Optimizing sDTW for AMD GPUs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929272021581824 |
|---|---|
| author | Latta-Lin, Daniel Munoz, Sofia Isadora Padilla |
| author_facet | Latta-Lin, Daniel Munoz, Sofia Isadora Padilla |
| contents | Subsequence Dynamic Time Warping (sDTW) is the metric of choice when performing many sequence matching and alignment tasks. While sDTW is flexible and accurate, it is neither simple nor fast to compute; significant research effort has been spent devising parallel implementations on the GPU that leverage efficient memory access and computation patterns, as well as features offered by specific vendors and architectures (notably NVIDIA's). We present an implementation of sDTW on AMD hardware using HIP and ROCm. Our implementation employs well-known parallel patterns, as well as lower-level features offered by ROCm. We use shuffling for intra-wavefront communication and shared memory to transfer data between consecutive wavefronts. By constraining the input data to batches of 512 queries of length 2,000, we optimized for peak performance the width of reference elements operated on by a single thread. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_06931 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Optimizing sDTW for AMD GPUs Latta-Lin, Daniel Munoz, Sofia Isadora Padilla Distributed, Parallel, and Cluster Computing Subsequence Dynamic Time Warping (sDTW) is the metric of choice when performing many sequence matching and alignment tasks. While sDTW is flexible and accurate, it is neither simple nor fast to compute; significant research effort has been spent devising parallel implementations on the GPU that leverage efficient memory access and computation patterns, as well as features offered by specific vendors and architectures (notably NVIDIA's). We present an implementation of sDTW on AMD hardware using HIP and ROCm. Our implementation employs well-known parallel patterns, as well as lower-level features offered by ROCm. We use shuffling for intra-wavefront communication and shared memory to transfer data between consecutive wavefronts. By constraining the input data to batches of 512 queries of length 2,000, we optimized for peak performance the width of reference elements operated on by a single thread. |
| title | Optimizing sDTW for AMD GPUs |
| topic | Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2403.06931 |