{ "benchmark_version": "2", "generated_at": "2026-04-17T15:41:19.563694878Z", "hostname": "debian", "server_model": "CG480-S5063", "benchmark_profile": "standard", "parallel_gpus": true, "ramp_total": 8, "ramp_run_id": "ramp-20260417-151651", "overall_status": "FAILED", "selected_gpu_indices": [ 0, 1, 2, 3, 4, 5, 6, 7 ], "findings": [ "0 of 8 GPU(s) passed the benchmark.", "GPU 0 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 1 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 2 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 3 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 4 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 5 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 6 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED.", "GPU 7 had incomplete precision coverage: int8:FAILED, fp8:FAILED, fp16:FAILED, fp32:FAILED." ], "normalization": { "status": "full", "gpus": [ { "index": 0, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 1, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 2, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 3, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 4, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 5, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 6, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" }, { "index": 7, "persistence_mode": "applied", "gpu_clock_lock_mhz": 1785, "gpu_clock_lock_status": "applied", "memory_clock_lock_mhz": 3201, "memory_clock_lock_status": "applied" } ] }, "host_config": { "cpu_model": "Intel(R) Xeon(R) 6505P", "cpu_sockets": 2, "cpu_cores": 24, "cpu_threads": 48, "mem_total_gib": 125.17120742797852 }, "cpu_load": { "avg_pct": 13.3, "max_pct": 20.1, "p95_pct": 19.1, "samples": 65, "status": "ok" }, "cooling": { "available": true, "avg_fan_rpm": 12551.062569676678, "fan_duty_cycle_available": true, "fan_duty_cycle_estimated": true, "avg_fan_duty_cycle_pct": 74.86946212697437, "p95_fan_duty_cycle_pct": 96.68469785575046, "notes": [ "fan duty cycle is estimated from the highest fan RPM observed since boot; treat it as an approximation, not a direct PWM reading" ] }, "gpus": [ { "index": 0, "uuid": "GPU-9514fbcc-bb3d-8faf-3e8b-f59b3dcd3bfb", "name": "NVIDIA H200 NVL", "bus_id": "00000000:10:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 42.5, "p95_temp_c": 43.85, "avg_power_w": 97.315, "p95_power_w": 98.0395, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0.8272106047371934, "temp_cv_pct": 3.5294117647058822, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 61.8, "p95_temp_c": 65.8, "avg_power_w": 468.82244444444456, "p95_power_w": 507.108, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 95.55555555555556, "avg_mem_usage_pct": 43.111111111111114, "clock_cv_pct": 0, "power_cv_pct": 19.23666310098599, "temp_cv_pct": 7.6629128954382075, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 56.69230769230769, "p95_temp_c": 64, "avg_power_w": 502.67538461538453, "p95_power_w": 533.738, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 79.07692307692308, "clock_cv_pct": 0, "power_cv_pct": 15.340700843700555, "temp_cv_pct": 13.206848694593024, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 66.1, "p95_temp_c": 67, "avg_power_w": 554.813, "p95_power_w": 563.286, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.6, "clock_cv_pct": 0, "power_cv_pct": 2.4256976823456045, "temp_cv_pct": 2.38725171528888, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 64.63636363636364, "p95_temp_c": 67, "avg_power_w": 516.5, "p95_power_w": 562.2349999999999, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 91.27272727272727, "avg_mem_usage_pct": 42.63636363636363, "clock_cv_pct": 0, "power_cv_pct": 9.649207157978053, "temp_cv_pct": 3.862033816524888, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 63, "p95_temp_c": 63, "avg_power_w": 495.315, "p95_power_w": 497.57349999999997, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 44.25, "clock_cv_pct": 0, "power_cv_pct": 0.3479587927999076, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 499976, "sw_thermal_slowdown_us": 499976, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 79.1773266423819, "thermal_sustain_score": 77.01126131368538, "stability_score": 99.81049064033768, "thermal_throttle_pct": 0.09475467983116138, "power_cap_throttle_pct": 0.09475467983116138, "temp_headroom_c": 24.200000000000003, "interconnect_score": 0, "server_quality_score": 86.78077264295526, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 