▣ AI Vitals ▣ Spark vLLM Telemetry ▣ Qwen3.6-35B-A3B-FP8 ▣ Live data from the DGX Spark ▣ Booting telemetry… ▣

▣ AI Vitals ▣

Real-time telemetry from the DGX Spark
Qwen/Qwen3.6-35B-A3B-FP8

▣ Cumulative Tokens (Rolling 30 Days)

▣ Tokens Per Hour (Trend)

▣ Prompt Tokens

accumulated

▣ Gen Tokens

accumulated

▣ Requests

total successful

▣ Spec Drafts

MTP speculative

▣ Cache Hit %

prefix caching

▣ Memory

resident set
Host: gx10-fe52 · CUDA 13.0 · ARM64
GPU: GB10 Blackwell · RAM: 121.6 GB · Repo: vLLM cu130-nightly