Main image of article ‘Tech Connects’ Video, Transcript: Tony Chan Carusone on Semiconductors

The semiconductor industry has drawn a lot of attention, thanks in large part to President Biden’s CHIPS act, which is designed to bring a good deal of chip manufacturing back to the United States. Tech pros are increasingly interested in semiconductors as a potential source of new job opportunities, and they’re curious about the skills they need to break into this market.

On the Tech Connects podcast, we recently interviewed Tony Chan Carusone, who’s a Professor at the University of Toronto and CTO of Alphawave Semi, which gives him some interesting perspective into the current evolution of the semiconductor market. The audio version of the episode is available on Dice Career Advice, along with podcast channels such as Acast and Spotify. Here’s the video:

We’ve also included some transcript excerpts. Follow along as Carusone discusses how there’s no better time to get into the semiconductor industry, and the skills you’ll need to help build the world’s next generation of chips.

Last August, we had the CHIPS Act, which is designed to stimulate the semiconductor industry in the U.S. with lots of investment. There's also a potential for the semiconductor industry to be short a million employees by the end of the decade, which is an incredible shortfall. It's such a critical industry and a critical technology, so what's behind the shortage and the talent pipeline?

I've been thinking about this problem for a long time because, before I was ETO at Alphawave, I have been a professor at the University Of Toronto, teaching integrated circuit design semiconductors for over twenty years. At Alphawave, we see the tremendous push for new semiconductor technology, connectivity technology, specifically to move data around in datacenters. That bottlenecks A.I., which is something that is on the tip of everyone's tongue these days.

Semiconductors have been in people's hands and pockets and on their laps for a long time now, and yet we've always faced this problem that somehow it's really hard to make people understand the way in which so much of the technology that they care about and love rides on semiconductor technology. It's somehow abstract and invisible to them in a way that software is not, I think. Young people get exposed to software very early: it's easy to play with, get stuff up and running. It's kind of ironic in a way, because the truth is that software is actually more abstract, whereas hardware is actually physically real, and yet people feel like they can touch and play with software, I think, more than they feel that with semiconductor technology.

So, I think that's one of the challenges. Everything you care about these days in terms of technology probably rides in semiconductors. 

With somebody who's potentially interested in getting into semiconductors, what do they need to learn? What are the top skills, and does it vary considerably between specializations? Where do you begin if you’re a student interested in it?

I would say that there's so many different applications of semiconductors because it does touch so many different technology areas. It depends on the application area you want to combine, and an understanding of semiconductor technology and circuits, along with maybe another application area expertise in an application area for which you have some personal passion.

So, if you're interested in biomedical applications and electronics, you might couple that with an understanding of semiconductors and the associated circuits. It's always been connectivity and high-speed communication. It motivates me to see these new technologies come online, and what the miracle it is that we can have a video chat like this or on your phone even while I'm riding the train. That just seems magical to me, so that's always captured my passion.

If you’re coupling [your focus], it’s all the integrated circuit design. The microchip design is where the rubber hits the road, and then you've got stacked on top of that mathematics and signal processing. It's really applied math that's riding on top of that. 

A great thing about working in this area, it's a kind of platform on which all these other things ride. If it's artificial intelligence, there's on-processor technology that's implemented in semiconductors. I think that would be my advice to a young person, and when I speak to my own kids—they're getting the age where they're starting to pick future direction for themselves—I can't help but encourage them to follow their interests. But at the same time, there's these kind of table stakes, understandings that you need to have along with software, which I think now people generally accept some comfort with software as table stakes for working in a lot of different fields. I think we should get to a place where some understanding of hardware and electronics is also table stakes for doing a lot of the work that you think you want to do in the future. 

When you look at the data produced by the U.S. Bureau of Labor Statistics (BLS) or CompTIA, you see this incredible demand across the entire tech spectrum for all kinds of skills, whether it's people who can build machine learning models or program in whatever language or what have you, and a lot of companies are reacting to that considerable demand by loosening their previous requirements for jobs. Does the same thing apply for people who want to get into the semiconductor industry—as in, can you master a particular skillset and not have formal education to break into it? Or is it an industry that demands people be properly credentialed with the right certifications and degrees?

The trend that's emerging now is the fact that so much software work can be done with open-source tools, openly available tools. That has traditionally not been the case when it comes to microchip design. Interestingly, there is now a lot of movement on that front to make available open-source tools for microchip design to enable younger people not inside a big institution with a lot less money to be able to design a chip and actually get it fabricated and get it back.

There are programs to allow people to code a chip, right? And just to essentially describe a whole chip with a similar kind of language as you would a software program. I think that's happening for a number of different reasons. One of them is to open up the doors of innovation, product innovation, to more people; and it's also to pitch a bigger tent to include people with different kind of interests and different kind of skills, right? And to allow them to participate in in this exciting area. That’s something that's traditionally been a barrier to allowing people with different backgrounds into this field, but that is changing over time as these tools become more common.