Runtime Reconnguration Techniques for Eecient General Purpose Computation

نویسندگان

  • Bingxiong Xu
  • David H. Albonesi
چکیده

Because of their widespread use, general purpose microprocessors are called upon to execute an increasingly diverse set of applications. Due to their static organization , these devices often operate very ineeciently under these conditions, which results in lost performance opportunities and extra energy dissipation. By exploiting the partitioning of major microprocessor hardware structures, and applying runtime reconnguration techniques , microprocessor eeciency can be greatly improved while retaining the beneets of fast clock speed, dense functionality, and mature software and chip development tools. While the functionality of each hardware structure remains xed, its complexity is conngured to match application requirements at runtime. Two applications of this approach demonstrate signiicant improvements in both energy eeciency and performance.

برای دانلود رایگان 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

Loop Pipelining and Optimization for Run Time Reconnguration ?

Lack of automatic mapping techniques is a signiicant hurdle in obtaining high performance for general purpose computing on recon-gurable hardware. In this paper, we develop techniques for mapping loop computations from applications onto high performance pipelined conng-urations. Loop statements with generalized directed acyclic graph dependencies are mapped onto multiple pipeline segments. Each...

متن کامل

Eecient Communication Operations in Reconngurable Parallel Computers Eecient Communication Operations in Reconngurable Parallel Computers

Reconnguration is largely an unexplored property in the context of parallel models of computation. However, it is a powerful concept as far as massively parallel architectures are concerned, because it overcomes the constraints due to the bissection width arising in most of distributed memory machines. In this paper, we show how to use reconnguration in order to improve communication operations...

متن کامل

Eecient Communication Operations in Reconngurable Parallel Computers

Reconnguration is largely an unexplored property in the context of parallel models of computation. However, it is a powerful concept as far as massively parallel architectures are concerned, because it overcomes the constraints due to the bissection width arising in most of distributed memory machines. In this paper, we show how to use reconnguration in order to improve communication operations...

متن کامل

Runtime Mechanisms for E cient Dynamic

High performance on distributed memory machines for programming models with dynamic thread creation and multithreading requires eecient thread management and communication. Traditional mul-tithreading runtimes, consisting of few general-purpose, bundled mechanisms that assume minimal compiler and hardware support, are suitable for computations involving coarse-grained threads but provide low ee...

متن کامل

Dynamic Reconnguration with I/o Abstraction Dynamic Reconnguration with I/o Abstraction

Dynamic reconnguration is explored in the context of I/O abstraction, a new programming model that deenes the communication structure of a system in terms of connections among well-deened data interfaces for the modules in the system. The properties of I/O abstraction, particularly the clear separation of computation from communication and the availability of a module's state information, help ...

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

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2000