Fig. 4From: Analyzing big datasets of genomic sequences: fast and scalable collection of k-mer statisticsComparison of execution times of FastKmer on the 32GB for k∈{28,55}, using default (left) or custom (right) MPS-based partitioning method, for an increasing parallelism level. The number of bins is set to 8192. For the custom partitioning scheme the performance of the signature granularity is also shown, marked with a dashed lineBack to article page