New C++ Datatypes Accelerate Algorithm Validation by 10X

By Dylan McGrath

July 3, 2006

(New C++ Datatypes Accelerate Algorithm Validation :  Page 1 of 1 )

Mentor Graphics has made available new high-speed datatypes based on ANSI C++ that the company claims can accelerate algorithm validation by 10 fold.

Algorithmic C datatypes are described as arbitrary-bit-width datatypes and are said to enable algorithm, system and hardware designers to precisely model bit-true behavior in C++ specifications while accelerating simulation speeds by 10-200x. Mentor said it is making the C++ datatypes immediately available to designers and EDA tool vendors through the company's Web site.

According to Mentor, algorithm and system designers need to specify bit widths in order to create bit-accurate models used for system and algorithm validation. Previous bit-width specifications require compromiseýdesigners must either use native processor precision (64-bits for integer), requiring them to truncate the results and thus introduce artificial corner cases, or create a model using an arbitrary bit width resulting in much slower simulation speeds, according to Mentor.

Based on ANSI C++, the integer and fixed-point Algorithmic C datatypes offer "the best of both worlds," according to Mentor, algorithm and system designers to specify arbitrary bit widths while improving simulation performance by 10-200x versus other datatypes such as SystemC. Arbitrary bit widths are also critical for high-level synthesis, as they allow hardware designers to make tradeoffs between hardware size and numerical precision, Mentor said.

"By offering efficient fixed-point C++ datatypes, Mentor is adding a valuable link in the C-to-RTL design chain," said Dave Burow, president and CEO of Catalytic, in a statement. "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,"

According to Mentor, the new Algorithmic C datatypes also solve problems of semantic consistency inherent in previous datatypes.

 Page 1 of 1