{ "device": "NVIDIA GeForce RTX 3090", "seconds": 71.4, "note": "tools/tp2_tune.py: [shape, num_sms] per matrix key (in x out x K) and row bucket", "kernels": { "5120x64768xK6": { "head": [ 3, 82 ] }, "15360x5120xK5": { "8": [ 3, 82 ], "16": [ 3, 80 ], "32": [ 5, 82 ], "64": [ 7, 82 ], "128": [ 7, 82 ] }, "5120x512xK5": { "8": [ 2, 32 ], "16": [ 2, 32 ], "32": [ 5, 32 ], "64": [ 7, 24 ], "128": [ 7, 24 ], "5": [ 2, 32 ], "10": [ 2, 32 ], "15": [ 2, 32 ], "20": [ 5, 32 ] }, "5120x1280xK5": { "5": [ 2, 80 ], "10": [ 2, 80 ], "15": [ 2, 80 ], "20": [ 5, 80 ] }, "1280x16384xK5": { "5": [ 4, 80 ], "10": [ 4, 80 ], "15": [ 3, 80 ], "20": [ 5, 82 ] }, "4096x5120xK5": { "5": [ 3, 80 ], "10": [ 3, 80 ], "15": [ 2, 80 ], "20": [ 5, 80 ] }, "5120x1152xK5": { "5": [ 2, 72 ], "10": [ 2, 72 ], "15": [ 2, 72 ], "20": [ 5, 72 ] }, "1152x5120xK5": { "5": [ 2, 82 ], "10": [ 2, 80 ], "15": [ 2, 80 ], "20": [ 5, 80 ] }, "5120x64512xK6": { "head": [ 4, 82 ] }, "5120x129280xK6": { "head": [ 0, 82 ] }, "1280x32768xK5": { "5": [ 4, 80 ], "10": [ 4, 80 ], "15": [ 4, 80 ], "20": [ 5, 82 ] }, "8192x5120xK5": { "5": [ 3, 82 ], "10": [ 3, 80 ], "15": [ 3, 80 ], "20": [ 5, 82 ] }, "5120x2304xK5": { "5": [ 2, 80 ], "10": [ 2, 80 ], "15": [ 2, 80 ], "20": [ 5, 72 ] }, "2304x5120xK5": { "5": [ 2, 80 ], "10": [ 2, 80 ], "15": [ 2, 80 ], "20": [ 5, 80 ] } }, "detail": { "5120x64768xK6": { "head": { "best": [ 3, 82 ], "us": 310.32, "default_us": 335.51, "candidates": 37 } }, "15360x5120xK5": { "8": { "best": [ 3, 82 ], "us": 88.99, "default_us": 91.9, "candidates": 49 }, "16": { "best": [ 3, 80 ], "us": 93.7, "default_us": 118.27, "candidates": 49 }, "32": { "best": [ 5, 82 ], "us": 139.11, "default_us": 234.8, "candidates": 73 }, "64": { "best": [ 7, 82 ], "us": 216.17, "default_us": 470.68, "candidates": 85 }, "128": { "best": [ 7, 82 ], "us": 428.03, "default_us": 947.81, "candidates": 85 } }, "5120x512xK5": { "8": { "best": [ 2, 32 ], "us": 18.89, "default_us": 25.91, "candidates": 49 }, "16": { "best": [ 2, 32 ], "us": 22.12, "default_us": 34.41, "candidates": 49 }, "32": { "best": [ 5, 32 ], "us": 29.59, "default_us": 66.15, "candidates": 73 }, "64": { "best": [ 7, 24 ], "us": 49.51, "default_us": 124.46, "candidates": 85 }, "128": { "best": [ 7, 24 ], "us": 95.69, "default_us": 255.64, "candidates": 85 }, "5": { "best": [ 2, 32 ], "us": 18.38, "default_us": 23.86, "candidates": 49 }, "10": { "best": [ 2, 32 ], "us": 19.81, "default_us": 28.46, "candidates": 49 }, "15": { "best": [ 2, 32 ], "us": 21.86, "default_us": 33.08, "candidates": 49 }, "20": { "best": [ 5, 32 ], "us": 25.7, "default_us": 54.53, "candidates": 61 } }, "5120x1280xK5": { "5": { "best": [ 2, 80 ], "us": 18.74, "default_us": 19.05, "candidates": 37 }, "10": { "best": [ 2, 80 ], "us": 20.63, "default_us": 21.56, "candidates": 37 }, "15": { "best": [ 2, 80 ], "us": 22.37, "default_us": 23.91, "candidates": 37 }, "20": { "best": [ 5, 80 ], "us": 26.72, "default_us": 40.29, "candidates": 