Fig. 5From: Accelerating metagenomic read classification on CUDA-enabled GPUsMerge kernel. Each thread combines two compacted target scores for the same sequence into oneBack to article page