GLM 5.2 Quantized Runs Locally at 16 tokens/s on M3 Ultra
antirez · x · 2026-07-05
Ivan Fioravanti shared a video demo showing that a 4bit quantized version of Zhipu's GLM 5.2 can run locally at about 16 tokens/second on a single Apple M3 Ultra equipped with 512GB of RAM, while the ds4-eval video runs on the q2 quantized version. The demo was reshared by Redis creator antirez, showcasing the real-world performance of running large models locally on high-memory Macs.
Related event: Quantized GLM-5.2 Runs Locally on M3 Ultra at 16 Tokens/s(2 posts)→
More from Infra
- Sail equivalence proof and riscv-coq cross-checks uncover real RISC-V spec bugs — xennygrimmato_ · 2026-10-27
- xAI's Colossus 1 hits ~14.6 exaflops FP64, ~5x Top500 leader — cis_female · 2026-10-27
- Lambda runs 19 B200 nodes in a 16-node power budget, boosting throughput 24% — TheZachMueller · 2026-10-27
- Podcast: Micron's HBM roadmap, Meta's AI infra, and NVIDIA's supply chain lock-in — BenBajarin · 2026-10-27
- SGLang MegaKernel + PD Pushes the TCO-Latency Frontier on GB300, SemiAnalysis Benchmarks Show — AccBalanced · 2026-10-27
- Scaling self-hosted Convex: compiling TypeScript Convex functions to WASM — leventov · 2026-10-27