Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Efficient Execution of Recursive Programs on Commodity Vector Hardware

Conference ·

The pursuit of computational efficiency has led to the proliferation of throughput-oriented hardware, from GPUs to increasingly-wide vector units on commodity processors and accelerators. This hardware is designed to efficiently execute data-parallel computations in a vectorized manner. However, many algorithms are more naturally expressed as divide-and-conquer, recursive, task-parallel computations; in the absence of data parallelism, it seems that such algorithms are not well-suited to throughput-oriented architectures. This paper presents a set of novel code transformations that expose the data-parallelism latent in recursive, task-parallel programs. These transformations facilitate straightforward vectorization of task-parallel programs on commodity hardware. We also present scheduling policies that maintain high utilization of vector resources while limiting space usage. Across several task-parallel benchmarks, we demonstrate both efficient vector resource utilization and substantial speedup on chips using Intel's SSE4.2 vector units as well as accelerators using Intel's AVX512 units.

Research Organization:
Pacific Northwest National Laboratory (PNNL), Richland, WA (US)
Sponsoring Organization:
USDOE
DOE Contract Number:
AC05-76RL01830
OSTI ID:
1194297
Report Number(s):
PNNL-SA-107984; KJ0402000
Country of Publication:
United States
Language:
English

Similar Records

Extracting SIMD Parallelism from Recursive Task-Parallel Programs
Journal Article · Sun Dec 01 23:00:00 EST 2019 · ACM Transactions on Parallel Computing · OSTI ID:1592696

Exploiting Vector and Multicore Parallelsim for Recursive, Data- and Task-Parallel Programs
Conference · Wed Jan 25 23:00:00 EST 2017 · OSTI ID:1349171

A Length Adaptive Algorithm-Hardware Co-design of Transformer on FPGA Through Sparse Attention and Dynamic Pipelining
Conference · Tue Aug 23 00:00:00 EDT 2022 · OSTI ID:1891848

Related Subjects