SCBI_MapReduce, a New Ruby Task-Farm Skeleton for Automated Parallelisation and Distribution in Chunks of Sequences: The Implementation of a Boosted Blast+
Figure 4
Behaviour of SCBI_DISTRIBUTED_BLAST. (a) BLAST+ speed-up in chunks of 100X in two different clusters using both different network protocols and queue systems. Theoretical speed-up corresponds to the one that equals the number of cores used. Speed-up was calculated dividing the time spent using 1 core by the time of the corresponding number of cores. The following execution times were used: for 50 000 reads from AC# SRR069473 in the “x86” cluster, 25.8 h (92 880 s; 1 core), 27 600 s (2 cores), 13 980 s (4 cores), 6960 s (8 cores), 3540 s (16 cores), and 1740 s (32 cores); for the 261 304 reads of AC# SRR069473 in the “x86 upgraded” cluster, 88.6 h (318 960 s; 1 core), 115 161 s (2 cores), 56 385 s (4 cores), 28 180 s (8 cores), 14 123 s (16 cores), and 7068 s (32 cores). (b) Threaded BLAST+ and SCBI_DISTRIBUTED_BLAST use differently the 8 cores available in the same computer. BLAST+ was executed with the -num_threads 8 option, and SCBI_DISTRIBUTED_BLAST was executed with the − 8 option using chunks of 100X by default in the “x86” cluster.
