Module state

Module state 

Source
Expand description

2026-09-26: HardwareState, what the box was doing at one instant, and HardwareStateDelta, what changed between the captures before and after a run.

Owner: bench hardware. Invariants:

  • A reading that does not parse is None, never a default number (the parsers in super::parse).
  • A throttle counter that went backwards between captures yields None in the delta, never zero.

super::Hardware answers “which box”; this answers “what state was that box in”. Two boxes with identical fingerprints can differ in chassis temperature, throttle time and clock under load, and only the state shows it.

Memory comes from /proc/meminfo: nvidia-smi reports GB10’s framebuffer usage as N/A (fixtures/gb10_performance.txt). Cached is kept beside MemAvailable rather than folded into it.

Structs§

GpuComputeApp
2026-09-26: One GPU-resident compute process, as --query-compute-apps reports it.
HardwareState
2026-09-26: Everything the collector could read about the box at one instant. The executor captures it before a run and again when the run’s terminal frame is ready; HardwareStateDelta is the difference.
HardwareStateDelta
2026-09-26: What changed across the run.
MachineIdentity
2026-09-26: Which individual box this is. super::Hardware::gate_key gives the silicon class ("gb10") that baselines are keyed by; this tells two boxes of one class apart.
ThermalZone
2026-09-26: One /sys/class/thermal/thermal_zone* reading.
ThrottleActive
2026-09-26: Throttle reasons asserted at the instant of capture.
ThrottleCounters
2026-09-26: Cumulative throttle time per reason, in the microseconds nvidia-smi -q -d PERFORMANCE reports under Clocks Event Reasons Counters.