Abstract | ||
---|---|---|
The FM-index is a data structure which is seeing more and more pervasive use, in particular in the field of high-throughput bioinformatics. Algorithms based on it show a pseudo-random memory access pattern. As a consequence, they are usually bound by memory bandwidth rather than CPU usage. Naive GPU implementations are no exception. Here we show that the combination of a compact design of the FM-index and a thread-cooperative approach can be used to restore a proper balance. The resulting solution is less memory-bandwidth intensive, and allows full exploitation of the computational resources of the GPU across several GPU architectures. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1109/ISPA.2014.10 | ISPA |
Keywords | Field | DocType |
gpgpu, bioinformatics, fm-index, fine-grain parallelism, memory-level parallelism,instruction sets,indexes,bioinformatics,memory management,registers,algorithm design and analysis | Interleaved memory,Uniform memory access,Memory bandwidth,Computer science,Parallel computing,Out-of-core algorithm,Memory management,Memory map,Flat memory model,Memory footprint | Conference |
ISSN | Citations | PageRank |
2158-9178 | 1 | 0.36 |
References | Authors | |
11 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Alejandro Chacón | 1 | 44 | 2.87 |
Santiago Marco-Sola | 2 | 18 | 2.12 |
Antonio Espinosa | 3 | 43 | 7.56 |
Paolo Ribeca | 4 | 33 | 2.99 |
Juan Carlos Moure | 5 | 82 | 13.31 |