Find Articles in:
All
Business
Reference
Technology
News
Lifestyle

Manufacturing Industry

VLSI Technology to use fuzzy logic from Togai InfraLogic in ASIC line

Electronic News, July 20, 1992 by J. Robert Lineback, Robert Ristlehueber

MUNICH--VLSI Technology last week said it will attempt to incorporate fuzzy logic in its ASIC offerings through a deal with Togai InfraLogic Inc. of Irvine, Calif.

As expected (Antenna, July 13), the two firms signed a joint development and marketing deal which will combine Togai's expertise in fuzzy logic with VLSI's ASIC capabilities. The two are particularly aiming at embedded control applications.

Later this year, VLSI also plans to use Togai's new Fuzzy Computational Acceleration (FCA) technology to offer standard products such as fuzzy coprocessors for its 32-bit ARM RISC processors.

The announcement, made in Munich, is the latest move to drive fuzzy logic technology into embedded control applications, which today are primarily served by binary-based microcontrollers. Fuzzy logic--with its ability to deal with inferences and process imprecise inputs--offers performance improvements in a wide range of control applications and is spreading to home appliances, military avionics and automotive systems.

The Togai core is scalable to offer a fuzzy processing block that is anywhere from 8 to 32 bits wide. In a 10-bit configuration, the FCA core is capable of handling over 1 million fuzzy rule evaluations per second at 20MHz. That performance is about three to five times faster than the core of Togai's existing FC110 stand-alone fuzzy logic processor, which has been designed into a number of consumer product applications in Japan, said Masaki Togai, chairman and CEO of Togai InfraLogic.

The area requirement for the FCA block is also small, a key requirement in the ASIC arena. Dr. Togai estimated that adding an FCA core to a standard microcontroller would increase the die size by about 20 percent. In very high-volume ASIC applications, the addition of FCA could add as little as 10 cents to the component's selling price, Dr. Togai estimated.

VLSI Technology will offer the FCA standard cell in both 5-volt and 3-volt 1-micron CMOS versions. The core has also been designed for fabrication in a 0.8-micron process for even higher performance and lower power dissipation.

The companies are working to create fuzzy-based solutions at several levels, said Jeff Hendy, director of new product and business development for VLSI. Togai is working on a new version of its Fuzzy Programming Language compiler, expected to be available by September, which would allow the 32-bit microprocessor made by Advanced RISC Machines Ltd. (ARM) to utilize fuzzy algorithms. ARM is a joint venture of VLSI, Apple Computer and Acorn Computer.

VLSI will incorporate the FCA and the ARM microprocessor core into ASIC devices. "The ARM core is small and has low power dissipation, making it ideal to be implanted in an ASIC," Mr. Hendy said.

VLSI and Togai anticipate their products will be used in consumer electronics, automotive, telecommunications, computer speech and character recognition, and military/aerospace applications.

Togai offers a development system for fuzzy-based products. The tools run on IBM-compatible PCs and Sun Microsystems workstations, and are priced between $500 and $5,000 depending on the configuration.

A number of major microcontroller makers around the world have formed alliances with fuzzy logic specialty houses: Motorola with Aptronix Inc., San Jose, Calif.; South Korea's Samsung Electronics with NeuralLogix Inc., Sanford, Fla.; and Japan's NEC with Omron, Kyoto. Separate European fuzzy logic projects are under way at SGS-Thomson Microelectronics and Siemens.

COPYRIGHT 1992 Reed Business Information, Inc. (US)
COPYRIGHT 2008 Gale, Cengage Learning
 

BNET TalkbackShare your ideas and expertise on this topic

The following tags are supported in BNET comments:
<b></b> <i></i> <u></u> <pre></pre>

Leave a Reply

  1. You are currently a guest | Login?
Go
advertisement
  • Click Here
  • Click Here
advertisement

Content provided in partnership with http://findarticles.com/source//