ResearchTasksLinear Kernel PerformanceLinear Kernel PerformanceFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updated256x256x1 Matrix Multiplication NVIDIA H800 GPU (test)BWTA_QK5.54Latency (us)8Apr 7, 2026