Fully Parallel Hardware / Software
نویسندگان
چکیده
The design of multi-dimensional systems using hard-ware/software codesign allows a signiicant improvement in the development cycle. This paper presents a technique that enables a design to have arbitrarily high throughput by using multi-dimensional retiming techniques while adjusting the composition of hardware and multiple software elements in order to satisfy the area requirements. A multi-dimensional graph representing the problem is transformed and scheduled such that all nodes are executed in a fully parallel way. The techniques presented are applicable to any problem which can be represented as a multi-dimensional data ow graph. Results are shown which illustrate the eeciency of the system as well as the savings achieved.
منابع مشابه
The Ramp Project’s Goal Is to Enable the Intensive, Multidisciplinary Innovation That the Computing Industry Will Need to Tackle the Problems of Parallel
......In 2005, the computer hardware industry took a historic change of direction: The major microprocessor companies all announced that their future products would be single-chip multiprocessors, and that future performance improvements would rely on software-specified parallelism rather than additional software-transparent parallelism extracted automatically by the microarchitecture. Several ...
متن کاملHigh Performance Software Coherence for Current and Future Architectures
Shared memory provides an attractive and intuitive programming model for large-scale parallel computing, but requires a coherence mechanism to allow caching for performance while ensuring that processors do not use stale data in their computation. Implementation options range from distributed shared memory emulations on networks of workstations to tightly-coupled fully cachecoherent distributed...
متن کاملHigh Performance Software Coherence for Current and Future Architectures1
Shared memory provides an attractive and intuitive programming model for large-scale parallel computing, but requires a coherence mechanism to allow caching for performance while ensuring that processors do not use stale data in their computation. Implementation options range from distributed shared memory emulations on networks of workstations to tightly coupled fully cache-coherent distribute...
متن کاملAn Intelligent Computer Interface Utilizing Parallel Picocontrollers (TECHNICAL NOTE)
The design of an interface unit is described, in which RS232 serial data is converted to latched parallel data on 22 independent lines. The data direction of each line is programmable through the serial port. Two picocontrollers are employed in a parallel processing mode to give the required number of I/O pins, and data on the shared serial line is coded to separate data streams to the individu...
متن کاملA Scalable and Efficient Peer-to-Peer Run-Time System for a Hardware Independent Software Platform
This paper introduces a run-time system that is scalable, efficient, and implements an intermediate format for hardware-independent parallel software. It is a peer-to-peer system, having no centralized functions, with peers organized into a tree-of-fully-connected-graphs. The coordination overhead (processing plus number of control messages) grows logarithmically with the number of processing n...
متن کاملذخیره 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 1996