Loo.py: transformation-based code generation for GPUs and CPUs

Abstract

Today's highly heterogeneous computing landscape places a burden on programmers wanting to achieve high performance on a reasonably broad cross-section of machines. To do so, computations need to be expressed in many different but mathematically equivalent ways, with, in the worst case, one variant per target machine. Loo.py, a programming system embedded in Python, meets this challenge by defining a data model for array-style computations and a library of transformations that operate on this model. Offering transformations such as loop tiling, vectorization, storage management, unrolling, instruction-level parallelism, change of data layout, and many more, it provides a convenient way to capture, parametrize, and re-unify the growth among code variants. Optional, deep integration with numpy and PyOpenCL provides a convenient computing environment where the transition from prototype to high-performance implementation can occur in a gradual, machine-assisted form.

References in corpus (3)

• The NumPy array: a structure for efficient numerical computation
• PyCUDA and PyOpenCL: A Scripting-Based Approach to GPU Run-Time Code Generation
• Loo.py: transformation-based code generation for GPUs and CPUs

Cited by in corpus (16)

• Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions
• TVM: An Automated End-to-End Optimizing Compiler for Deep Learning
• Fast matrix-free evaluation of discontinuous Galerkin finite element operators
• Loo.py: transformation-based code generation for GPUs and CPUs
• Gauge preheating with full general relativity
• A study of vectorization for matrix-free finite element methods
• A shared compilation stack for distributed-memory parallelism in stencil DSLs
• ConfuciuX: Autonomous Hardware Resource Assignment for DNN Accelerators using Reinforcement Learning
• Using SIMD and SIMT vectorization to evaluate sparse chemical kinetic Jacobian matrices and thermochemical source terms
• Loo.py: From Fortran to performance via transformation and substitution rules
• FusionStitching: Deep Fusion and Code Generation for Tensorflow Computations on GPUs
• Array Program Transformation with Loo.py by Example: High-Order Finite Elements
• A mechanism for balancing accuracy and scope in cross-machine black-box GPU performance modeling
• A Unified, Hardware-Fitted, Cross-GPU Performance Model
• Teaching An Old Dog New Tricks: Porting Legacy Code to Heterogeneous Compute Architectures With Automated Code Translation
• Investigating the OPS intermediate representation to target GPUs in the Devito DSL