Attendance is up in seminars about low power techniques. More questions are being asked about power islands and multiple voltages. And in the verification world, engineers everywhere are rolling up their sleeves.

What’s happened is that power-saving techniques are now becoming mainstream. As more mainstream companies move to more advanced processes they are wrestling with leakage from both active and static current. What used to be the headache for companies like Qualcomm, Intel, Freescale and Broadcom is now a collective pain across the semiconductor industry.

This is all good news. It provides more input on solving complex problems, a ready market for more advanced tools, and jobs for people who understand these issues. It also makes these techniques available across a broad swath of companies, which has a cumulative effect on conserving power in devices ranging from consumer electronics to white goods all the way up to cars.

In addition, transitions always create new opportunities—for both people with the necessary skill sets and the startups that can create tools to fill in the gap. That tends to keep the industry vibrant and creative, and it tends to make what were previously difficult engineering tasks more common with better tools to solve the problems.

The downside is this will require some massive re-tooling of the engineering workforce, which costs money and takes time. Just because the problems are mainstream doesn’t mean everyone will catch up quickly.

–Ed Sperling

Designing and building semiconductors represents the very best of science and mathematics: physics, geometry, thermal mechanics, Boolean logic, complex mathematical modeling, not to mention a good deal of finger-crossing at tapeout.

While much of what has been accomplished is nothing short of amazing, particularly given the time frame in which some of these advances were made, the hardest work is still ahead. The next step isn’t so much creating new dielectrics or layout techniques—although that work will have to continue for Moore’s Law to survive. It’s taking advantage of what’s already been developed in many places and integrating it together into even more complex devices.

This is easier said than done. Just because technologies can co-exist on separate chips or even within the same package doesn’t mean they can be loaded onto the same piece of silicon without creating significant new problems. Heat, which now comes from both static and active power leakage, is one issue that has to be dealt with. Signal integrity is another. Overall reliability is a third. Flexibility, a more recent phenomenon driven by changing profitability requirements, is a fourth.

It’s this flexibility that is raising the most red flags recently. Profitable designs of the future need to be able to span multiple uses and multiple derivatives in each segment in which they are used—different vertical markets, applications that span multiple markets and in various configurations for the same market. And they have to do this without impacting performance or raising the power budget.

As an industry, we are used to creating chips linearly. We have flows, models and business models to match—a monolithic approach to a multi-dimensional problem. The next step will be to re-invent the method while still keeping the rest of the pieces working properly—a combination of the old and the new, and a set of skills to match. The old stuff won’t certainly go away, but the growing opportunity is in integrating it in new ways.

—Ed Sperling

Concern for power is moving from “nice to have” to “must have” in all designs—not just semiconductors. What’s interesting is that, at least from a utility standpoint, not all designs require it.

Over the next decade, almost everything in our world will be re-engineered with an emphasis on lowering the power requirements. Low-power sells, even if consumers can’t recoup the benefits for years. The entire solar industry is a case in point. For homeowners in particular, insulation has a much quicker return on investment. Nevertheless, solar installations are booming.

The same goes for household devices that normally are plugged into the wall. The amount of current these devices draw normally doesn’t even register with consumers, most of whom never turn off their televisions or computers at night because standby power is considered good enough. But they’re willing to plunk down money on devices that can communicate with each other wirelessly to turn down the heat or turn off lights remotely.

All of this is good for the environment, of course. It ultimately has an effect on energy bills, and may sidestep the need for building new power plants. It may even ultimately reduce the amount of oil being consumed in some regions, although that is likely to be offset by consumption in developing economies such as India and China. But perhaps even more, it represents a new mindset that emphasizes saving power at every turn.

Chips that can add an extra hour or two of battery life to devices will give way to chips that can add an extra hour or two of life to power-intensive applications. And applications that can intelligently use power will win over those that suck it down like power was still a penny a kilowatt/hour. (The price ranges between 5 cents and 29 cents in the United States, according to the U.S. Department of Energy.)

All of this is very good news for semiconductor companies. There is huge upside in lowering the power. But it also means that chip design will only get more complicated, with verification and software development turning into incredibly challenging opportunities. Getting to tapeout first is still a huge opportunity, but getting to tapeout with a much lower power envelope may be equally important over the long run—and possibly even more lucrative.

—Ed Sperling

The problems now coming to light in Toyota’s Prius line are a grim reminder of just how far things can go wrong in low-power engineering.

For years, the Prius has been a shining example of how to reduce power consumption in an highly integrated array of complex systems, each geared for maximum performance while also reducing power consumption. But developing a low-power system is challenging enough. When you connect multiple systems together, each with a focus on low power, it becomes even more challenging.

So far, validation of concept and verification of an end product has been confined to individual blocks. When software is added in, the problem moves from block to subsystem. And when networking and power management of multiple systems is included, it becomes an exponentially larger challenge with an enormous number of hidden corner cases based upon state, unexpected interactions, breakdowns caused by harsh conditions—potholes or excessive heat, for example—and potential shutdowns.

All of this is tolerable in a consumer device like a smart phone. Users can live with dropped calls or a glitch in a handheld game. They simply call back or reboot. It’s not as important to solve the verification problems in these devices because no one’s life depends on them.

It’s a lot different when it comes to medical or industrial devices, but even those are generally limited-function devices. A pacemaker may be critical, but it also has one purpose. The same is true of a fire alarm. Still, in the future more of these devices will be tied together and increasingly interdependent, which means reliability testing will have to be done in all states and with all possible permutations being considered.

This is brain-bending stuff in the simplest of configurations, where it can be done on a spreadsheet. It’s the work of server farms in complex systems, based upon models that have to be developed at the architectural level. And it’s the kind of problem that has to be tested continually throughout the development cycle to ensure reliability and to incorporate any changes.

All of this adds cost, of course. And it requires more engineers with a broader focus. Once you get beyond that hurdle, it requires an interplay between groups that generally don’t speak the same language and which have, at least so far, had little in common. But like complex systems, all the pieces must work together. That starts with the people involved in making the systems.

Toyota had the right idea in developing the Prius. Unfortunately it didn’t have all the right tools or procedures for testing all of the possible things that could go wrong. It’s no longer just a chip or a software problem. It’s a much bigger engineering issue, and low power is now a central part of that problem.

–Ed Sperling