Efficiency, Expressivity, and Extensibility in a Close-to-Metal NPU Programming Interface

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Hunhoff, Erika, Melber, Joseph, Denolf, Kristof, Bisca, Andra, Bayliss, Samuel, Neuendorffer, Stephen, Fifield, Jeff, Lo, Jack, Vasireddy, Pranathi, James-Roxby, Phil, Keller, Eric
Format:	Preprint
Published:	2025
Subjects:	Software Engineering
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866909594461143040
author	Hunhoff, Erika Melber, Joseph Denolf, Kristof Bisca, Andra Bayliss, Samuel Neuendorffer, Stephen Fifield, Jeff Lo, Jack Vasireddy, Pranathi James-Roxby, Phil Keller, Eric
author_facet	Hunhoff, Erika Melber, Joseph Denolf, Kristof Bisca, Andra Bayliss, Samuel Neuendorffer, Stephen Fifield, Jeff Lo, Jack Vasireddy, Pranathi James-Roxby, Phil Keller, Eric
contents	Accelerators such as neural processing units (NPUs) deliver an enticing balance of performance and efficiency compared to general purpose compute architectures. However, effectively leveraging accelerator capabilities is not always simple: low-level programming toolkits may require substantial developer effort while high-level programming toolkits may abstract critical optimization features. This work aims to increase efficiency of designers using IRON, a toolkit for close-to-metal NPU performance engineers. We provide an updated programmer interface to IRON containing new and refined programming constructs. The new interface includes extensible features for placement and data transformation. These contributions are evaluated in terms of 1) efficiency, with analysis showing ~26% average reduction in lines of code and decreases in Halstead metrics for a variety of designs; 2) expressivity, demonstrating the new interface supports the wide range of features and patterns already supported by IRON; and 3) extensibility, illustrating the new tooling for placement and tiling can be extended to accommodate common use-cases.
format	Preprint
id	arxiv_https___arxiv_org_abs_2504_18430
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	Efficiency, Expressivity, and Extensibility in a Close-to-Metal NPU Programming Interface Hunhoff, Erika Melber, Joseph Denolf, Kristof Bisca, Andra Bayliss, Samuel Neuendorffer, Stephen Fifield, Jeff Lo, Jack Vasireddy, Pranathi James-Roxby, Phil Keller, Eric Software Engineering Accelerators such as neural processing units (NPUs) deliver an enticing balance of performance and efficiency compared to general purpose compute architectures. However, effectively leveraging accelerator capabilities is not always simple: low-level programming toolkits may require substantial developer effort while high-level programming toolkits may abstract critical optimization features. This work aims to increase efficiency of designers using IRON, a toolkit for close-to-metal NPU performance engineers. We provide an updated programmer interface to IRON containing new and refined programming constructs. The new interface includes extensible features for placement and data transformation. These contributions are evaluated in terms of 1) efficiency, with analysis showing ~26% average reduction in lines of code and decreases in Halstead metrics for a variety of designs; 2) expressivity, demonstrating the new interface supports the wide range of features and patterns already supported by IRON; and 3) extensibility, illustrating the new tooling for placement and tiling can be extended to accommodate common use-cases.
title	Efficiency, Expressivity, and Extensibility in a Close-to-Metal NPU Programming Interface
topic	Software Engineering
url	https://arxiv.org/abs/2504.18430