49 } }, "1280x16384xK5": { "5": { "best": [ 4, 80 ], "us": 25.09, "default_us": 25.29, "candidates": 49 }, "10": { "best": [ 4, 80 ], "us": 28.06, "default_us": 30.16, "candidates": 49 }, "15": { "best": [ 3, 80 ], "us": 29.03, "default_us": 34.71, "candidates": 49 }, "20": { "best": [ 5, 82 ], "us": 41.83, "default_us": 58.57, "candidates": 61 } }, "4096x5120xK5": { "5": { "best": [ 3, 80 ], "us": 28.52, "default_us": 34.0, "candidates": 49 }, "10": { "best": [ 3, 80 ], "us": 30.77, "default_us": 47.1, "candidates": 49 }, "15": { "best": [ 2, 80 ], "us": 32.61, "default_us": 62.41, "candidates": 49 }, "20": { "best": [ 5, 80 ], "us": 39.52, "default_us": 91.85, "candidates": 61 } }, "5120x1152xK5": { "5": { "best": [ 2, 72 ], "us": 18.69, "default_us": 19.15, "candidates": 25 }, "10": { "best": [ 2, 72 ], "us": 20.37, "default_us": 20.94, "candidates": 25 }, "15": { "best": [ 2, 72 ], "us": 22.07, "default_us": 22.94, "candidates": 25 }, "20": { "best": [ 5, 72 ], "us": 25.8, "default_us": 38.81, "candidates": 37 } }, "1152x5120xK5": { "5": { "best": [ 2, 82 ], "us": 13.26, "default_us": 15.16, "candidates": 49 }, "10": { "best": [ 2, 80 ], "us": 13.88, "default_us": 18.07, "candidates": 49 }, "15": { "best": [ 2, 80 ], "us": 14.13, "default_us": 22.22, "candidates": 49 }, "20": { "best": [ 5, 80 ], "us": 16.64, "default_us": 34.05, "candidates": 61 } }, "5120x64512xK6": { "head": { "best": [ 4, 82 ], "us": 295.67, "default_us": 295.68, "candidates": 49 } }, "5120x129280xK6": { "head": { "best": [ 0, 82 ], "us": 613.58, "default_us": 613.58, "candidates": 37 } }, "1280x32768xK5": { "5": { "best": [ 4, 80 ], "us": 40.29, "default_us": 42.96, "candidates": 49 }, "10": { "best": [ 4, 80 ], "us": 42.85, "default_us": 44.08, "candidates": 49 }, "15": { "best": [ 4, 80 ], "us": 45.36, "default_us": 47.31, "candidates": 49 }, "20": { "best": [ 5, 82 ], "us": 76.95, "default_us": 87.4, "candidates": 61 } }, "8192x5120xK5": { "5": { "best": [ 3, 82 ], "us": 46.8, "default_us": 49.0, "candidates": 49 }, "10": { "best": [ 3, 80 ], "us": 49.05, "default_us": 63.69, "candidates": 49 }, "15": { "best": [ 3, 80 ], "us": 51.0, "default_us": 79.41, "candidates": 49 }, "20": { "best": [ 5, 82 ], "us": 71.68, "default_us": 125.34, "candidates": 61 } }, "5120x2304xK5": { "5": { "best": [ 2, 80 ], "us": 23.14, "default_us": 23.19, "candidates": 37 }, "10": { "best": [ 2, 80 ], "us": 24.42, "default_us": 24.52, "candidates": 37 }, "15": { "best": [ 2, 80 ], "us": 25.34, "default_us": 25.86, "candidates": 37 }, "20": { "best": [ 5, 72 ], "us": 30.72, "default_us": 45.93, "candidates": 49 } }, "2304x5120xK5": { "5": { "best": [ 2, 80 ], "us": 20.58, "default_us": 20.74, "candidates": 49 }, "10": { "best": [ 2, 80 ], "us": 21.25, "default_us": 23.6, "candidates": 49 }, "15": { "best": [ 2, 80 ], "us": 21.61, "default_us": 26.88, "candidates": 49 }, "20": { "best": [ 5, 80 ], "us": 26.01, "default_us": 46.59, "candidates": 61 } } } }