Fig. 6From: Accelerating metagenomic read classification on CUDA-enabled GPUsResult kernel. Each thread gets the index and the score of the two best targets for a sequence. It also calculates the sum of all scores for this sequenceBack to article page