-
Notifications
You must be signed in to change notification settings - Fork 1
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
[FEATURE] GPU Utilization Monitoring During Training
area:monitoringGPU monitoring and diagnosticsGPU monitoring and diagnosticsdifficulty:lowEasy to implement (1-3 days)Easy to implement (1-3 days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#11 In SergioZ3R0/nvprobe;[FEATURE] Compute Throughput Per Precision
area:performancePerformance benchmarking and scoringPerformance benchmarking and scoringdifficulty:mediumMedium complexity (3-5 days)Medium complexity (3-5 days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#10 In SergioZ3R0/nvprobe;[FEATURE] Multimodal Throughput Benchmark
area:aiAI/ML workloads and benchmarksAI/ML workloads and benchmarksdifficulty:highHard to implement (5+ days)Hard to implement (5+ days)enhancementNew feature or requestNew feature or requestpriority:lowLow priorityLow priorityStatus: Open.#8 In SergioZ3R0/nvprobe;[FEATURE] Inference Latency Metrics (TTFT/TPOT)
area:performancePerformance benchmarking and scoringPerformance benchmarking and scoringdifficulty:mediumMedium complexity (3-5 days)Medium complexity (3-5 days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#7 In SergioZ3R0/nvprobe;[FEATURE] MLPerf Training Benchmark
area:aiAI/ML workloads and benchmarksAI/ML workloads and benchmarksdifficulty:highHard to implement (5+ days)Hard to implement (5+ days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#6 In SergioZ3R0/nvprobe;[FEATURE] Agentic Inference Benchmark (miperf-agentic)
area:aiAI/ML workloads and benchmarksAI/ML workloads and benchmarksdifficulty:highHard to implement (5+ days)Hard to implement (5+ days)enhancementNew feature or requestNew feature or requestpriority:lowLow priorityLow priorityStatus: Open.#5 In SergioZ3R0/nvprobe;[FEATURE] vLLM Inference Benchmark
area:aiAI/ML workloads and benchmarksAI/ML workloads and benchmarksdifficulty:highHard to implement (5+ days)Hard to implement (5+ days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#4 In SergioZ3R0/nvprobe;[FEATURE] NCCL Networking Benchmark
area:networkingNetwork and inter-GPU communicationNetwork and inter-GPU communicationdifficulty:mediumMedium complexity (3-5 days)Medium complexity (3-5 days)enhancementNew feature or requestNew feature or requestpriority:mediumMedium priorityMedium priorityStatus: Open.#3 In SergioZ3R0/nvprobe;