1, "uuid": "GPU-bd48250e-65fc-018c-cad1-38b51deb8654", "name": "NVIDIA H200 NVL", "bus_id": "00000000:11:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 45, "p95_temp_c": 46.8, "avg_power_w": 101.125, "p95_power_w": 101.9665, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0.9245982694684818, "temp_cv_pct": 4.444444444444445, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 65.33333333333333, "p95_temp_c": 69, "avg_power_w": 488.67600000000004, "p95_power_w": 528.658, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 95.55555555555556, "avg_mem_usage_pct": 43.24444444444445, "clock_cv_pct": 0, "power_cv_pct": 18.798043777423782, "temp_cv_pct": 7.553777764621119, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 59.23076923076923, "p95_temp_c": 67, "avg_power_w": 514.8653846153846, "p95_power_w": 545.828, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 79, "clock_cv_pct": 0, "power_cv_pct": 15.74138305807383, "temp_cv_pct": 13.541413400246311, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 69.8, "p95_temp_c": 71, "avg_power_w": 578.1009999999999, "p95_power_w": 587.2345, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.7, "clock_cv_pct": 0, "power_cv_pct": 2.6721987904928524, "temp_cv_pct": 2.200901360420805, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 68.63636363636364, "p95_temp_c": 71, "avg_power_w": 539.5336363636363, "p95_power_w": 587.6099999999999, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 91.18181818181819, "avg_mem_usage_pct": 43.27272727272727, "clock_cv_pct": 0, "power_cv_pct": 9.66483315772293, "temp_cv_pct": 3.6369616470850277, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 67, "p95_temp_c": 67, "avg_power_w": 518.345, "p95_power_w": 520.1735, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 47.25, "clock_cv_pct": 0, "power_cv_pct": 0.41658246692557344, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 599991, "sw_thermal_slowdown_us": 599991, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 78.62875189508861, "thermal_sustain_score": 77.33866670613665, "stability_score": 99.77258126347434, "thermal_throttle_pct": 0.11370936826283333, "power_cap_throttle_pct": 0.11370936826283333, "temp_headroom_c": 21, "interconnect_score": 0, "server_quality_score": 86.69925808575731, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 2, "uuid": "GPU-eb2c2fb8-6cfb-8cc4-cfd7-9c5d2e6af0fe", "name": "NVIDIA H200 NVL", "bus_id": "00000000:37:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 44.5, "p95_temp_c": 45.85, "avg_power_w": 100.03, "p95_power_w": 100.876, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0.9397180845746254, "temp_cv_pct": 3.3707865168539324, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 63.48888888888889, "p95_temp_c": 67, "avg_power_w": 476.53400000000005, "p95_power_w": 515.906, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 95.55555555555556, "avg_mem_usage_pct": 43.48888888888889, "clock_cv_pct": 0, "power_cv_pct": 19.191615084599853, "temp_cv_pct": 7.591104158300821, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 57, "p95_temp_c": 65, "avg_power_w": 494.6261538461539, "p95_power_w": 533.464, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 72.92307692307692, "clock_cv_pct": 0, "power_cv_pct": 23.348279478191564, "temp_cv_pct": 15.433074708770722, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 67.3, "p95_temp_c": 68, "avg_power_w": 558.755, "p95_power_w": 566.4755, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.5, "clock_cv_pct": 0, "power_cv_pct": 2.30306505897872, "temp_cv_pct": 1.7643896117441185, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 66.72727272727273, "p95_temp_c": 68, "avg_power_w": 537.6118181818182, "p95_power_w": 567.235, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 49.09090909090909, "clock_cv_pct": 0, "power_cv_pct": 5.885991260052247, "temp_cv_pct": 2.1281334266775622, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 65, "p95_temp_c": 65, "avg_power_w": 502.59499999999997, "p95_power_w": 504.221, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 46.25, "clock_cv_pct": 0, "power_cv_pct": 0.3359471213122686, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 499938, "sw_thermal_slowdown_us": 499938, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 76.0950378110989, "thermal_sustain_score": 77.22668752509753, "stability_score": 99.81050504374038, "thermal_throttle_pct": 0.09474747812981255, "power_cap_throttle_pct": 0.09474747812981255, "temp_headroom_c": 23, "interconnect_score": 