Price per stack and per GB
HBM3 24GB 12-Hi Stack: $200 per stack, $8.34/GB (reported). HBM3E 36GB 12-Hi Stack: $2,094 per stack, $58.17/GB (reported). HBM4 48GB 16-Hi Stack: $3,493 per stack, $72.77/GB (reported).
HBM3 is a contract reference level. The HBM3E and HBM4 levels are scarce-lot spot indications reported by TrendForce in September 2026, citing Korean press. TrendForce put them at roughly four to five times long-term-agreement pricing. They are real reported prints, but they show what a buyer without allocation pays, not what Nvidia or a hyperscaler pays under contract.
What changes between generations
HBM3 (24GB, 12-Hi here) set the baseline for the first wave of large AI accelerators. HBM3E raises per-pin speed and capacity (36GB at 12-Hi) and is the volume part in current GPUs. HBM4 doubles the interface width to 2,048 bits, moves the base die to a logic process, and reaches 48GB at 16-Hi.
Each step adds stacking layers and packaging complexity, which lowers yield. That, more than raw die cost, is why price per GB rises from one generation to the next.
How the generations interact
Suppliers shift wafers and packaging lines toward the newest generation, so older HBM becomes scarcer before demand for it fades. The first sign of a softer HBM market usually shows up in HBM3: a sustained negative 30-day move there, before HBM3E or HBM4 react.
HBM4 has no meaningful year-over-year history yet. Its chart line starts from the current reported level and is reconstructed backwards, so read its long-window shape with caution.