high performance computing on graphics processing units: hgpu.org

hgpu.org » Heterogeneous systems

Demystifying Cost-Efficiency in LLM Serving over Heterogeneous GPUs

Youhe Jiang, Fangcheng Fu, Xiaozhe Yao, Guoliang He, Xupeng Miao, Ana Klimovic, Bin Cui, Binhang Yuan, Eiko Yoneki

View

Download (PDF)

Source codes

Tags: Benchmarking, Cloud, Computer science, Heterogeneous systems, LLM, nVidia, nVidia GeForce RTX 4090, nVidia H100, nVidia RTX A6000, Package

February 10, 2025 by hgpu

Towards autonomous resource management: Deep learning prediction of CPU-GPU load balancing

Inigo Gabirondo Lopez

View

Download (PDF)

Tags: AMD Radeon HD 7970, Artificial intelligence, ATI, Computer science, Deep learning, Heterogeneous systems, load balancing, nVidia, nVidia GeForce GTX 970, OpenCL

February 10, 2025 by hgpu

CPU-GPU co-execution through the exploitation of hybrid technologies via SYCL

Nozal Raúl, Jose Luis Bosque

View

Download (PDF)

Tags: Computer science, CUDA, Heterogeneous systems, Hybrid computing, LLVM, load balancing, nVidia, nVidia GeForce GT 1030, oneAPI, OpenCL, performance portability, SYCL

February 3, 2025 by hgpu

Exploring data flow design and vectorization with oneAPI for streaming applications on CPU+GPU

Cristian Campos, Rafael Asenjo, Angeles Navarro

View

Download (PDF)

Source codes

Tags: Computer science, Data parallelism, Heterogeneous systems, Intel, Intel UHD 770, oneAPI, Package, SYCL

January 27, 2025 by hgpu

Asynchronous-Many-Task Systems: Challenges and Opportunities – Scaling an AMR Astrophysics Code on Exascale machines using Kokkos and HPX

Gregor Daiß, Patrick Diehl, Jiakun Yan, John K. Holmen, Rahulkumar Gayatri, Christoph Junghans, Alexander Straub, Jeff R. Hammond, Dominic Marcello, Miwako Tsuji, Dirk Pflüger, Hartmut Kaiser

View

Download (PDF)

Source codes

Tags: AMD Radeon Instinct MI100, AMD Radeon Instinct MI250X, Astrophysics, ATI, Computer science, CUDA, Heterogeneous systems, HIP, HPC, nVidia, nVidia A100, Package, performance portability, Physics

December 29, 2024 by hgpu

Utilizing Tensor Cores in Futhark

Kristoffer August Kortbæk, Rune Ejnar Bang Lejbølle

View

Download (PDF)

Source codes

Tags: Benchmarking, Computer science, CUDA, Heterogeneous systems, High-level Languages, Matrix multiplication, nVidia, nVidia A100, Package, PTX

December 24, 2024 by hgpu

Accelerating Sparse Graph Neural Networks with Tensor Core Optimization

Ka Wai Wu

View

Download (PDF)

Tags: Computer science, CUDA, Graph, Heterogeneous systems, Neural networks, nVidia, PyTorch, Tesla V100

December 24, 2024 by hgpu

Analyzing the Performance Portability of SYCL across CPUs, GPUs, and Hybrid Systems with Protein Database Search

Manuel Costanzo, Enzo Rucci, Carlos García-Sánchez, Marcelo Naiouf, Manuel Prieto-Matías

View

Download (PDF)

Source codes

Tags: AMD Radeon RX 6700 XT, AMD Radeon RX Vega 6, ATI, Bioinformatics, Biology, Computer science, CUDA, Databases, Heterogeneous systems, HPC, Intel, Intel Arc A770, Intel UHD 630, Intel UHD 770, nVidia, nVidia GeForce GTX 1080, nVidia GeForce GTX 980, nVidia GeForce RTX 2070, nVidia GeForce RTX 3070, nVidia GeForce RTX 3090, oneAPI, Package, performance portability, SYCL, Tesla V100

