Improved C++ data types expedite algorithm validation -- ADTmag

Improved C++ data types expedite algorithm validation

By ADT Staff
June 30, 2006

Developers now have new tools for building bit-true behavior in C++ for algorithm systems and hardware. And the new data types operate at simulation speeds 10 to 200 times faster than traditional bit-width specifications.

Mentor Graphics released Algorithmic C data types this week. Until now developers often resorted to native processor precision or a created model using a random bit width¾both resulting in slow simulation speeds. But reps at the Wilsonville, Ore.-based company say their new integer and fixed-point data type enables users to deploy arbitrary bit widths all while improving performance.

“By offering efficient fixed-point C++ data types, Mentor is adding a valuable link in the C-to-RTL design chain. Mentor’s open standard enables Catalytic’s hardware synthesis users to conduct faster algorithm validation as they convert their floating-point MATLAB descriptions to fixed-point ANSI C++ models and ensure bit-width precision throughout the entire implementation flow,” said Dave Burow, president and CEO of Catalytic.

Algorithmic C is also said to solve problems stemming from semantic inconsistency. For those with experience in ANSI C++, the Algorithmic C semantics are easily to adjust to. Before, designers would find stumbling blocks between System C and arbitrary precision data types, and had to learn coding for both data types before integration.

Mentor is scheduled to offer a suite seminar on the new AC data types at the 43rd Annual Design Automation Conference, in San Francisco. Algorithmic C is available now, free of charge, for electronics designers and electronic design automation tool vendors.