How Will the Death of Moore's Law Affect Engineers?
Borrowing a line from Mark Twain, the death of Moore's Law has been greatly exaggerated. Many times over the years. But have the death prognosticators gotten it right this time? And, if so, what will it mean for engineers?
In case you are one of the few technology industry citizens who is unfamiliar with Moore's Law, it is the observation that (in general terms) computing power doubles every two years.
As the fabrication processes to produce new chips have approached the atomic level, it has seemed pretty clear to many, including me, that Moore's Law cannot continue to hold. A recent article ("Moore’s law really is dead this time") asserts we are just about there.
These difficulties mean that the Moore's law-driven roadmap is now at an end. ITRS decided in 2014 that its next roadmap would no longer be beholden to Moore's "law," and Nature writes that the next ITRS roadmap, published next month, will instead take a different approach.
Engineers have benefited tremendously from Moore's Law. Let's think about what it means if computing power growth does slow down dramatically.
Engineering software is a different breed. It is computational and depends on numerical processing speed. It is graphical and depends on video processing speed. Reductions in the growth of new bandwidth for engineering tools has to have some kind of impact on engineers.
It seems to me there are two conflicting paths that will emerge. One path is the path of innovation. Software tool developers will have to get more innovative, more creative, and more focused on maximizing what they do. They cannot just keep slapping new features onto their software tools. They will have to take what they have developed and to re-think it to make it fit better into a fixed size box.
Engineers who use these tools will also have to also get more innovative and creative. The capability of their tools will not grow as quickly. They cannot count on this increased capability to solve ever more complex problems. They instead will have to learn the tools that exist better than ever before and learn how to apply them in more innovative and creative ways. Which means squeezing every ounce of information from the tools that they can.
The other path is the path of status quo and stagnation. New, innovative software tool development companies will not be able to capitalize on some new computing capability and introduce some innovation that helps them overtake an established competitor. This will make existing tool developers safer and feel less push from competition. They may become satisfied with their position and not feel the need to innovate to maintain their competitive edge.
Engineers who use these stagnant tools will also enjoy greater protection from the competition. No longer will the established engineer feel pushed to improve him or herself in order to stay ahead of the younger generation. The younger generation of engineers will feel less excitement from the technologies they learn in school. They will not be learning cutting edge tools - just ones that feel old and established.
Which path is more likely to emerge? When I ask myself this question I am taken back to a time before computers or even calculators even existed. When engineers used slide rules to perform their calculations. Did engineering stagnate then?
Well, I do not know for certain because I was not there. But I am old enough to have worked with and learned from engineers who were there. Such engineers built the Hoover Dam (one of my engineering professors did this) and developed rockets that took mankind into space (I worked with some of these "old timers" in my first job).
During this time of relatively slow growth in engineering tools engineers got very creative. They developed nomographs and specialized charts to overcome the limitations of their computational tools. They developed creative ways to solve equations in approximate form (think boundary layer theory).
Of the two paths I have outlined - one the path of innovation and the other the path of stagnation - my money is on the path of innovation. Engineers will figure out how to make good progress with the tools that they have. And software tool developers will doubly focus on how to rewrite their software to make it more efficient.
And somewhere along the way I suspect there will be a breakthrough in processor technology that allows Moore's Law to become relevant again. Whether it is 3-D computer chips, quantum computing, optical computing or something else we have not discovered yet. My money is on engineers and their innovation.
Great article, Trey. I think any slowing of hardware advances, will only provide us a chance to catch up with the computing power that is already available! Moore's law will still apply to the increases in innovation and application of technology.
Intriguing !! I am one of the guys who started with a slide rule. I purchased my first HP Engineering Calculator while in college, had to have it to keep up with fellow students in test sessions. I am still working and find myself missing the cadre of engineers who knew so much about everything that was happening in the chemical plants that I served. The young guys are much more calculation and statistical methods oriented...and seem to depend on what they are able to Google in response to practical issues that are fundamental to keeping a plant operating safely and reliably.
I hope I get to see the results of the future prospects that Trey addressed, it is wonderfully fun to observe how we are as engineers and as people who respond to all the influences of the digital lives that we now have opportunity to live.