Tensor Processing Primitives: A Programming Abstraction for Efficiency and Portability in Deep Learning & HPC Workloads
arXiv:2104.05755 · doi:10.1145/3458817.3476206
Abstract
During the past decade, novel Deep Learning (DL) algorithms, workloads and hardware have been developed to tackle a wide range of problems. Despite the advances in workload and hardware ecosystems, the programming methodology of DL systems is stagnant. DL workloads leverage either highly-optimized, yet platform-specific and inflexible kernels from DL libraries, or in the case of novel operators, reference implementations are built via DL framework primitives with underwhelming performance. This work introduces the Tensor Processing Primitives (TPP), a programming abstraction striving for efficient, portable implementation of DL workloads with high-productivity. TPPs define a compact, yet versatile set of 2D-tensor operators (or a virtual Tensor ISA), which subsequently can be utilized as building-blocks to construct complex operators on high-dimensional tensors. The TPP specification is platform-agnostic, thus code expressed via TPPs is portable, whereas the TPP implementation is highly-optimized and platform-specific. We demonstrate the efficacy and viability of our approach using standalone kernels and end-to-end DL & HPC workloads expressed entirely via TPPs that outperform state-of-the-art implementations on multiple platforms.
References in corpus (14)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
- CP2K: An Electronic Structure and Molecular Dynamics Software Package -- Quickstep: Efficient and Accurate Electronic Structure Calculations
- Inductive Representation Learning on Large Graphs
- Instance Normalization: The Missing Ingredient for Fast Stylization
- cuDNN: Efficient Primitives for Deep Learning
- Deep Learning Recommendation Model for Personalization and Recommendation Systems
- PyFR: An Open Source Framework for Solving Advection-Diffusion Type Problems on Streaming Architectures using the Flux Reconstruction Approach
- Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions
- The Deep Learning Compiler: A Comprehensive Survey
- A Survey of Deep Learning for Scientific Discovery
- Stripe: Tensor Compilation via the Nested Polyhedral Model
- Deep Graph Library Optimizations for Intel(R) x86 Architecture