@inproceedings{anderson_optimal_2018,
	location = {Vienna, Austria},
	title = {Optimal {DNN} primitive selection with partitioned boolean quadratic programming},
	isbn = {978-1-4503-5617-6},
	url = {http://dl.acm.org/citation.cfm?doid=3168805},
	doi = {10.1145/3168805},
	abstract = {Deep Neural Networks ({DNNs}) require very large amounts of computation both for training and for inference when deployed in the field. Many different algorithms have been proposed to implement the most computationally expensive layers of {DNNs}. Further, each of these algorithms has a large number of variants, which offer different trade-offs of parallelism, data locality, memory footprint, and execution time. In addition, specific algorithms operate much more efficiently on specialized data layouts and formats. We state the problem of optimal primitive selection in the presence of data format transformations, and show that it is {NP}-hard by demonstrating an embedding in the Partitioned Boolean Quadratic Assignment problem ({PBQP}). We propose an analytic solution via a {PBQP} solver, and evaluate our approach experimentally by optimizing several popular {DNNs} using a library of more than 70 {DNN} primitives, on an embedded platform and a general purpose platform. We show experimentally that significant gains are possible versus the state of the art vendor libraries by using a principled analytic solution to the problem of layout selection in the presence of data format transformations.},
	eventtitle = {the 2018 International Symposium},
	pages = {340--351},
	booktitle = {Proceedings of the 2018 International Symposium on Code Generation and Optimization  - {CGO} 2018},
	publisher = {{ACM} Press},
	author = {Anderson, Andrew and Gregg, David},
	urldate = {2018-08-22},
	date = {2018},
	langid = {english},
}