Run GLM 5.2 Models at Scale for Low Cost
Aizkmusic · x · 2026-07-17
The development team demonstrated how to run the GLM 5.2 model at scale at an extremely low cost (less than $0.1 per million tokens). By using 4 AMD MI325X GPUs, they achieved a generation speed of 1482 tokens/s, making the cost 3 times lower than B300 and 10 times lower than Opus.
More from Infra
- Encrypted GPT-2 speedrun challenges FHE inference: best is 3.031 s/token, goal sub-1s — tensorqt · 2026-10-27
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27
- NVIDIA claims HGX Rubin NVL8 delivers 10x tokens per megawatt vs B200 — thursdai_pod · 2026-10-27
- NVIDIA Claims HGX Rubin NVL8 Delivers 10x More Tokens per Megawatt Than B200 — thursdai_pod · 2026-10-27
- Together lets multiple LoRA post-training experiments share the same GPU capacity — togethercompute · 2026-10-27
- Commoditization dilemma: either human labor or GPUs become the commodity — _AustinCalvert_ · 2026-10-27