0, "server_quality_score": 85.9207196183551, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 3, "uuid": "GPU-554e6f2c-13e9-fc9e-6aff-3bfe8148e3ac", "name": "NVIDIA H200 NVL", "bus_id": "00000000:38:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 43, "p95_temp_c": 44.8, "avg_power_w": 97.98, "p95_power_w": 98.88000000000001, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 1.0206164523372117, "temp_cv_pct": 4.651162790697675, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 62.044444444444444, "p95_temp_c": 66, "avg_power_w": 469.41733333333315, "p95_power_w": 506.536, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 95.55555555555556, "avg_mem_usage_pct": 43.53333333333333, "clock_cv_pct": 0, "power_cv_pct": 18.603656113661604, "temp_cv_pct": 7.844248396404321, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 55.61538461538461, "p95_temp_c": 63, "avg_power_w": 480.88615384615383, "p95_power_w": 518.248, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 71.3076923076923, "clock_cv_pct": 0, "power_cv_pct": 23.280253452924775, "temp_cv_pct": 15.282132577519375, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 65.9, "p95_temp_c": 67, "avg_power_w": 547.229, "p95_power_w": 556.2850000000001, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 52.7, "clock_cv_pct": 0, "power_cv_pct": 2.662484706533015, "temp_cv_pct": 2.394496788779893, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 64.9090909090909, "p95_temp_c": 67, "avg_power_w": 531.3863636363636, "p95_power_w": 555.175, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 70, "clock_cv_pct": 0, "power_cv_pct": 4.786766568300306, "temp_cv_pct": 3.565244274890156, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 62, "p95_temp_c": 62, "avg_power_w": 504.3775, "p95_power_w": 506.0845, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 97.75, "clock_cv_pct": 0, "power_cv_pct": 0.27955383595035244, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 599992, "sw_thermal_slowdown_us": 599992, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 76.74320607721866, "thermal_sustain_score": 76.46725481078704, "stability_score": 99.77258088443742, "thermal_throttle_pct": 0.11370955778128987, "power_cap_throttle_pct": 0.11370955778128987, "temp_headroom_c": 24, "interconnect_score": 0, "server_quality_score": 85.87217062017668, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 4, "uuid": "GPU-fcacb1de-4691-5cce-a38e-a193d88d7f77", "name": "NVIDIA H200 NVL", "bus_id": "00000000:8F:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 45, "p95_temp_c": 46.8, "avg_power_w": 102.435, "p95_power_w": 103.38449999999999, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 1.0299214135793429, "temp_cv_pct": 4.444444444444445, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 67.04444444444445, "p95_temp_c": 70, "avg_power_w": 503.0633333333334, "p95_power_w": 539.196, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 94.13333333333334, "avg_mem_usage_pct": 47.06666666666667, "clock_cv_pct": 0, "power_cv_pct": 20.99256889123879, "temp_cv_pct": 7.4938176418875795, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 57.69230769230769, "p95_temp_c": 66, "avg_power_w": 506.95846153846145, "p95_power_w": 551.5600000000001, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 72.92307692307692, "clock_cv_pct": 0, "power_cv_pct": 23.340903336422265, "temp_cv_pct": 15.657444093962319, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 68.9, "p95_temp_c": 70, "avg_power_w": 581.2620000000001, "p95_power_w": 590.2625, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.6, "clock_cv_pct": 0, "power_cv_pct": 2.54415611867442, "temp_cv_pct": 2.2902371317938313, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 68.72727272727273, "p95_temp_c": 71, "avg_power_w": 563.2963636363637, "p95_power_w": 591.37, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 70.36363636363636, "clock_cv_pct": 0, "power_cv_pct": 5.34325196906135, "temp_cv_pct": 3.4033545847271447, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 65.25, "p95_temp_c": 65.85, "avg_power_w": 528.115, "p95_power_w": 529.024, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 98, "clock_cv_pct": 0, "power_cv_pct": 0.17344903498675804, "temp_cv_pct": 0.6636209990685353, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 500008, "sw_thermal_slowdown_us": 500008, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 76.4486796556414, "thermal_sustain_score": 77.51854707433726, "stability_score": 