Move NCCL and NVBandwidth into validate mode

audit/internal/platform/sat.go

@@ -366,12 +366,14 @@ func (s *System) ResetNvidiaGPU(index int) (string, error) {
 	return string(raw), err
 }
 
-// RunNCCLTests runs nccl-tests all_reduce_perf across all NVIDIA GPUs.
+// RunNCCLTests runs nccl-tests all_reduce_perf across the selected NVIDIA GPUs.
 // Measures collective communication bandwidth over NVLink/PCIe.
-func (s *System) RunNCCLTests(ctx context.Context, baseDir string, logFunc func(string)) (string, error) {
-	// detect GPU count
-	out, _ := exec.Command("nvidia-smi", "--query-gpu=index", "--format=csv,noheader").Output()
-	gpuCount := len(strings.Split(strings.TrimSpace(string(out)), "\n"))
+func (s *System) RunNCCLTests(ctx context.Context, baseDir string, gpuIndices []int, logFunc func(string)) (string, error) {
+	selected, err := resolveDCGMGPUIndices(gpuIndices)
+	if err != nil {
+		return "", err
+	}
+	gpuCount := len(selected)
 	if gpuCount < 1 {
 		gpuCount = 1
 	}
@@ -380,7 +382,7 @@ func (s *System) RunNCCLTests(ctx context.Context, baseDir string, logFunc func(
 		satJob{name: "02-all-reduce-perf.log", cmd: []string{
 			"all_reduce_perf", "-b", "512M", "-e", "4G", "-f", "2",
 			"-g", strconv.Itoa(gpuCount), "--iters", "20",
-		}},
+		}, env: nvidiaVisibleDevicesEnv(selected)},
 	), logFunc)
 }