Technology Tsunami Approaches

How many times have we heard the saying that technology advancements are accelerating and that inevitably the older generation will have increasing problems keeping up with the new advancements?

This happened to me with software development methodologies over fifteen years ago. I still program, when people actually let me, using basically the same techniques I learned when I was in my teens. While the languages that I used changed and my skills improved, they all used the same underlying principles – that of procedural programming. I was not able to make the jump to the new object-oriented paradigm and programming methodology. The younger generation could do it much better than I could.

When I look on the technology horizon, I see more change barreling towards us than I have at any time in my career and I believe that many waves are about to converge and hit us in one enormous technology tsunami. I expect many people will be washed to the sides and not just users but hardware and software developers who are unable to handle the new realities. Engineers, like myself, who cannot switch quickly enough to a new set of realities. Many will resist the changes and hold companies back, but hopefully they will all get moved into management and allow progress to get made before too much damage is done.

Some of the reasons for this tsunami are almost a cliché, others a lot more subtle and some are due to pent up need for change, but reluctance to allow such change because the costs will be enormous. But they will change out of necessity because of the growing costs associated with not making the change.

Just consider a few factors leading up to this tsunami. There are the semiconductor production issues. I am not going to call it a slowing of Moore’s law, instead the reaching of physical limits along one path that are forcing the industry to explore alternative paths. Just one of them could cause significant change in the way systems are constructed, but the industry is not just looking at one change, they are putting investments into many. Examples include 3D integration, carbon nanotubes and optical interconnect and processing.

has been stagnant for decades, except for simple scaling, but new memories are on the horizon that will do much more than just replace the existing memories. They will cause the industry to re-examine memory architectures, the way that information is moved around a system, the architecture of the system themselves and basically upset the whole notion of computing. We cannot afford to make the entire memory space coherent across multiple processors when it is only required for a few exceptional pieces of code.

The computing paradigm has been in place for 50 years without any major change, but the time has come for it to move away from Von Neumann architecture. This is limiting advancement, has difficulty scaling to heterogeneous computing architectures, consumes too much power and is basically too slow. Alternatives are known but resisted because of the huge software base that is currently installed and the cost of rewriting would bankrupt many companies.

The , in whatever way you choose to think about it, is a game changer because it adds a new layer of architecture into the world, one that has never existed in the past. Information will be flowing in a global manner for 100s of millions of sensors, the communications infrastructure that allows that raw data to be transmitted in real time, the processing systems that can turn that data into information and knowledge, and the ways in which that knowledge can be used. Each “industrial revolution” that has happened historically has been when a new form of communication emerges and enables mobility that did not exist previously. It happened with trains, it happened with the Internet, and I believe the IoT is on the same scale as those advancement.

I am about to start writing a series of articles about these changes and the impacts that they could have on the industry. What do you see that will be completely different in ten or fifteen years? I would love to talk to you about your views and get them incorporated into those articles. How will these changes affect the design and verification of systems? What other impacts will they have? Please send me an email and get the conversation started. You can reach me at [email protected].