Optical Chip Boosts Inference Speed 100x
Neil_at_HackerEarth · reddit · 2026-07-15
A new achievement by Chinese researchers: Peking University developed an optical chip system claimed to boost AI inference speed by over 100x while using about one-ninth of conventional compute.
The focus is on a shift from 'stacking GPUs' to optical chip interconnects, potentially reshaping discussions on data centers and power demand. The tone remains cautious, noting significance hinges on scalability, with no further verification details.
Related event: Chinese Photonic Chips Claim 100x AI Speedup(2 posts)→
More from Infra
- vLLM eyes Rust migration as inference engines race to rewrite the stack — AccBalanced · 2026-10-27
- Anthropic's S-1 reveals up to $567B compute spend and self-built data centers — jeremyakahn · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- Anthropic S-1: CSP marketplace revenue jumped from 11% to 48% in three years — AccBalanced · 2026-10-27
- Alchemy adds effect/Workflow support across Durable Objects, Workflows and Lambda Durable Functions — samgoodwin89 · 2026-10-27