zhaoxiahust/GPGPU-sim_LowCostNoC
In this project, I design a Low-Cost NoC for GPGPU based on its traffic characteristics. After my paper is accepted, the source code would be published to the public
A version of GPGPU-sim that can use multiple compute nodes to accelerate its simulation speed. It can accelerate the simulation by a factor of up to 3.9 with a maximum error rate of 0.11\% on four simulation hosts. The detailed information is in "Exploiting Parallelism in the Simulation of GPGPU Program"
Public repository record indexed from GitHub. Explore verified star velocity metrics, source code repositories, and curated developer tool directories across the GitHubRepo ecosystem.
In this project, I design a Low-Cost NoC for GPGPU based on its traffic characteristics. After my paper is accepted, the source code would be published to the public
In this project, I recover the work of Micro10 ``Throughput-effective on-chip networks for manycore accelerators''.
In this project, I use pthread to parallel the simulation of GPGPU-sim. If you want to learn more detail information, you can find my paper "Exploiting Parallelism in the Simulation of GPGPU Program"
This repo is for demonstration purposes only. Comments and issues may or may not be responded to.