December 15, 2024 by hgpu

FortranX: Harnessing Code Generation, Portability, and Heterogeneity in Fortran

Sanil Rao, Mike Franusich, , Mohammad Alaul Haque Monil, Het Mankad, Jeffrey S. Vetter, Franz Franchetti

View

Download (PDF)

Tags: AMD, ATI, Code generation, Computer science, CUDA, Differential equations, Fortran, Heterogeneous systems, HIP, nVidia, OpenCL, OpenMP, Partial differential equations, PDEs

December 8, 2024 by hgpu

Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data

Jaroslav Oľha

View

Download (PDF)

Tags: AMD Radeon RX Vega 56, ATI, Auto-Tuning, Computer science, Heterogeneous systems, nVidia, nVidia GeForce GTX 1070, nVidia GeForce GTX 680, nVidia GeForce GTX 750, nVidia GeForce RTX 2080 Ti, OpenCL, Performance, Tesla K20, Thesis

November 3, 2024 by hgpu

General-Purpose Computing on Tensor Processors

Kuan-Chieh Hsu

View

Download (PDF)

Source codes

Tags: Artificial intelligence, Computer science, CUDA, Heterogeneous systems, Machine learning, nVidia, nVidia GeForce RTX 2080, nVidia GeForce RTX 3090, nVidia Jetson Nano, Package, Sparse matrix, Thesis, TPU

October 27, 2024 by hgpu

Optimized Code Generation for Parallel and Polyhedral Loop Nests using MLIR

Arun Thangamani

View

Download (PDF)

Tags: AMD Radeon Instinct Mi50, ATI, Code generation, Computer science, CUDA, Heterogeneous systems, nVidia, nVidia A100, OpenCL, Thesis

October 13, 2024 by hgpu

Agentic Code Optimization via Compiler-LLM Cooperation

MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

Device Virtual Machine (DVM)

DVM: Real-Time Kernel Generation for Dynamic AI Models

AutoKernel: Autoresearch for GPU kernels. Give it any PyTorch model, go to sleep, wake up to optimized Triton kernels

AutoKernel: Autonomous GPU Kernel Optimization via Iterative Agent-Driven Search

SOL-ExecBench: Speed-of-Light Benchmarking for Real-World GPU Kernels Against Hardware Limits

Triton-Sanitizer: A Fast and Device-Agnostic Memory Sanitizer for Triton with Rich Diagnostic Context

* * *

high performance computing on graphics processing units: hgpu.org

Towards autonomous resource management: Deep learning prediction of CPU-GPU load balancing

CPU-GPU co-execution through the exploitation of hybrid technologies via SYCL

Exploring data flow design and vectorization with oneAPI for streaming applications on CPU+GPU

Asynchronous-Many-Task Systems: Challenges and Opportunities – Scaling an AMR Astrophysics Code on Exascale machines using Kokkos and HPX

Utilizing Tensor Cores in Futhark

Accelerating Sparse Graph Neural Networks with Tensor Core Optimization

Analyzing the Performance Portability of SYCL across CPUs, GPUs, and Hybrid Systems with Protein Database Search

FortranX: Harnessing Code Generation, Portability, and Heterogeneity in Fortran

Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data

General-Purpose Computing on Tensor Processors

Optimized Code Generation for Parallel and Polyhedral Loop Nests using MLIR

Recent source codes

Agentic Code Optimization via Compiler-LLM Cooperation

MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

Device Virtual Machine (DVM)

AutoKernel: Autoresearch for GPU kernels. Give it any PyTorch model, go to sleep, wake up to optimized Triton kernels

SOL-ExecBench: Speed-of-Light Benchmarking for Real-World GPU Kernels Against Hardware Limits

Triton-Sanitizer: A Fast and Device-Agnostic Memory Sanitizer for Triton with Rich Diagnostic Context

LLM.Q: Quantized LLM training in pure CUDA/C++

True 4-Bit Quantized CNN Training on CPU

cuFuzz: A GPU-oriented coverage-guided fuzzer for userland CUDA application

KernelSkill: A Multi-Agent Framework for GPU Kernel Optimization

Most viewed papers (last 30 days)