99.81047851115646, "thermal_throttle_pct": 0.09476074442177093, "power_cap_throttle_pct": 0.09476074442177093, "temp_headroom_c": 20, "interconnect_score": 0, "server_quality_score": 86.1143594234562, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 5, "uuid": "GPU-ecb4d90e-0ece-a1e7-c0d5-271d304c4699", "name": "NVIDIA H200 NVL", "bus_id": "00000000:90:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 43.5, "p95_temp_c": 44.85, "avg_power_w": 93.75, "p95_power_w": 97.611, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 4.576000000000007, "temp_cv_pct": 3.4482758620689653, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 64.42222222222222, "p95_temp_c": 67, "avg_power_w": 486.561111111111, "p95_power_w": 518.428, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 93.91111111111111, "avg_mem_usage_pct": 47.422222222222224, "clock_cv_pct": 0, "power_cv_pct": 20.87603045254485, "temp_cv_pct": 7.788767452172025, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 56.23076923076923, "p95_temp_c": 64, "avg_power_w": 493.1276923076922, "p95_power_w": 532.274, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 72.92307692307692, "clock_cv_pct": 0, "power_cv_pct": 23.41905944690355, "temp_cv_pct": 15.72414013187525, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 66.6, "p95_temp_c": 68, "avg_power_w": 559.5139999999999, "p95_power_w": 567.345, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.5, "clock_cv_pct": 0, "power_cv_pct": 2.440278136766094, "temp_cv_pct": 2.1445731016645193, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 66, "p95_temp_c": 68, "avg_power_w": 541.8245454545454, "p95_power_w": 568.165, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 70.36363636363636, "clock_cv_pct": 0, "power_cv_pct": 5.253750976085873, "temp_cv_pct": 3.3570406630727105, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 63, "p95_temp_c": 63, "avg_power_w": 510.2925, "p95_power_w": 512.052, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 97.75, "clock_cv_pct": 0, "power_cv_pct": 0.28016761975385435, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 600022, "sw_thermal_slowdown_us": 600022, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 76.45505786536796, "thermal_sustain_score": 76.63369764348393, "stability_score": 99.77256951333003, "thermal_throttle_pct": 0.11371524333498631, "power_cap_throttle_pct": 0.11371524333498631, "temp_headroom_c": 23, "interconnect_score": 0, "server_quality_score": 85.83565445798757, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 6, "uuid": "GPU-3a515237-7b67-0348-2d57-baa3936aca1c", "name": "NVIDIA H200 NVL", "bus_id": "00000000:B7:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 43.5, "p95_temp_c": 44.85, "avg_power_w": 91.945, "p95_power_w": 98.1415, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 7.48817227690467, "temp_cv_pct": 3.4482758620689653, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 63.888888888888886, "p95_temp_c": 68, "avg_power_w": 479.58555555555563, "p95_power_w": 519.832, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 95.55555555555556, "avg_mem_usage_pct": 45.44444444444444, "clock_cv_pct": 0, "power_cv_pct": 19.36181817888047, "temp_cv_pct": 7.612545609822358, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 56.76923076923077, "p95_temp_c": 65, "avg_power_w": 492, "p95_power_w": 533.042, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 71.61538461538461, "clock_cv_pct": 0, "power_cv_pct": 23.323904491680764, "temp_cv_pct": 15.953522362540005, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 67.3, "p95_temp_c": 68.55, "avg_power_w": 547.4779999999998, "p95_power_w": 569.2435, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 54.7, "clock_cv_pct": 0, "power_cv_pct": 7.537292735081571, "temp_cv_pct": 1.9990526072917847, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 67.63636363636364, "p95_temp_c": 69, "avg_power_w": 543.4218181818181, "p95_power_w": 569.115, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 48.45454545454545, "clock_cv_pct": 0, "power_cv_pct": 5.105532606248768, "temp_cv_pct": 2.0206043519318424, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 65.5, "p95_temp_c": 66, "avg_power_w": 514.1374999999999, "p95_power_w": 515.1445, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 47, "clock_cv_pct": 0, "power_cv_pct": 0.19371300061215918, "temp_cv_pct": 0.7633587786259541, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 600038, "sw_thermal_slowdown_us": 600038, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 