2028: The end user is always right - Embedded.com

2028: The end user is always right

November 2028 is the 40th anniversary of ESD. Click here read other 2028 lookbacks.

Just last week we celebrated the shipment of our 100 billionth PIC Microcontroller since the inception of Microchip Technology almost 40 years ago! This milestone, and the 10 billion PIC microcontrollers we shipped in the last 12 months, gives me pause to reflect on what an incredible ride the last 20 years have been. While many continue to talk about how the pace of change can't possibly continue at the current rate, the last 20 years have borne witness to exactly the opposite.

So, with my imminent retirement almost upon us, please indulge me as I wax eloquent about the last 20 years and reflect on the broad themes of change and innovation that occurred over these glorious years.

The market

  • The single largest growth driver has been the expansion and pervasiveness of green technology and solutions to enable energy efficiency. In retrospect, what started as a niche for embedded applications 20 years ago turned into a tremendous growth engine that changed the way microcontrollers, digital signal controllers, and advanced analog products were used to control motors, power supplies, lighting, solar and wind energy conversion, batteries for electric and hybrid cars, utility meters, and a host of other applications so as to make them all significantly more energy efficient.
  • Another major market driver has been the explosive growth and innovation of human-interface solutions in embedded systems. We take various forms of touch sensing and rich display technologies (even in harsh environments) for granted today, but these were embryonic areas 20 years ago that rode the engine of silicon, software, and material innovation.
  • A final area of explosive growth has been intelligent sensors. The ability to pack very-high-performance microcontrollers that consume ultra-low power into very small packages, the rapid innovation in microelectromechanical systems-based sensors, and the emergence of embedded networking standards (especially wireless networking standards) has enabled very-low-cost sensor networks to be ubiquitously deployed in a broad range of distributed but connected applications.

The technology

  • Silicon and system design have become much more about software in one way shape or form, as the hardware design has been largely abstracted out by design tools that automatically configure and optimize the hardware. Application software itself has evolved to where it looks and reads like normal English, with software tools that handle all the complexity underneath this high-level abstraction in which most engineers work. This simplicity, in turn, has greatly increased software reliability and reduced the documentation workload.
  • Just about all microcontrollers now employ multiple cores to deliver the level of performance required at a level of power consumption and thermal dissipation that is acceptable for embedded systems. The software development challenge to make multicore microcontrollers was extremely painful, but after several false starts, the industry was finally able to make this work quite effectively.
  • Finally, wireless networks for embedded systems made tremendous progress from substantially fragmented and proprietary solutions 20 years ago, to standards-based solutions that provide user flexibility and configurability. The breakthrough came with the advent of software-defined radios that enabled flexible, low-cost, and ultra-low-power radio solutions.

The market

  • The single largest growth driver has been the expansion and pervasiveness of green technology and solutions to enable energy efficiency. In retrospect, what started as a niche for embedded applications 20 years ago turned into a tremendous growth engine that changed the way microcontrollers, digital signal controllers, and advanced analog products were used to control motors, power supplies, lighting, solar and wind energy conversion, batteries for electric and hybrid cars, utility meters, and a host of other applications so as to make them all significantly more energy efficient.
  • Another major market driver has been the explosive growth and innovation of human-interface solutions in embedded systems. We take various forms of touch sensing and rich display technologies (even in harsh environments) for granted today, but these were embryonic areas 20 years ago that rode the engine of silicon, software, and material innovation.
  • A final area of explosive growth has been intelligent sensors. The ability to pack very-high-performance microcontrollers that consume ultra-low power into very small packages, the rapid innovation in microelectromechanical systems-based sensors, and the emergence of embedded networking standards (especially wireless networking standards) has enabled very-low-cost sensor networks to be ubiquitously deployed in a broad range of distributed but connected applications.

The technology

  • Silicon and system design have become much more about software in one way shape or form, as the hardware design has been largely abstracted out by design tools that automatically configure and optimize the hardware. Application software itself has evolved to where it looks and reads like normal English, with software tools that handle all the complexity underneath this high-level abstraction in which most engineers work. This simplicity, in turn, has greatly increased software reliability and reduced the documentation workload.
  • Just about all microcontrollers now employ multiple cores to deliver the level of performance required at a level of power consumption and thermal dissipation that is acceptable for embedded systems. The software development challenge to make multicore microcontrollers was extremely painful, but after several false starts, the industry was finally able to make this work quite effectively.
  • Finally, wireless networks for embedded systems made tremendous progress from substantially fragmented and proprietary solutions 20 years ago, to standards-based solutions that provide user flexibility and configurability. The breakthrough came with the advent of software-defined radios that enabled flexible, low-cost, and ultra-low-power radio solutions.

Product development

  • The way we conduct business globally has also changed dramatically. Teams have become highly decentralized and scattered across the globe. The communication tools to enable true virtual teams across the globe have evolved to the point where no one even gives this a second thought any more. This, in turn, has enabled companies to readily access a global and flexible pool of talent without having to relocate people from, or start operations in, the countries where talent is available.
  • Moore's Law came to an end many years ago–not because the engineers couldn't continue to find ways to shrink transistors, but because the economics became prohibitive. Yet, the rate of innovation didn't slow down as engineers found new ways to design value-added solutions combining multiple cores, highly complex mixed-signal functions, and innovative memory technologies.

Of course, the real proof is in how end-user systems and results have changed. Technology for the sake of technology no longer supports itself. Users expect systems that improve and simplify their lives, or they will not spend their hard-earned money. Having replaced all the mechanical water valves in my house with programmable, touchscreen systems means we select the desired water temperature, eliminating any burning hazard. Long gone are the traffic jams caused by human driving errors. The new smart automobiles, coupled with intelligent highways, have all but eliminated driver errors. Now, roadways operate at optimal capacity at all times, adjusting to the number and size of automobiles, weather conditions, and final destinations. Even traffic accidents have been greatly reduced, via computer-controlled driving, which has substantially lowered insurance costs. Probably the biggest success is that we finally conquered the air-traffic-control system problems. Having the system totally adaptive to traffic demand, wind and weather conditions, and aircraft type allows for optimal control of airspace. Real-time capacity planning, using all of the built-in system environmental monitors, has paid for itself many times over.

In closing, while the details of how things used to be back in 2008 are hazy at best, very little of where we are and what we do today was envisioned with any clarity back then. Change and innovation, over the long run, are rarely linear processes. So, the constant that we can expect for the next 20 years is the continued relentless pace of change and innovation, and that human ingenuity will put this innovation to good use to improve the quality of our lives. If the pace of innovation over the next 20 years is anything like the last 20 years, we can safely conclude that our crystal balls remain as they have always been–as clear as mud! Onward!

Ganesh Moorthy is the executive vice president of Microchip Technology Incorporated. He has 19 years of experience with Intel after which he become the CEO of Cybercilium, a small Tempe-based software start-up company. He joined Microchip as vice president in 2001.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.