ResearchDatasetsSynthetic LLM WorkloadSynthetic LLM WorkloadFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLLM Inference EfficiencySynthetic LLM Workload Input 4096 Output 4096160.32Latency (s)4LLM Inference EfficiencySynthetic LLM Workload Input 8192 Output 4096162.78Latency (s)2