Case Study: Getting More Functionality from Existing Chips

Submitted by BDTI on Wed, 05/18/2011 - 21:00

These days, few system design teams can afford the expense and time of developing a custom chip tailored to their specific needs.  Therefore, most system designers are forced to rely on off-the-shelf chips that are a less-than-perfect fit for their needs.  And even teams that are able to design their own chips must minimize the frequency of doing so, given the costs of creating a new chip.

For these reasons, embedded system designers often find themselves needing to shoehorn new functionality into old chips—whether those chips are of their own design, or designed by a chip supplier.

In some cases, diligent optimization of existing or new software can be used to fit new functionality into an existing design.  In other cases, a chip is already so heavily loaded that adding functionality requires a more creative approach.

Recently, a manufacturer of ultra-low-power audio devices approached BDTI with such a dilemma.  The company’s existing product supported limited wireless connectivity, intended to be used to configure the device.  Now, in order to support a valuable new feature, the company needed a much higher performance wireless capability that could be used continuously.  The company’s next-generation ASIC wouldn’t be available for a couple of years, so engineers needed to find a way to implement the new functionality using the existing chips.

The manufacturer’s design team had come up with an initial approach for implementing the new functionality within the constraints of the existing chip, but didn’t have the bandwidth to fully evaluate their proposal—they were busy implementing a variety of other new features.  So, the company engaged BDTI to bring a fresh perspective and additional engineering resources to bear.

BDTI’s engineers began by working with the manufacturer’s design team to clarify their requirements.  Then BDTI studied the existing chip and system design—in enough detail to enable addressing the issue at hand, but without getting bogged down by reams of irrelevant information.  Armed with this knowledge, BDTI was able to quickly perform a hand analysis of the client’s initial solution, followed by high-level simulations.  The analysis and simulations showed that the initial solution, while reasonable at first glance, wouldn’t work.

Fortunately, BDTI was aware of a technique that had been used in wired networking to solve a related problem.  BDTI proposed adapting that approach to enable the existing chip to meet the client’s new wireless connectivity requirements with minimal demands on the existing chip.  The client’s engineers found this unconventional approach intriguing, but were uncertain whether it would work in practice.  So, BDTI performed a second series of simulations—in this case demonstrating that the new technique would in fact yield the needed performance.

The net result was that in a few weeks, the client was able to identify a viable solution for implementing the new wireless functionality on its existing chip, enabling it to bring a valuable new product feature to market years earlier than would have been possible if a new chip had been required.  And, the client was able to obtain this solution without diverting its own design team from other critical efforts.

Sometimes, a key added feature can make the difference between an also-ran product and a home run.  BDTI enables design teams to address such opportunities without diverting resources from other priorities.  Find out how BDTI can help you find the best ways to get more functionality from your existing chips:  contact Jeremy Giddings at +1 (925) 954 1411 or giddings@BDTI.com

Add new comment

Log in to post comments