{"model":"glm-5.3-flash","title":"PP8/c11 live deployment receipt","status":"positive","summary":"NVFP4 target + DFlash2 K=7 live on TP1/PP8 (GPUs 1,0,2,3,4,5,7,8; GPU6 excluded): 256K max request, max-num-seqs 11, GPU cache 3,200 blocks = 2,964,172 tokens (~11.3x 256K concurrency); FULL_DECODE_ONLY graphs with DFlash buckets 8-88; healthy LMCache sidecar. Explicitly not a claim that c11…","occurred_at":"2026-09-03T18:00:00Z","config":{"serving":"TP1/PP8, max-num-seqs 11, scheduler batch cap 4,608","engine":"vLLM 0.1.dev20051+g487ecf187, image local/glm53-lmcache-hma:6950fb8ae5a8","kv":"NVFP4 target + DFlash2 K=7; 2,964,172-token cache"},"benchmarks":[{"metric":"kv_pool_tokens","value":2964172,"unit":"tokens","context":"PP8, 3,200 blocks, 256K max request"}]}