Event ·Official first-hand

北大中科院相变忆阻器芯片比NVIDIA A100快478倍

北京大学与中国科学院联合团队在《科学》发表40nm相变忆阻器神经动力学芯片，实现2.12ms计算延迟。该芯片在脑皮层重建任务上性能超越NVIDIA A100 GPU达50至478倍。研究展示了忆阻器在类脑计算中的巨大潜力。

Current read

北大和中科院发了款新芯片，算脑皮层比A100快最多478倍，还上了《科学》封面。

8 sources72° AI heatLast updated 2026/7/14 07:58:13

Evidence

8 sources

Primary sourcePandaily

2026/7/14 07:58:13View source

Related source 1Together AI

2026/7/13 03:00:27View source

Related source 2ollama

2026/7/14 00:01:12View source

Related source 3NVIDIA AI

2026/7/14 01:14:39View source

Related source 4vLLM

2026/7/14 05:55:48View source

Related source 5Fireworks AI

2026/7/14 14:51:38View source

Related source 6elvis

2026/7/15 14:48:02View source

Related source 7Fei-Fei Li

2026/7/15 20:55:33View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story