HALO 1.0: A Hardware-agnostic Accelerator Orchestration Framework for Enabling Hardware-agnostic Programming with True Performance Portability for Heterogeneous HPC
This paper presents HALO 1.0, an open-ended extensible multi-agent software framework that implements a set of proposed hardware-agnostic accelerator orchestration (HALO) principles. HALO implements a novel compute-centric message passing interface (C^2MPI) specification for enabling the performance portable execution of a hardware-agnostic host application across heterogeneous accelerators. The experiment results of evaluating eight widely used HPC subroutines based on Intel Xeon E5-2620 CPUs, Intel Arria 10 GX FPGAs, and NVIDIA GeForce RTX 2080 Ti GPUs show that HALO 1.0 allows for a unified control flow for host programs to run across all the computing devices with a consistently top performance portability score, which is up to five orders of magnitude higher than the OpenCL-based solution.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
HALO: Hardware-aware quantization with low critical-path-delay weights for LLM acceleration
Quantization is critical for efficiently deploying large language models (LLMs). Yet conventional methods remain hardware-agnostic, limited to bit-width constraints, and do not account for intrinsic circuit characteristi…
QuantizationAdaptive Orchestration for Large-Scale Inference on Heterogeneous Accelerator Systems Balancing Cost, Performance, and Resilience
The surge in generative AI workloads has created a need for scalable inference systems that can flexibly harness both GPUs and specialized accelerators while containing operational costs. This paper proposes a hardware-a…
Union: A Unified HW-SW Co-Design Ecosystem in MLIR for Evaluating Tensor Operations on Spatial Accelerators
To meet the extreme compute demands for deep learning across commercial and scientific applications, dataflow accelerators are becoming increasingly popular. While these "domain-specific" accelerators are not fully progr…
HALOC: Hardware-Aware Automatic Low-Rank Compression for Compact Neural Networks
Low-rank compression is an important model compression strategy for obtaining compact neural network models. In general, because the rank values directly determine the model complexity and model accuracy, proper selectio…
GPULow-rank compressionModel CompressionHigh-Resolution Image Reconstruction with Unsupervised Learning and Noisy Data Applied to Ion-Beam Dynamics for Particle Accelerators
Image reconstruction in the presence of severe degradation remains a challenging inverse problem, particularly in beam diagnostics for high-energy physics accelerators. As modern facilities demand precise detection of be…
Image Reconstruction