COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sudusinghe, Chamika, Gerogiannis, Gerasimos, Lenadora, Damitha, Block, Charles, Torrellas, Josep, Mendis, Charith
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915342841806848
author Sudusinghe, Chamika
Gerogiannis, Gerasimos
Lenadora, Damitha
Block, Charles
Torrellas, Josep
Mendis, Charith
author_facet Sudusinghe, Chamika
Gerogiannis, Gerasimos
Lenadora, Damitha
Block, Charles
Torrellas, Josep
Mendis, Charith
contents Sparse tensor programs are essential in deep learning and graph analytics, driving the need for optimized processing. To meet this demand, specialized hardware accelerators are being developed. Optimizing these programs for accelerators is challenging for two reasons: program performance is highly sensitive to variations in sparse inputs, and early-stage accelerators rely on expensive simulators. Therefore, ML-based cost models used for optimizing such programs on general-purpose hardware are often ineffective for early-stage accelerators, as they require large datasets for proper training. To this end, we introduce COGNATE, a novel framework that leverages inexpensive data samples from general-purpose hardware (e.g., CPUs) to train cost models, followed by few-shot fine-tuning on emerging hardware. COGNATE exploits the homogeneity of input features across hardware platforms while effectively mitigating heterogeneity, enabling cost model training with just 5% of the data samples needed by accelerator-specific models to achieve comparable performance. We conduct extensive experiments to demonstrate that COGNATE outperforms existing techniques, achieving average speedups of 1.47x (up to 5.46x) for SpMM and 1.39x (up to 4.22x) for SDDMM.
format Preprint
id arxiv_https___arxiv_org_abs_2506_00424
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning
Sudusinghe, Chamika
Gerogiannis, Gerasimos
Lenadora, Damitha
Block, Charles
Torrellas, Josep
Mendis, Charith
Machine Learning
Artificial Intelligence
Hardware Architecture
Emerging Technologies
Sparse tensor programs are essential in deep learning and graph analytics, driving the need for optimized processing. To meet this demand, specialized hardware accelerators are being developed. Optimizing these programs for accelerators is challenging for two reasons: program performance is highly sensitive to variations in sparse inputs, and early-stage accelerators rely on expensive simulators. Therefore, ML-based cost models used for optimizing such programs on general-purpose hardware are often ineffective for early-stage accelerators, as they require large datasets for proper training. To this end, we introduce COGNATE, a novel framework that leverages inexpensive data samples from general-purpose hardware (e.g., CPUs) to train cost models, followed by few-shot fine-tuning on emerging hardware. COGNATE exploits the homogeneity of input features across hardware platforms while effectively mitigating heterogeneity, enabling cost model training with just 5% of the data samples needed by accelerator-specific models to achieve comparable performance. We conduct extensive experiments to demonstrate that COGNATE outperforms existing techniques, achieving average speedups of 1.47x (up to 5.46x) for SpMM and 1.39x (up to 4.22x) for SDDMM.
title COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning
topic Machine Learning
Artificial Intelligence
Hardware Architecture
Emerging Technologies
url https://arxiv.org/abs/2506.00424