Sophon PFG-1: a monolithic-3D AI ASIC with 330 GB of on-die DRAM and no HBM
The Sophon PFG-1 is a new monolithic 3D AI ASIC that integrates 330 GB of on-die DRAM, eliminating the need for off-die HBM. This architecture allows for significantly higher throughput in low-batch training and inference compared to current GPU standards.
Why it matters
This represents a major technical shift in AI hardware design, potentially overcoming memory bandwidth bottlenecks in large-scale model training.
PFG-1 "Sophon" is a unified training-and-inference die on a 750 mm², 32-tier 2D Transition-Metal Dichalcogenide (TMD) Monolithic 3D (M3D) platform. Weights, gradients, and optimizer state reside in on-die 2T0C 2D-TMD gain-cell DRAM; because the array is fully read-write, the same silicon executes BF16 forward/backward training passes and serves low-batch decode at the compute-bound rate.
Technical analysis of hardware specifications without political or commercial bias.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in