Silicon Lattice — Compute Dashboard

GPU utilization

87.4%

+4.2 pts · 24h

Tokens / second

1.84M

+12.4% wow

P99 latency

214ms

−18ms 24h

Active fleet

3,206GPU

42 draining

Inference endpoints

Live · updated 7s ago

All regions Production Canary

Endpoint

Method

P99 latency

Throughput

Status

/v1/llm/generate

POST

186 ms

412k tok/s

Healthy

/v1/vision/embed

POST

94 ms

238k req/s

Healthy

/v1/diffusion/render

POST

612 ms

18.4k img/s

Degraded

/v1/speech/transcribe

POST

228 ms

96.2k req/s

Healthy

/v1/agents/orchestrate

GET

142 ms

54.8k req/s

Draining

/v1/index/retrieve

GET

38 ms

1.21M req/s

Healthy