GLM-4.5-Air 106B (12B active)

User reports GLM-4.5-Air FP8 at 133 tokens/sec with SGLang on an RTX 6000 Pro Blackwell, compared with about 78 tokens/sec in vLLM. Setup is SGLang with FP8 weights, fp8_e5m2 KV cache, 128000 context length and tensor parallelism across 4 GPUs. The user asks how to reach the same 133 tokens/sec in vLLM with Triton, and whether the Cutlass design is slower than Triton.