[ { "50.00 percentile latency (ns)": 360247788736, "90.00 percentile latency (ns)": 648454740812, "95.00 percentile latency (ns)": 684498852793, "97.00 percentile latency (ns)": 698887350571, "99.00 percentile latency (ns)": 713271381909, "99.90 percentile latency (ns)": 719784917997, "Max latency (ns)": 720498927017, "Mean latency (ns)": 360223574191, "Min duration satisfied": "Yes", "Min latency (ns)": 8141802, "Min queries satisfied": "Yes", "Mode": "PerformanceOnly", "Result is": "VALID", "SUT name": "FuriosaQueueSUT", "Samples per second": 4122.14, "Scenario": "offline", "accelerator_frequency": "", "accelerator_host_interconnect": "PCIe 4.0 x8", "accelerator_interconnect": "", "accelerator_interconnect_topology": "", "accelerator_memory_capacity": "32 GB", "accelerator_memory_configuration": "LPDDR4X", "accelerator_model_name": "FuriosaAI Warboy", "accelerator_on-chip_memories": "32 MB", "accelerators_per_node": 1, "accuracy_log_probability": 0, "accuracy_log_rng_seed": 0, "accuracy_log_sampling_target": 0, "characteristics.mAP": 22.805, "characteristics.samples_per_second": 4122.14, "characteristics.samples_per_second.normalized_per_core": 4122.14, "characteristics.samples_per_second.normalized_per_processor": 4122.14, "ck_system": "warboy_intel", "ck_used": false, "cooling": "active", "dataset": "COCO 2017 (300x300)", "dataset_link": "https://github.com/ctuning/ck/blob/master/docs/mlperf-automation/datasets/coco2017.md", "dim_x_default": "characteristics.samples_per_second", "dim_x_maximize": true, "dim_y_default": "characteristics.mAP", "dim_y_maximize": true, "division": "closed", "formal_model": "ssd-mobilenet", "formal_model_accuracy": 99.0, "formal_model_link": "https://github.com/mlcommons/ck-mlops/tree/main/package", "framework": "FuriosaAI SDK", "host_memory_capacity": "32 GB", "host_memory_configuration": "2x16 GB DDR4-3200", "host_networking": "", "host_networking_topology": "", "host_processor_caches": "16 MB IntelR Smart Cache", "host_processor_core_count": 8, "host_processor_frequency": "3.5GHz", "host_processor_interconnect": "", "host_processor_model_name": "Intel Core i9-11900k", "host_processors_per_node": 1, "host_storage_capacity": "500 GB", "host_storage_type": "NVMe SSD", "hw_notes": "", "informal_model": "ssd-mobilenet", "input_data_types": "int8", "key.accuracy": "characteristics.mAP", "max_async_queries": 1, "max_duration (ms)": 0, "max_query_count": 0, "min_duration (ms)": 600000, "min_query_count": 1, "mlperf_version": 1.1, "normalize_cores": 1, "normalize_processors": 1, "note_code": "https://github.com/mlcommons/inference_results_v1.1/tree/master/closed/FuriosaAI/code", "note_details": "https://github.com/mlcommons/inference_results_v1.1/tree/master/closed/FuriosaAI/results/warboy_intel", "number_of_nodes": 1, "operating_system": "Ubuntu 20.04.2 LTS", "other_software_stack": "", "performance_issue_same": 0, "performance_issue_same_index": 0, "performance_issue_unique": 0, "performance_sample_count": 256, "print_timestamps": 0, "problem": false, "qsl_rng_seed": 1624344308455410291, "retraining": "N", "sample_index_rng_seed": 517984244576520566, "samples_per_query": 2970000, "schedule_rng_seed": 10051496985653635065, "starting_weights_filename": "ssd_mobilenet_v1.pytorch", "status": "preview", "submitter": "FuriosaAI", "submitter_link": "https://github.com/ctuning/ck-mlperf-inference/tree/main/bench.mlperf.submitter/FuriosaAI", "sw_notes": "", "system_link": "https://github.com/ctuning/ck-mlperf-inference/tree/main/bench.mlperf.system/warboy_intel", "system_name": "Warboy", "system_type": "edge", "target_latency (ns)": 0, "target_qps": 4500, "task": "object detection", "task2": "object detection", "total_cores": 8, "uid": "7dc67cbed4aabb02", "use_accelerator": true, "weight_data_types": "int8", "weight_transformations": "quantization and