| |
 Math Products
GoFast®
Floating Point Library
GoFast® is a family of fast, IEEE 754 floating point libraries designed for embedded applications. They are fast because they are written in assembly language and specifically optimized for each processor. They are much faster than floating point libraries included with many compilers — see the tables in the datasheets in the sidebar. They
are ANSI C compatible and are designed to directly replace a C compiler’s runtime floating point support
(library or coprocessor). GoFast® boosts the performance of an application’s math calculations or
eliminates the need for hardware floating point coprocessors, in order to reduce product manufacturing
cost. GoFast® libraries are reentrant.
Floating Point Technology
GoFast is based on “Architecture Independent Technology” (AIT) and proven floating point
algorithms that were developed for over a decade. The algorithms have been thoroughly tested
using automated methods. GoFast was specifically optimized and integrated for enhanced performance on
each supported processor.
Source Library Solutions
The GoFast IEEE 754 Floating Point Libraries are composed of single precision (32-bit format) and double
precision (64-bit format) floating point routines delivered with full source code. Each library has been
optimized for speed and space. Users are provided with immediate access to an array of floating point
operations including add, subtract, multiply, divide, square root, sine, cosine, tangent, arctangent,
exponentiation, log, natural log, and ASCII and integer conversion routines to and from the floating point
format. Since source code is delivered, a GoFast® library can be customized to fit specific user application
requirements.
Conformance and Testing
The accuracy of each GoFast Floating Point Library is within one (least significant) bit for arithmetic
functions and two bits for transcendental functions, in most cases. The IEEE 754 Floating Point Format
defines special representations for underflow, overflow, and invalid operation. The GoFast library routines
use these formats and adhere to the IEEE 754 error handling procedures in all applicable cases. Quality
assurance and testing procedures have assured proper product operation. In addition, each delivery
includes target specific test programs assuring confidence of product installation and operation.
Target Processor and Compiler Support
See the list in the sidebar. GoFast libraries are specifically integrated for “link and go” operation with each compiler.
Performance Notes
- The overall benefit depends on your application. It may do a lot of math, but use mostly the simpler functions, for which the speed increase is smaller.
- Times for some functions can vary for different ranges of arguments because different code is used.
- We sell on 30-day free trial, so you can test it in your application.
Want more information?
|
|
| |
| |
 |
Register for More Info |
| |
| |
Datasheets
- x86 real mode
- x86 protected mode
Performance: See notes to the left. |
| |
"The performance of the [GoFast floating point] library has been exceptional, making our software run 3 times faster. It is not often you get a silver bullet in software, but in this case we did."
|
John Saunders
ResMed
Bella Vista NSW Australia |
|
|
eheap
