Operating System Abstractions of Hardware Accelerators on Field-programmable Gate Arrays

نویسنده

  • SIMON FRASER
چکیده

Traditionally, one of the main functions of the Operating System (OS) is to abstract the programming model from the low level details of the specific HW platform resources. However, in an FPGA-based SoC with HW accelerators, even with an OS layer, there is no unified HW/SW framework that provides: 1) transparency to the SW designer at the application level; and 2) an interface and OS support for easy HW accelerator integration by the HW designer at the platform level. This thesis presents a Front-end USEr framework, called FUSE, that introduces a set of policies and mechanisms for HW accelerator abstraction. We illustrate FUSE as an API for an embedded Linux OS with POSIX threads on Xilinx’s MicroBlaze on a Virtex5 FPGA. For three different applications and HW accelerators, we achieve performance speedups ranging from 5.8-9.0x.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Field Programmable Gate Array–based Implementation of an Improved Algorithm for Objects Distance Measurement (TECHNICAL NOTE)

In this work, the design of a low-cost, field programmable gate array (FPGA)-based digital hardware platform that implements image processing algorithms for real-time distance measurement is presented. Using embedded development kit (EDK) tools from Xilinx, the system is developed on a spartan3 / xc3s400, one of the common and low cost field programmable gate arrays from the Xilinx Spartan fami...

متن کامل

A Methodology to Design Pipelined Simulated Annealing Kernel Accelerators in Space-borne Field-Programmable Gate Arrays

A Methodology to Design Pipelined Simulated Annealing Kernel Accelerators on Space-borne Field-Programmable Gate Arrays

متن کامل

Efficient CUDA

Recent work in deep learning has created a voracious demand for more compute cycles, but with the slowdown of Moore’s law, CPUs cannot keep up with the demand. Thus, attention has turned to massively-parallel hardware accelerators, ranging from new processing units designed specifically for deep learning to the repurposing of existing technologies like Fieldprogrammable gate arrays (FPGAs) and ...

متن کامل

A Case for FPGA Accelerators in the Cloud

Hardware accelerators use custom architectures to significantly speed up computations in a wide range of domains. As performance scaling in server-class CPUs slows, we propose integration of hardware accelerators in the cloud as a way to maintain a positive performance trend. Field programmable gate arrays represent the ideal resource for such integration since they can be reprogrammed as needs...

متن کامل

A new FPGA and multi-core processor architecture for high-performance signal processing and control systems

Tenova I2S’ technology for high-performance signal processing and control systems gives unparalleled performance by implementing the systems’ real-time algorithms, processes and computations directly within a unified framework of generic programmable hardware (field programmable gate arrays), overseen by a real-time operating system supervisor. The systems’ many components and interconnections ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2011