PyPOD-GP: Using PyTorch for Accelerated Chip-Level Thermal Simulation of the GPU

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: He, Neil, Cheng, Ming-Cheng, Liu, Yu
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917861629362176
author He, Neil
Cheng, Ming-Cheng
Liu, Yu
author_facet He, Neil
Cheng, Ming-Cheng
Liu, Yu
contents The rising demand for high-performance computing (HPC) has made full-chip dynamic thermal simulation in many-core GPUs critical for optimizing performance and extending device lifespans. Proper orthogonal decomposition (POD) with Galerkin projection (GP) has shown to offer high accuracy and massive runtime improvements over direct numerical simulation (DNS). However, previous implementations of POD-GP use MPI-based libraries like PETSc and FEniCS and face significant runtime bottlenecks. We propose a $\textbf{Py}$Torch-based $\textbf{POD-GP}$ library (PyPOD-GP), a GPU-optimized library for chip-level thermal simulation. PyPOD-GP achieves over $23.4\times$ speedup in training and over $10\times$ speedup in inference on a GPU with over 13,000 cores, with just $1.2\%$ error over the device layer.
format Preprint
id arxiv_https___arxiv_org_abs_2412_06041
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PyPOD-GP: Using PyTorch for Accelerated Chip-Level Thermal Simulation of the GPU
He, Neil
Cheng, Ming-Cheng
Liu, Yu
Computational Engineering, Finance, and Science
The rising demand for high-performance computing (HPC) has made full-chip dynamic thermal simulation in many-core GPUs critical for optimizing performance and extending device lifespans. Proper orthogonal decomposition (POD) with Galerkin projection (GP) has shown to offer high accuracy and massive runtime improvements over direct numerical simulation (DNS). However, previous implementations of POD-GP use MPI-based libraries like PETSc and FEniCS and face significant runtime bottlenecks. We propose a $\textbf{Py}$Torch-based $\textbf{POD-GP}$ library (PyPOD-GP), a GPU-optimized library for chip-level thermal simulation. PyPOD-GP achieves over $23.4\times$ speedup in training and over $10\times$ speedup in inference on a GPU with over 13,000 cores, with just $1.2\%$ error over the device layer.
title PyPOD-GP: Using PyTorch for Accelerated Chip-Level Thermal Simulation of the GPU
topic Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2412.06041