HPTT: a high-performance tensor transposition C++ library
Authors:
Open Access
- 18 June 2017
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 56-62
Abstract
Recently we presented TTC, a domain-specific compiler for tensor transpositions. Despite the fact that the performance of the generated code is nearly optimal, due to its offline nature, TTC cannot be utilized in all the application codes in which the tensor sizes and the necessary tensor permutations are determined at runtime. To overcome this limitation, we introduce the open-source C++ library High-Performance Tensor Transposition (HPTT). Similar to TTC, HPTT incorporates optimizations such as blocking, multi-threading, and explicit vectorization; furthermore it decomposes any transposition into multiple loops around a so called micro-kernel. This modular design-inspired by BLIS-makes HPTT easy to port to different architectures, by only replacing the hand-vectorized micro-kernel (e.g.,a 4 x 4 transpose). HPTT also offers an optional autotuning framework-guided by performance heuristics-that explores a vast search space of implementations at runtime (similar to FFTW). Across a wide range of different tensor transpositions and architectures (e.g., Intel Ivy Bridge, ARMv7, IBM Power7), HPTT attains a bandwidth comparable to that of SAXPY, and yields remarkable speedups over Eigen's tensor transposition implementation. Most importantly, the integration of HPTT into the Cyclops Tensor Framework (CTF) improves the overall performance of tensor contractions by up to 3.1x.Keywords
All Related Versions
Funding Information
- DFG (GSC 111)
This publication has 18 references indexed in Scilit:
- BLIS: A Framework for Rapidly Instantiating BLAS FunctionalityACM Transactions on Mathematical Software, 2015
- Efficient 3D Transpositions in Graphics Processing UnitsInternational Journal of Parallel Programming, 2015
- An efficient tensor transpose algorithm for multicore CPU, Intel Xeon Phi, and NVidia Tesla GPUComputer Physics Communications, 2015
- Optimizing matrix transposes using a POWER7 cache model and explicit prefetchingACM SIGMETRICS Performance Evaluation Review, 2012
- P3DFFT: A Framework for Parallel Computations of Fourier Transforms in Three DimensionsSIAM Journal on Scientific Computing, 2012
- Coupled-cluster theory in quantum chemistryReviews of Modern Physics, 2007
- Design and performance of a scalable parallel community climate modelParallel Computing, 1995
- Automatic benchmark generation for cache optimization of matrix operations1995
- A fast algorithm for transposing large multidimensional image data setsUltramicroscopy, 1991
- A fifth-order perturbation comparison of electron correlation theoriesChemical Physics Letters, 1989