GLM-5.3 Flash: 320B total params, 18B active, deployed on Ascend
zephyr_z9 · x · 2026-08-26
GLM-5.3 Flash features 320B total parameters with 18B active. The FP8 weights require approximately 328 GB of storage. The model is deployed on Ascend hardware and does not yet use NVFP4.
More from Infra
- Proposal: DGX Spark-class devices sold on $200/mo contracts could mesh into a giant cheap inference network — jasonkneen · 2026-08-27
- Glean reveals model routing scores: GPT-5.6 Luna leads at $0.08 — testingcatalog · 2026-08-27
- Serving frontier models at scale on purely Chinese hardware — tokumin · 2026-08-27
- Firecrawl Launches Startup Deal: Up to $30k in Credits — devdigest · 2026-08-27
- Long Read: AI Is Buying the Data of Dead Companies — rvp · 2026-08-27
- Antirez: High prefill speed makes LLMs feel 10x more powerful — antirez · 2026-08-27