Aspex Semiconductor Home Page
 
 
 
Linedancer
Overview
Functional Description
Linedancer HD
Overview
Functional Description
Accelera
Accelera 3000 Product Family
ASProCore IP
Development Tools
Activate Software Toolchain
WiMAX Development Kit
OpenRL Development Kit
Libraries
Image Processing Library
Open RL
Reference Code
Accelera Advanced Video Encoder
Advanced Video Compression
JPEG 2000
WIMAX (802.16d/e) PHY
3GPP Infrastructure
Product Brochures
 
 
 
 
Linedancer - Overview
Overview
Linedancer is a software programmable Extreme Processor, which can deliver the performance levels traditionally associated with expensive high-end FPGAs, but at a fraction of the cost.

Based around Aspex’s patented 6th generation ASProCore Associative String Processing array, Linedancer takes advantage of the natural parallelism present in many image and signal processing applications to perform thousands of calculations in parallel.

Unlike traditional DSPs and CPUs, which typically only process up to 8 or 16 data items in parallel, Linedancer contains 4,096 processing elements in a Single Instruction, Multiple Data (SIMD) configuration, allowing all 4,096 processors to calculate results in parallel.


Key benefits
  • Faster time-to-market
  • Lower risk
  • Lower development cost
  • Scalable price:performance
Linedancer’s 100% software-programmable model allows developers to start hardware development early in the cycle, or avoid custom hardware development altogether by using one of Aspex’s standard Accelera plug-in cards.
 

Unlike cumbersome FPGA design flows, Linedancer’s C-based development flow allows the developer to compile, run, and debug within minutes, allowing product development to take weeks instead of months.

A software based platform allows new features to be added late in the design cycle, or even deployed in the field after products have shipped, leading to significantly reduced risk.

   
 Return to top

Scalable price:performance
The one-dimensional “String” topology of ASProCore allows the processing array to be extended across multiple chips, with an increase in performance that is close to linear. Neighbouring chips can share data, and Aspex’s libraries allow the programmer to write code that is independent of the number of chips in the system.

Straightforward C-based programming model
Linedancer is programmed in C, with a small number of extensions to support parallel processing.
A Windows or Linux toolchain provides compilation and support in a PC-based development environment, based around the industry standard GNU toolset.


 

 
 
 
 Return to top
 Login / register
 Site Index
Japanese home page
 Legal
 Home
 Products
 Encoders
 Markets
 Technology
 Company
 Contact Us
 Extranet