72.87966787478256, "thermal_sustain_score": 77.16236317053293, "stability_score": 99.77256344873942, "thermal_throttle_pct": 0.11371827563029109, "power_cap_throttle_pct": 0.11371827563029109, "temp_headroom_c": 22, "interconnect_score": 0, "server_quality_score": 84.92163469309041, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] }, { "index": 7, "uuid": "GPU-5ebe5195-b37b-2a61-3563-b50fd3b3b736", "name": "NVIDIA H200 NVL", "bus_id": "00000000:B8:00.0", "vbios": "96.00.D9.00.0E", "compute_capability": "9.0", "status": "UNSUPPORTED", "power_limit_w": 600, "default_power_limit_w": 600, "max_graphics_clock_mhz": 1785, "max_memory_clock_mhz": 3201, "locked_graphics_clock_mhz": 1785, "locked_memory_clock_mhz": 3201, "baseline": { "duration_sec": 17.70195119, "samples": 2, "avg_temp_c": 42.5, "p95_temp_c": 43.85, "avg_power_w": 88.02, "p95_power_w": 97.371, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 11.804135423767327, "temp_cv_pct": 3.5294117647058822, "clock_drift_pct": 0 }, "steady": { "duration_sec": 527.65309417, "samples": 45, "avg_temp_c": 64.31111111111112, "p95_temp_c": 67, "avg_power_w": 483.15111111111116, "p95_power_w": 513.216, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 94.04444444444445, "avg_mem_usage_pct": 47.17777777777778, "clock_cv_pct": 0, "power_cv_pct": 20.70718453101131, "temp_cv_pct": 7.9545354916488, "clock_drift_pct": 0 }, "precision_steady": [ { "precision": "int8", "status": "FAILED", "steady": { "duration_sec": 58.529044441, "samples": 13, "avg_temp_c": 55.92307692307692, "p95_temp_c": 64, "avg_power_w": 492.15923076923076, "p95_power_w": 530.96, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 92.3076923076923, "avg_mem_usage_pct": 72.92307692307692, "clock_cv_pct": 0, "power_cv_pct": 23.35025789536203, "temp_cv_pct": 15.923934991725334, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp8", "status": "FAILED", "steady": { "duration_sec": 115.978129508, "samples": 10, "avg_temp_c": 66.3, "p95_temp_c": 67.55, "avg_power_w": 557.518, "p95_power_w": 565.1195, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 53.4, "clock_cv_pct": 0, "power_cv_pct": 2.3506960550068565, "temp_cv_pct": 2.02920423032786, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp16", "status": "FAILED", "steady": { "duration_sec": 179.082805727, "samples": 11, "avg_temp_c": 65.9090909090909, "p95_temp_c": 68, "avg_power_w": 540.4381818181818, "p95_power_w": 565.825, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 70.18181818181819, "clock_cv_pct": 0, "power_cv_pct": 5.09693172900808, "temp_cv_pct": 3.51115091347803, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" }, { "precision": "fp32", "status": "FAILED", "steady": { "duration_sec": 204.385589702, "samples": 4, "avg_temp_c": 63, "p95_temp_c": 63, "avg_power_w": 509.42, "p95_power_w": 511.6965, "avg_graphics_clock_mhz": 1635, "p95_graphics_clock_mhz": 1635, "avg_memory_clock_mhz": 3201, "p95_memory_clock_mhz": 3201, "avg_usage_pct": 100, "avg_mem_usage_pct": 97.25, "clock_cv_pct": 0, "power_cv_pct": 0.3365042478140501, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "notes": "phase produced no output" } ], "precision_failures": [ "int8:FAILED", "fp8:FAILED", "fp16:FAILED", "fp32:FAILED" ], "cooldown": { "duration_sec": 0, "samples": 0, "avg_temp_c": 0, "p95_temp_c": 0, "avg_power_w": 0, "p95_power_w": 0, "avg_graphics_clock_mhz": 0, "p95_graphics_clock_mhz": 0, "avg_memory_clock_mhz": 0, "p95_memory_clock_mhz": 0, "avg_usage_pct": 0, "avg_mem_usage_pct": 0, "clock_cv_pct": 0, "power_cv_pct": 0, "temp_cv_pct": 0, "clock_drift_pct": 0 }, "throttle_counters": { "sw_power_cap_us": 499987, "sw_thermal_slowdown_us": 499987, "sync_boost_us": 0, "hw_thermal_slowdown_us": 0, "hw_power_brake_slowdown_us": 0 }, "ecc": { "corrected": 0, "uncorrected": 0 }, "scores": { "compute_score": 0, "power_sustain_score": 76.64920755460673, "thermal_sustain_score": 76.13639352505359, "stability_score": 99.81048647093164, "thermal_throttle_pct": 0.09475676453418341, "power_cap_throttle_pct": 0.09475676453418341, "temp_headroom_c": 23, "interconnect_score": 0, "server_quality_score": 85.75987491227076, "composite_score": 0 }, "notes": [ "precision plan failed: signal: killed" ] } ], "interconnect": { "status": "UNSUPPORTED", "attempted": true, "supported": false, "selected_gpu_indices": [ 0, 1, 2, 3, 4, 5, 6, 7 ], "notes": [ "" ] }, "server_power": { "available": false, "notes": [ "IPMI power reading unavailable; server-side power characterization skipped" ] } }