OptiML: An Implicitly Parallel Domain-Specific Language for Machine Learning
Abstract
As the size of datasets continues to grow, machine learning applications are becoming increasingly limited by the amount of available computational power. Taking advantage of modern hardware requires using multiple parallel programming models targeted at different devices (e.g. CPUs and GPUs). However, programming these devices to run efficiently and correctly is difficult, error-prone, and results in software that is harder to read and maintain. We present OptiML, a domain-specific language (DSL) for machine learning. OptiML is an implicitly parallel, expressive and high performance alternative to MATLAB and C++. OptiML performs domain-specific analyses and optimizations and automatically generates CUDA code for GPUs. We show that OptiML outperforms explicitly parallelized MATLAB code in nearly all cases.
Cite
Text
Sujeeth et al. "OptiML: An Implicitly Parallel Domain-Specific Language for Machine Learning." International Conference on Machine Learning, 2011.Markdown
[Sujeeth et al. "OptiML: An Implicitly Parallel Domain-Specific Language for Machine Learning." International Conference on Machine Learning, 2011.](https://mlanthology.org/icml/2011/sujeeth2011icml-optiml/)BibTeX
@inproceedings{sujeeth2011icml-optiml,
title = {{OptiML: An Implicitly Parallel Domain-Specific Language for Machine Learning}},
author = {Sujeeth, Arvind K. and Lee, HyoukJoong and Brown, Kevin J. and Rompf, Tiark and Chafi, Hassan and Wu, Michael and Atreya, Anand R. and Odersky, Martin and Olukotun, Kunle},
booktitle = {International Conference on Machine Learning},
year = {2011},
pages = {609-616},
url = {https://mlanthology.org/icml/2011/sujeeth2011icml-optiml/}
}