high performance computing on graphics processing units: hgpu.org

Jianhui Li, Zhennan Qin, Yijie Mei, Jingze Cui, Yunfei Song, Ciyong Chen, Yifei Zhang, Longsheng Du, Xianhang Cheng, Baihui Jin, Jason Ye, Eric Lin, Dan Lavery

View

Download (PDF)

Source codes

Tags: Code generation, Compilers, Computer science, Deep learning, OpenCL, Package

January 8, 2023 by hgpu

Kernel-as-a-Service: A Serverless Interface to GPUs

Nathan Pemberton, Anton Zabreyko, Zhoujie Ding, Randy Katz, Joseph Gonzalez

View

Download (PDF)

Source codes

Tags: Cloud, Computer science, CUDA, Deep learning, Distributed computing, nVidia, Package, Tesla V100

December 25, 2022 by hgpu

A Study on the Intersection of GPU Utilization and CNN Inference

Jack Kosaian, Amar Phanishayee

View

Download (PDF)

Tags: CNN, Computer science, Deep learning, Neural networks, nVidia, Performance, Tesla V100

December 19, 2022 by hgpu

Precise Energy Consumption Measurements of Heterogeneous Artificial Intelligence Workloads

René Caspart, Sebastian Ziegler, Arvid Weyrauch, Holger Obermaier, Simon Raffeiner, Leon Pascal Schuhmacher, Jan Scholtyssek, Darya Trofimova, Marco Nolden, Ines Reinartz, Fabian Isensee, Markus Götz, Charlotte Debus

View

Download (PDF)

Source codes

Tags: Artificial intelligence, Computer science, Deep learning, Energy-efficient computing, Heterogeneous systems, nVidia, nVidia A100, nVidia V100, Package

December 11, 2022 by hgpu

* * *

high performance computing on graphics processing units: hgpu.org

Pgx: Hardware-accelerated parallel game simulation for reinforcement learning

ARK: GPU-driven Code Execution for Distributed Deep Learning

A Deep Learning Model for Loop Interchange

DeepAxe: A Framework for Exploration of Approximation and Reliability Trade-offs in DNN Accelerators

TLP: A Deep Learning-based Cost Model for Tensor Program Tuning

CMLCompiler: A Unified Compiler for Classical Machine Learning

AutoDDL: Automatic Distributed Deep Learning with Asymptotically Optimal Communication

Cramming: Training a Language Model on a Single GPU in One Day

oneDNN Graph Compiler: A Hybrid Approach for High-Performance Deep Learning Compilation

A Study on the Intersection of GPU Utilization and CNN Inference

Precise Energy Consumption Measurements of Heterogeneous Artificial Intelligence Workloads

Recent source codes

Kokkidio: Kokkos+Eigen - write fast, readable multi-platform code

LLaMA-Inference-Bench

MATAR

Over-synchronization in GPU Programs

profile_util: simple C++ library to report memory, timing, thread affinity

MaC: Mamba CPU performance predictor

Mixed-precision finite element kernels

SHMT: The Simultaneous and Heterogenous Multithreading project

Superpipeline: A Universal Approach for Reducing GPU Memory Usage in Large Models

EnergyUCB-Bandit

Most viewed papers (last 30 days)