GLM-5.3-Flash (FP8) + DFlash2 speculative decoding on 4x NVIDIA DGX Spark / ASUS GX10: vLLM TP4, 256K context, switchless ConnectX-7 ring with patched NCCL
Switchless RoCE collectives, NCCL profiles and fabric tooling for DGX Spark inference recipes
GLM-5.3-Flash (FP8) + DFlash2 speculative decoding on 4x NVIDIA DGX Spark / ASUS GX10: vLLM TP4, 256K context, switchless ConnectX-7 ring with patched NCCL
Switchless RoCE collectives, NCCL profiles and fabric tooling for DGX Spark inference recipes