calibration" }, { "50.00 percentile latency (ns)": 357469139515, "90.00 percentile latency (ns)": 643455869903, "95.00 percentile latency (ns)": 679199414526, "97.00 percentile latency (ns)": 693514122033, "99.00 percentile latency (ns)": 707820841312, "99.90 percentile latency (ns)": 714246761511, "Max latency (ns)": 714949537037, "Mean latency (ns)": 357439149079, "Min duration satisfied": "Yes", "Min latency (ns)": 15574154, "Min queries satisfied": "Yes", "Mode": "PerformanceOnly", "Result is": "VALID", "SUT name": "FuriosaAI-BasicSUT", "Samples per second": 73.8514, "Scenario": "offline", "accelerator_frequency": "", "accelerator_host_interconnect": "PCIe 4.0 x8", "accelerator_interconnect": "", "accelerator_interconnect_topology": "", "accelerator_memory_capacity": "32 GB", "accelerator_memory_configuration": "LPDDR4X", "accelerator_model_name": "FuriosaAI Warboy", "accelerator_on-chip_memories": "32 MB", "accelerators_per_node": 1, "accuracy_log_probability": 0, "accuracy_log_rng_seed": 0, "accuracy_log_sampling_target": 0, "characteristics.mAP": 22.129, "characteristics.samples_per_second": 73.8514, "characteristics.samples_per_second.normalized_per_core": 73.8514, "characteristics.samples_per_second.normalized_per_processor": 73.8514, "ck_system": "warboy_intel", "ck_used": false, "cooling": "active", "dataset": "COCO 2017 (300x300)", "dataset_link": "https://github.com/ctuning/ck/blob/master/docs/mlperf-automation/datasets/coco2017.md", "dim_x_default": "characteristics.samples_per_second", "dim_x_maximize": true, "dim_y_default": "characteristics.mAP", "dim_y_maximize": true, "division": "closed", "formal_model": "ssd-mobilenet", "formal_model_accuracy": 99.0, "formal_model_link": "https://github.com/mlcommons/ck-mlops/tree/main/package", "framework": "FuriosaAI SDK", "host_memory_capacity": "32 GB", "host_memory_configuration": "2x16 GB DDR4-3200", "host_networking": "", "host_networking_topology": "", "host_processor_caches": "16 MB IntelR Smart Cache", "host_processor_core_count": 8, "host_processor_frequency": "3.5GHz", "host_processor_interconnect": "", "host_processor_model_name": "Intel Core i9-11900k", "host_processors_per_node": 1, "host_storage_capacity": "500 GB", "host_storage_type": "NVMe SSD", "hw_notes": "", "informal_model": "ssd-resnet34", "input_data_types": "int8", "key.accuracy": "characteristics.mAP", "max_async_queries": 1, "max_duration (ms)": 0, "max_query_count": 0, "min_duration (ms)": 600000, "min_query_count": 1, "mlperf_version": 1.1, "normalize_cores": 1, "normalize_processors": 1, "note_code": "https://github.com/mlcommons/inference_results_v1.1/tree/master/closed/FuriosaAI/code", "note_details": "https://github.com/mlcommons/inference_results_v1.1/tree/master/closed/FuriosaAI/results/warboy_intel", "number_of_nodes": 1, "operating_system": "Ubuntu 20.04.2 LTS", "other_software_stack": "", "performance_issue_same": 0, "performance_issue_same_index": 0, "performance_issue_unique": 0, "performance_sample_count": 64, "print_timestamps": 0, "problem": false, "qsl_rng_seed": 1624344308455410291, "retraining": "N", "sample_index_rng_seed": 517984244576520566, "samples_per_query": 52800, "schedule_rng_seed": 10051496985653635065, "starting_weights_filename": "resnet34_tf.22.1.pb", "status": "preview", "submitter": "FuriosaAI", "submitter_link": "https://github.com/ctuning/ck-mlperf-inference/tree/main/bench.mlperf.submitter/FuriosaAI", "sw_notes": "", "system_link": "https://github.com/ctuning/ck-mlperf-inference/tree/main/bench.mlperf.system/warboy_intel", "system_name": "Warboy", "system_type": "edge", "target_latency (ns)": 0, "target_qps": 80, "task": "object detection", "task2": "object detection", "total_cores": 8, "uid": "dc934e5fa80d2558", "use_accelerator": true, "weight_data_types": "int8", "weight_transformations": "quantization and calibration" } ]