DeepSeek Engram debate: bandwidth, not storage, is the real constraint
bookwormengr · x · 2026-09-22
In a thread on DeepSeek's Engram, bookwormengr argues sparsity (MoE, sparse attention) is already at 90%; the bigger constraint is HBM bandwidth, which is costlier than storage. DeepSeek's designs target bandwidth, and Engram's extreme KV compression benefits both. He claims Huawei's "HBM" is closer to LPDDR; zephyrz9 counters that HiZQ is real HBM (cHBM).
Related event: DeepSeek Engram Is Semantic Retrieval, Can Cut HBM Needs 40-50%(4 posts)→
More from Infra
- Banks reportedly halting compute lending as AI credit crunch begins — citrini · 2026-09-22
- California signs seven bills making AI data centers pay their own utility costs — The Verge AI · 2026-09-22
- O'Reilly builds a working data vocabulary for the semantic era, from warehouses to ontologies — rseroter · 2026-09-22
- UK's most powerful government AI supercomputer cost £225m — same as one bridge — emax · 2026-09-22
- exe.dev wins over developers: SSH into root VMs, plus the underrated Shelley coding agent — davidcrawshaw · 2026-09-22
- NVIDIA hosts trilateral meeting as Artificial Analysis becomes Korea sovereign AI evaluator — ArtificialAnlys · 2026-09-22