Micron says it is working with Nvidia on a custom HBM4E product called NV-HBM

On the fiscal Q4 2026 call, Micron said it is working with Nvidia on a custom HBM4E implementation it calls NV-HBM, for next-generation GPUs and NVLink Fusion platforms. Micron described it as the first custom HBM4E product of its kind. HBM demand is growing faster than server DRAM demand, although HBM is currently less profitable than DRAM, and Micron says it is working to improve that margin.

• Custom HBM4E implementation named NV-HBM, developed with Nvidia
• Target platforms: next-generation GPUs and NVLink Fusion
• HBM demand growing faster than server DRAM demand, per the reporting
• HBM currently less profitable than DRAM; Micron is working to improve margins