Several weeks ago I was reading an article about Microchip Technology Inc. and a history of their growth and success in the embedded microcontroller market.  It was an interesting read for several reasons — first, because I’ve been using Microchip PIC devices in my work since the mid 1990s;  second,  their success in the market given the increasing diversity of their product line over the years is truly impressive, and lastly because I really enjoy designing with their devices — they are inexpensive, and the development tools are very easy to use and are well-documented.  At the bottom of the article in the comments section, someone posted a comment which went something like this:

Its so sad that people are still designing with PICs — we’ve got Raspberry Pi now!

Sigh.  Ten years ago my first reaction to a comment like this would be to fire up the pilot light on my cyber flamethrower and push the throttle to the “full open” position, however with age comes tolerance and understanding, so I decided to provide the well-intentioned, but woefully ignorant poster with a bit of constructive criticism.  I replied with:

Nobody is going to build one million pieces of ANYTHING with a Raspberry Pi…..

I have no axe to grind with Raspberry Pi — they are a very interesting set of products that have a tremendous following among techies, hobbyists, students, and others.  However, the implication that the mere existence of the Raspberry Pi should have Microchip Technology preparing their Chapter 11 filing brings to light two fundamental and deep misunderstandings about product development specifically, and more generally, the technology market.

The Raspberry Pi Foundation has tapped into a very interesting market — they have essentially taken the concept of a single board computer (which is nothing new, as SBCs have been around for two decades) and made it “cool”.  They deserve to be complemented for this, as they have brought an extremely versatile product to the masses at a very low price.  These tools make it extremely easy to build very sophisticated systems relatively quickly by individuals that possess a minimum of electrical engineering design talent.  A cursory web search for interesting Raspberry Pi projects brings up all sorts of things that are incredibly nifty — one individual actually launches Raspberry Pi’s into the upper atmosphere using balloons, and takes photos from the edge of space!  There is no question that the Raspberry Pi has revolutionized the junior high and senior high school science fair project offerings for many years to come — I wish I had access to one of these gizmos when I was in junior high school, as I could have used it to control the spacing between the electrodes of my carbon arc lamp (I won a blue ribbon in the school science fair for that project).

But, as the title of this blog entry states — it is easy to build one unit of anything, and “A prototype does not a product make.”  The Raspberry Pi would be an excellent low-cost platform to use for the development of a product proof-of-concept for a device which needed all of the functionality offered by a high-end ARM processor.  It would make no sense at all to actually develop a standalone product and stuff a Raspberry Pi inside the plastic — the cost associated with all the bells and whistles on the PCB would destroy the economic feasibility of the product.  That’s where the commenter’s train of thought went off the rails — having a really nifty computational platform that fits in the palm of your hand and contains more computing power than the command module on Apollo 11 does not necessarily make it applicable to be used in a product.  So much of what we do as product development engineers has nothing to do with hardware or firmware — the most important part of our job is to provide the simplest and most economical solution to the design problems that we are given to solve.  This requires us to survey the myriad of computational platform solutions available to us, and balance these against the multidimensional labyrinth of functionality, cost and price (the two are not the same!), manufacturability, development cost, testability, field support, scalability, etc.  It is in these aspects of the design process where the magic really lies — making complexity seem very simple, while eliminating as much superfluous “fluff” as possible.

The commenter also made me think about the current state of the technology market and how easy it is to lose sight of the entire forest because you are fixating on one particular tree.   While pretty much everyone around us walks around with a mobile device in their pocket, has at least one video game console at home, and goes about their errands while wearing a fitness tracker, the reality is that these systems and devices constitute only a small fraction of the entire embedded system market.  Much like an iceberg, the entirety of the embedded systems market is mostly invisible — comprised of billions (and I do mean BILLIONS) of very simple, very tiny bits of hardware and firmware that crank away constantly in devices all around us.  You’ve certainly got quite a few of these around you right now — the oven in your kitchen, your dishwasher, coffee maker, refrigerator, washing machine, thermostat, water heater, irrigation system, garage door opener — I could keep going, but you get the picture.  If you look a bit further afield there are even more — take a walk around your office, or the production floor of any factory, and you’ll find hundreds of places where some humble 4-bit or 8-bit microcontroller is whirring away silently, keeping its own little corner of the universe running smoothly.  You won’t find any Raspberry Pi’s here, but I’ll bet you dollars-to-doughnuts that you’ll find more than a few Microchip PIC processors.

It was about five years ago that Jack Malone*, my local Microchip Technology sales representative, told me that Microchip was shipping over one million of their low-end PIC16 microcontrollers each day.  You heard that right —- over one million microcontrollers each day.  Obviously, all that silicon is going somewhere — it is destined for that part of the embedded systems world that is “below the waterline”.

I don’t think Microchip has too much to worry about — their market space is alive and well.

Enjoy!

 

David

 

*For those of you in Southern California who knew Jack, he needs no further explanation.  He was a great guy who would go to the ends of the earth to help a customer in need.  In addition, he was a long-time veteran of Microchip who believed in his company’s products.  His passing several years ago was a sad day for many of us in the San Diego engineering community.