Abstract
This paper present contributions to the state-of-the art for graphics processing unit (GPU-based) embedded intelligence (EI) research for architectures and applications. This paper gives a comprehensive review and representative studies of the emerging and current paradigms for GPU-based EI with the focus on the architecture, technologies and applications: (1) First, the overview and classifications of GPU-based EI research are presented to give the full spectrum in this area that also serves as a concise summary of the scope of the paper; (2) Second, various architecture technologies for GPU-based deep learning techniques and applications are discussed in detail; and (3) Third, various architecture technologies for machine learning techniques and applications are discussed. This paper aims to give useful insights for the research area and motivate researchers towards the development of GPU-based EI for practical deployment and applications.
Highlights
Graphics Processing Unit (GPU)-Based Embedded IntelligenceEmbedded Intelligence (EI) in products or systems gives it the capability to reflect on its own operational performance
The authors proposed a circle-based pipeline where data are distributed at each computation node. Their experimental results showed that the proposed pipeline architecture and framework could reduce the overall training time by several hours compared to the baseline method
IB-MCAST header is stored in the host memory and the GPU data are collected in a single step via an IB gather operation with the GDR feature enabled
Summary
Embedded Intelligence (EI) in products or systems gives it the capability to reflect on its own operational performance. Machine learning [1], deep learning [2] and artificial intelligence (AI) have seen wide adoption across different platforms and impose new requirements on existing computing systems and architectures. They can exist solely as software, but in most cases, they require the use of hardware components to build standalone intelligent machines. Machine learning algorithms have seen wide adoption across different hardware platforms including GPU, for energy-efficiency, small formfactor and affordable devices and applications. The overview and classifications of embedded intelligence research and development on GPUs are shown in
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have