Add scalability/ramp-up labeling, ServerPower penalty in scoring, and report improvements

- Add RampStep/RampTotal/RampRunID to NvidiaBenchmarkOptions, taskParams, and
  NvidiaBenchmarkResult so ramp-up steps can be correlated across result.json files
- Add ScalabilityScore field to NvidiaBenchmarkResult (placeholder; computed externally
  by comparing ramp-up step results sharing the same ramp_run_id)
- Propagate ramp fields through api.go (generates shared ramp_run_id at spawn time),
  tasks.go handler, and benchmark.go result population
- Apply ServerPower penalty to CompositeScore when IPMI reporting_ratio < 0.75:
  factor = ratio/0.75, applied per-GPU with a note explaining the reduction
- Add finding when server power delta exceeds GPU-reported sum by >25% (non-GPU draw)
- Report header now shows ramp step N/M and run ID instead of "parallel" when in ramp mode;
  shows scalability_score when non-zero

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-04-12 22:30:47 +03:00
parent 58a6da9b44
commit 813e2f86a9
5 changed files with 50 additions and 3 deletions

View File

@@ -594,6 +594,7 @@ func (h *handler) handleAPIBenchmarkNvidiaRun(w http.ResponseWriter, r *http.Req
rampUp = false
} else {
now := time.Now()
rampRunID := fmt.Sprintf("ramp-%s", now.UTC().Format("20060102-150405"))
var allTasks []*Task
for step := 1; step <= len(resolved); step++ {
subset := resolved[:step]
@@ -611,6 +612,9 @@ func (h *handler) handleAPIBenchmarkNvidiaRun(w http.ResponseWriter, r *http.Req
BenchmarkProfile: body.Profile,
RunNCCL: runNCCL && step == len(resolved),
ParallelGPUs: true,
RampStep: step,
RampTotal: len(resolved),
RampRunID: rampRunID,
DisplayName: stepName,
},
}