high performance computing on graphics processing units: hgpu.org

Artificial intelligence (AI) depends on data. In sensitive domains – such as healthcare, security, finance, and many more – there is therefore tension between unleashing the power of AI and maintaining the confidentiality and security of the relevant data. This book – intended for researchers in academia and R&D engineers in industry – explains how […]

Intel oneAPI is a programming framework that accepts various accelerators such as GPUs, FPGAs, and multi-core CPUs, with a focus on HPC applications. Users can apply their code written in a single language, DPC++, to this heterogeneous programming environment. However, in practice, it is not easy to apply to different accelerators, especially for non-Intel devices […]

CIFAR-10 is among the most widely used datasets in machine learning, facilitating thousands of research projects per year. To accelerate research and reduce the cost of experiments, we introduce training methods for CIFAR-10 which reach 94% accuracy in 3.29 seconds, 95% in 10.4 seconds, and 96% in 46.3 seconds, when run on a single NVIDIA […]

Determining the maximum usage of random-access memory (RAM) on both the motherboard and on a graphical processing unit (GPU) over the lifetime of a computing task can be extremely useful for troubleshooting points of failure as well as optimizing memory utilization, especially within a high-performance computing (HPC) setting. While there are tools for tracking compute […]

Computational Fluid Dynamics (CFD) is the simulation of fluid flow undertaken with the use of computational hardware. The underlying equations are computationally challenging to solve and necessitate high performance computing (HPC) to resolve in a practical timeframe when a reasonable level of fidelity is required. The simulations are memory intensive, having previously been limited to […]

Modern GPUs are designed for regular problems and suffer from load imbalance when processing irregular data. Prior to our work, a domain expert selects the best kernel to map fine-grained irregular parallelism to a GPU. We instead propose Seer, an abstraction for producing a simple, reproduceable, and understandable decision tree selector model which performs runtime […]

While polyhedral compilers have shown success in implementing advanced code transformations, they still have challenges in selecting the most profitable transformations that lead to the best speedups. This has motivated the use of machine learning to build cost models to guide the search for polyhedral optimizations. State-of-the-art polyhedral compilers have demonstrated a viable proof-of-concept of […]

Modern HPC and AI Computing solutions regularly use GPUs as their main source of computational power. This creates a significant imbalance for storage operations for GPU applications, as every such storage operation has to be signalled to and handled by the CPU. In GPU4FS, we propose a radical solution to this imbalance: Move the file […]

In order to come close to peak performance, accelerators like GPUs require significant architecture-specific tuning that understand the availability of shared memory, parallelism, tensor cores, etc. Unfortunately, the pursuit of higher performance and lower costs have led to a significant diversification of architecture designs, even from the same vendor. This creates the need for performance […]

OpenMC is an open source Monte Carlo neutral particle transport application that has recently been ported to GPU using the OpenMP target offloading model. We examine the performance of OpenMC at scale on the Frontier, Polaris, and Aurora supercomputers, demonstrating that performance portability has been achieved by OpenMC across all three major GPU vendors (AMD, […]

This paper introduces a parallel implementation in CUDA/C++ of the Gaussian process with a decomposed kernel. This recent formulation, introduced by Joukov and Kulić (2022), is characterized by an approximated — but much smaller — matrix to be inverted compared to plain Gaussian process. However, it exhibits a limitation when dealing with higher-dimensional samples which […]

We investigate the solution of low-rank matrix approximation problems using the truncated SVD. For this purpose, we develop and optimize GPU implementations for the randomized SVD and a blocked variant of the Lanczos approach. Our work takes advantage of the fact that the two methods are composed of very similar linear algebra building blocks, which […]