Laguna XS Breaks 140 TPS on Apple Machines with New FAST Mode
gajesh · x · 2026-07-30
Laguna XS announced a new "FAST" mode for Apple machines. This mode doubled their previous benchmark improvements overnight, achieving over 140 decode tokens per second (TPS) on Apple devices for the first time—reaching the generally perceived threshold for "fast" user experiences. The author noted that with speculative decoding (SD) and multi-token prediction (MTP) added to their testing suite, another massive performance jump is expected.
Related event: Laguna XS Exceeds 140 TPS on Apple Silicon(2 posts)→
More from Infra
- Vera Rubin NVL72 Delivers 5x Inference Performance Per Watt Over GB200 — downingARK · 2026-07-30
- Report: Huawei Could Meet Half of China's AI Compute Demand by 2028 — teortaxesTex · 2026-07-30
- Italy and US to Sign 'Pax Silica' Agreement Securing AI Supply Chains — Polymarket · 2026-07-30
- Analysis: High HBM Profit Margins to Persist, China 5 Years Behind in Supply — JOBhakdi · 2026-07-30
- NVIDIA Unveils Cosmos-Dreams Neural Closed-Loop Simulator at SIGGRAPH — NVIDIAAI · 2026-07-30
- Resident Arrested for Clapping at Data Center Meeting Highlights Community Pushback — 404 Media · 2026-07-30