D-Matrix Raptor 3D-DRAM Accelerator for Generative Inference at Hot Chips 2026

Automated excerpt

Each tensor engine needs a 128B flit per access, and with 32B delivered per column access from 32B banks, that works out to needing 4 banks per channel. d-Matrix’s die has 840 banks, 768 after 72 spares, spread across 256 channels for just 3 banks per channel. Raptor posts about 32. 6 GB/s per mm2 compared with roughly 1. 5 GB/s for the HBM parts, around 20 times the bandwidth per square millimeter, and 2. 96 mW per GB/s against 40 mW, a 13. 5x improvement.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.