-Layer 48GB — 3.6 TB/s SK hynix HBM4E 48GB — Next-Gen AI Memory