UCERD & Microsoft Research Team Published Advanced Multi-accelerator Controller in ScienceDirect

The electronic forum deals with the topics related to analog and digital circuits and systems (i.e. ASIC, FPGA, DSPs, Microcontroller, Single/Multi Processors etc) and their programming such as HDL, C/C++, etc.
Forum rules
The electronic forum deals with the topics related to analog and digital circuits and systems (i.e. ASIC, FPGA, DSPs, Microcontroller, Single/Multi Processors, PCBs etc) and their programming such as HDL, C/C++, etc.

UCERD & Microsoft Research Team Published Advanced Multi-accelerator Controller in ScienceDirect

Unread postby UCERD.COM » Tue Nov 04, 2014 10:50 am

AMC: Advanced Multi-accelerator Controller


Image

Tassadaq Hussain a, b, c,
Amna Haider c,
Shakaib A. Gursal c,
Eduard Ayguadé a,
a, Barcelona Supercomputing Center,
b, Microsoft Research, Barcelona, Spain
c, Unal Center of Education Research & Development, Barcelona, Spain



Highlights
  • In this article, we propose AMC an intelligent memory system and efficient scheduler.
  • The AMC can operates without intervention of master core or Operating system.
  • It supports multi-accelerators designed by High Level Synthesis tools.
  • The system is evaluated with memory intensive accelerators tested on a Xilinx ML505 evaluation FPGA board.
  • Results show that the AMC system achieves 10.4x and 7x of speed-up compared to generic HLS multi-accelerator systems.



Abstract
The rapid advancement, use of diverse architectural features and introduction of High Level Synthesis (HLS) tools in FPGA technology have enhanced the capacity of data-level parallelism on a chip. A generic FPGA based HLS multi-accelerator system requires a microprocessor (master core) that manages memory and schedules accelerators. In a real environment, such HLS multi-accelerator systems do not give a perfect performance due to memory bandwidth issues. Thus, a system demands a memory manager and a scheduler that improves performance by managing and scheduling the multi-accelerator’s memory access patterns efficiently. In this article, we propose the integration of an intelligent memory system and efficient scheduler in the HLS-based multi-accelerator environment called Advanced Multi-accelerator Controller (AMC). The AMC system is evaluated with memory intensive accelerators, High Performance Computing (HPC) applications and implemented and tested on a Xilinx Virtex-5 ML505 evaluation FPGA board. The performance of the system is compared against the microprocessor-based systems that have been integrated with the operating system. Results show that the AMC based HLS multi-accelerator system achieves 10.4x and 7x of speedup compared to the MicroBlaze and Intel Core based HLS multi-accelerator systems.
Please Promote Our Work and website through yoursocial and professional network. We would highly appreciate your help.
Find us on FaceBook,Like and Share Please
User avatar
UCERD.COM
Site Admin
Site Admin
 
Posts: 958
Joined: Wed Jun 20, 2012 3:01 pm
Location: Barcelona
Has thanked: 13 times
Been thanked: 14 times
Blog: View Blog (1)

Return to Electronics

Who is online

Users browsing this forum: No registered users and 2 guests

cron