MyPhi: Efficient Levenshtein Distance Computation on Xeon Phi Based Architectures

Gpu Acceleration of Levenshtein Distance Computation between Long Strings

“…Hence, 𝑛 𝑚𝑎𝑥−𝑅𝑇 𝑋3090 < 1.5 G. 32 bit numbers are sufficient to store the pyramid values as ED cannot be bigger than the string length. Ignoring the amount of memory of the GPU, our implementation does not support strings longer than 2 32 . Longer strings would require a piece-wise strategy and the use a bigger integer types.…”

OpenCL-based FPGA Accelerator for Semi-Global Approximate String Matching Using Diagonal Bit-Vectors

“…Some mappers incorporate the idea of using less computationalintensive alternatives to SW and use pre-alignment filters to discard alignment candidates, hence, moving from the SE approach to the Seed-Filter-and-Extend (SFE) approach. Among other algorithms, the filtering phase might use Bit-Parallel Myers (BPM) [3], which has been optimized for CPUs [4], [5], GPUs [6], and FPGAs [7], [8]. However, the SFE approach introduce some constraints that allow optimizations to be applied that are not considered when using the BPM technique.…”