Scribd Logo
Upload

Welcome to Scribd!

  • 10 views

    Survey On CNN Hardware Accelerators: Keywords

    Uploaded by

    Srikanth
    The document surveys CNN hardware accelerators. It discusses how CNNs are used for tasks like image classification and require dedicated hardware. GPUs are commonly used but FPGAs are being explored as an alternative that can surpass GPUs in speed and efficiency. The survey presents the future of CNN accelerators on reconfigurable hardware and how domain specific architectures may be more efficient than GPUS.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Survey On CNN Hardware Accelerators: Keywords

    Uploaded by

    Srikanth
    10 views1 page
    The document surveys CNN hardware accelerators. It discusses how CNNs are used for tasks like image classification and require dedicated hardware. GPUs are commonly used but FPGAs are being explored as an alternative that can surpass GPUs in speed and efficiency. The survey presents the future of CNN accelerators on reconfigurable hardware and how domain specific architectures may be more efficient than GPUS.

    Original Title

    Survey on HA Orig

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document surveys CNN hardware accelerators. It discusses how CNNs are used for tasks like image classification and require dedicated hardware. GPUs are commonly used but FPGAs are being explored as an alternative that can surpass GPUs in speed and efficiency. The survey presents the future of CNN accelerators on reconfigurable hardware and how domain specific architectures may be more efficient than GPUS.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    10 views1 page

    Survey On CNN Hardware Accelerators: Keywords

    Uploaded by

    Srikanth
    The document surveys CNN hardware accelerators. It discusses how CNNs are used for tasks like image classification and require dedicated hardware. GPUs are commonly used but FPGAs are being explored as an alternative that can surpass GPUs in speed and efficiency. The survey presents the future of CNN accelerators on reconfigurable hardware and how domain specific architectures may be more efficient than GPUS.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 1

    Survey on CNN Hardware accelerators

    Abstract
    Deep Convolutional Neural Network (CNN) is the fastest growing segment of Artificial
    Neural Network (ANN), which has shown high performance on the emergence of many
    machine learning applications and their implementations on many platforms such as CPU,
    GPU’s FPGA’s and ASIC. Some of the exciting application areas of CNN include Image
    Classification and Segmentation, Object Detection, Video Processing, Natural Language
    Processing, and Speech Recognition. The large amount of processing required by CNNs calls
    for dedicated and tailored hardware support methods. Moreover, CNN function is well suited
    to reconfigurable hardware architectures such as FPGAs. It is difficult for CPU platforms to
    offer enough computation capacity. GPU platforms are the first choice for neural network
    processes because of its high computation capacity and easy-to-use development frameworks.
    However, FPGA-based neural network hardware accelerator is becoming a research topic.
    With specifically designed hardware, FPGA is the next possible solution to surpass GPU in
    speed and energy efficiency. Various FPGA-based accelerator designs have been proposed
    with software and hardware optimization techniques to achieve high speed and energy
    efficiency. This paper presents a futuristic of CNN hardware accelerators over reconfigurable
    hardware architectures. Future opportunities in computer architectures are evolving towards
    the concept of domain specific architectures (DSAs) because of inefficiency in GP
    architectures and parallelism, efficient memory subsystems and reduction of overhead. The
    methods and tools explored in this survey represent the new trends in domain specific
    hardware accelerators and will propellent the future advances on improving performance and
    accuracy of the architecture.
    Keywords: Convolutional Neural Networks, FPGA, Hardware accelerator, ASIC, DSA

    You might also like