high performance computing on graphics processing units: hgpu.org

Driven by the advancements in generative AI, large machine learning models have revolutionized domains such as image processing, audio synthesis, and speech recognition. While server-based deployments remain the locus of peak performance, the imperative for on-device inference, necessitated by privacy and efficiency considerations, persists. Recognizing GPUs as the on-device ML accelerator with the widest reach, […]

Deep Learning (DL) has become a cornerstone of modern Artificial Intelligence (AI), powering applications across healthcare, computer vision, and autonomous systems. However, executing DL inference on resource-constrained end devices—such as smartphones and IoT hardware—poses challenges due to limited computational resources, energy constraints, and real-time requirements. This thesis addresses the optimization of DL inference on Heterogeneous […]

Dynamic memory allocation is not traditionally available in kernels running on GPUs. This work aims to build on Ouroboros, an efficient dynamic memory management library for CUDA applications, by porting the code to SYCL, a cross-platform accelerator API. Since SYCL can be compiled to a CUDA backend, it is possible to compare the performance of […]

Modern cutting-edge AI applications are being developed over fast-evolving, heterogeneous, nascent hardware devices. This requires frequent reworking of the AI software stack to adopt bottom-up changes from new hardware, which takes time for general-purpose software libraries. Consequently, real applications often develop custom software stacks optimized for their specific workloads and hardware. Custom stacks help in […]

Introduction: Unity is a powerful and versatile tool for creating real-time experiments. It includes a built-in compute shader language, a C-like programming language designed for massively parallel General-Purpose GPU (GPGPU) computing. However, as Unity is primarily developed for multi-platform game creation, its compute shader language has several limitations, including the lack of multi-GPU computation support […]

The increasing scale of deep learning models has led to the development of various parallelization strategies for distributed training across accelerators. For example, fully sharded approaches like DeepSpeed ZeRO-3 and FSDP partition the parameters of each layer across multiple GPUs and gather them through communication when needed. These methods rely on optimizations such as prefetching, […]

Large language models (LLMs) have demonstrated significant potential in code generation tasks. However, there remains a performance gap between open-source and closed-source models. To address this gap, existing approaches typically generate large amounts of synthetic data for fine-tuning, which often leads to inefficient training. In this work, we propose a data selection strategy in order […]

The surging demand for GPUs in datacenters for machine learning (ML) has made efficient GPU utilization crucial. However, meeting the diverse needs of ML models while optimizing resource usage is challenging. To enable transparent, fine-grained GPU management that maximizes utilization and energy efficiency while maintaining strong isolation, an operating system (OS) approach is needed. This […]

Training large language models requires extensive processing, made possible by many high-performance computing resources. This study compares multi-node and multi-GPU environments for training large language models of electrocardiograms. It provides a detailed mapping of current frameworks for distributed deep learning in multinode and multi-GPU settings, including Horovod from Uber, DeepSpeed from Microsoft, and the built-in […]

CUDA (Compute Unified Device Architecture) parallel programming significantly improves computational efficiency across multiple fields. However, converting serial C code to CUDA poses challenges for non-experts, and traditional tools struggle with complex patterns. While LLMs (Large Language Models) enable automatic parallelization of complex patterns, they may generate CUDA code with synchronization and memory management issues. There […]

GigaAPI is a user-space API that simplifies multi-GPU programming, bridging the gap between the capabilities of parallel GPU systems and the ability of developers to harness their full potential. The API offers a comprehensive set of functionalities, including fundamental GPU operations, image processing, and complex GPU tasks, abstracting away the intricacies of low-level CUDA and […]

Compute nodes on modern heterogeneous supercomputing systems comprise CPUs, GPUs, and high-speed network interconnects (NICs). Parallelization is identified as a technique for effectively utilizing these systems to execute scalable simulation and deep learning workloads. The resulting inter-process communication from the distributed execution of these parallel workloads is one of the key factors contributing to its […]