中文 English

Three Technologies Enabling The Next Decade Of Hyperconnectivity

Some of the last decade’s big trends are still in full swing today.

popularity

As it has become a tradition in my 15 years of blogging, January is a month of both reflection and outlook. At the beginning of 2022, I am excited that key themes from 5 and 10 years ago—3D integration, artificial intelligence and machine learning (AI/ML), and ubiquitous needs for more connectivity driving 4G and 5G networks—clearly have exceeded expectations and forecasts from that time. Looking forward to the coming decade, some of these themes will even accelerate and are bound to lead us to new exciting highs.

It’s fascinating to me that in the IEEE Spectrum January 2012 issue, the list of technologies to watch in 2012 included 3D chips as a means to extend Moore’s Law with articles like “3-D Chips Grow Up”. Five years later, in 2017, “Intel Finds Moore’s Law’s Next Step at 10 Nanometers” similarly discussed the future of Moore’s Law. This year, the industry is talking about “More than Moore.” Clearly, what we used to know as systems on chip (SoCs) are disaggregating for more extensive and smaller designs. Complexities are often too close to reticle limits, and heterogeneous integration via chiplets allows faster turnaround times at all complexity levels. While the focus 10 years ago was on 3D-IC integration memory chips, we are today talking about co-design of chiplets. And it seems that memories are off to further systemic integration, so we find an article about in-memory processing accelerating AI in the IEEE Spectrum outlook at the end of December 2021.

AI/ML was the second big predicted trend five years ago. It has become a reality, both driving growth in the semiconductor market and increasing developer productivity. Five years ago, the IEEE Spectrum outlook conditioned us to Expect Deeper and Cheaper Machine Learning. And in 2022, the game is on for AI processing, meeting different performance, power, and latency requirements at numerous locations from sensors through far, middle, and near edges. The same is true for AI/ML increasing developer productivity, becoming more pervasive throughout the top-to-bottom design flows, extending beyond the high-impact verification and digital implementation areas in the news already. For example, we have already reported significant research in applying AI/ML to PCB design and system design. AI/ML is already pervasive. There is lots more to come.

The third trend—connectivity—was a common theme in the 2012 and 2017 issues of IEEE Spectrum. “Fantastic 4G” looked at investments in 4G LTE networks, and “Here Comes 5G—Whatever That Is” talked about 5G market aspects for a 2020 introduction. Today, while the 5G future has rung in a new era of ubiquitous hyperconnectivity in some areas, it is still very unevenly distributed. It has just touched the tip of its opportunities to date, and we are already talking about 6G targeting a rollout before this decade is over.

Looking forward, what’s in store for our hyperconnected, AI-enabled future, enabled by beyond-Moore semiconductors and systems?

We have arrived at the future of more autonomy in automotive, pretty much as predicted. Safety is critical, so “Brakes that Slam Themselves” is a timely part of this year’s IEEE Spectrum outlook feature discussing how some parts of the world mandate automatic emergency braking. This year’s latest news talks about making a “Star Trek”-style universal translator a reality in the “The AI Research SuperCluster could help the company develop real-time voice translations”.

Some transformations will be faster than others. Just as IEEE Spectrum’s “Augmented Reality: Forget the Glasses” talked five years ago about how reality prevailed, this year’s “What Can the Metaverse Learn From Second Life?” puts a cautionary note on the speed of technology adoption. In my mind, augmenting the “real” reality is a great steppingstone. I remember fondly when my daughter defended the Pokémon Go Gym of the Eiffel Tower while exploring Paris 5 years ago. Next step: Augmented Reality Glasses.

Some caution is warranted, too. It seems our responsibility as engineers is probably more significant than ever. Top of mind to me are:

The bottom line is that I am excited about our future! I am sure we will get there. Let’s not forget that Gen Z is now the dominant age group with about 32% of the global population. We are talking about the people born between 1996 and 2010 that are out to change the world. When talking to 3,000 consumers not too long ago about trends in hyperconnectivity, we found Gen Z to be “progressive skeptics.” They are generally more optimistic than Baby Boomers but less optimistic than Gen X and Y about emerging technologies’ impact on our lives. We speculated that the growth rates of cyberbullying, youth anxiety, and depression have forced Gen Z to reevaluate its relationship with technology. They will keep us honest. And next up, Gen Alpha (under 12 this year) is already standing by to keep Gen Z in check.

On a personal note, next month’s blog will be my 125th on Semiconductor Engineering, and I am in my 11th year on this platform. So, a huge thank you to Semiconductor Engineering. After reviewing some of the content I wrote, I have made a “new decade resolution regarding blog titles.” “Not All Software Is Like Elvis?” “The Art of Double-Indirect Sales and Product Marketing?”. Really, Frank? I herewith solemnly promise to title my blog posts well enough to understand them myself 10 years from now. Wiser dog, new tricks. Please keep me honest. Or let AI do it.

Cheers to a great 2022!



Leave a Reply


(Note: This name will be displayed publicly)