Micron details custom base-die HBM and vertical HBM to boost tokens per watt
BenBajarin · x · 2026-10-21
Ben Bajarin stresses that HBM is the key memory segment that is co-designed and designed-in with customers—often misunderstood. Micron's slide shows three architecture stages: standard HBM (memory beside compute on an interposer), custom HBM (custom base-die logic for differentiated features), and vertical HBM (memory stacked directly on compute, shortest memory-compute path), each step raising bandwidth and tokens per watt.
More from Infra
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Jensen Huang lays out NVIDIA's full Vera Rubin agentic system: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- ONNX to Transformers.js: running models directly in your browser — JosephJacks_ · 2026-10-23
- MiniMax M3.1 Flash free for 3 days on 0G, with 1M context and 512K output — MiniMax_AI · 2026-10-23
- "We bet on Cloudflare and it paid off": R2 and D1 at billion scale — dinasaur_404 · 2026-10-23
- Public data is drying up; private training data is turning into a browsable catalog — SucceededMind · 2026-10-23