How to Future-Proof A Hardware Designer

Just because skills become commoditized doesn’t mean they’re not a valuable foundation.

popularity

I’m at DAC this week, where there is a lot of interest and discussion on what’s going on in design and what’s going to happen to designers. One conversation with a university professor gave me a “eureka” moment.

The professor had a student who really loved RTL design. The student asked him where he could get a job doing this, and the professor suggested the student move to India.  While this is not entirely true, I believe the sentiment brings home a significant issue hardware designers are facing today. RTL design is becoming, or perhaps has already become, a commodity.

As a hardware engineer in a world where my profession is a commodity, how can I expect to survive?  I believe the secret to success—the secret to “future-proofing” your professional capabilities—is to understand that an engineer’s value comes from his ability to solve a problem. If RTL design is becoming less and less of a problem then we have to figure out where the hard part of design is today.

I believe the current challenges in design are in the area of defining the architecture. We need to understand what we are going to build and we need to characterize the performance and prove the efficiency our hardware will achieve before we build it.

By understanding the design at a more abstract level I can make tradeoffs at the system level, define performance characteristics, make tradeoffs between hardware and software and between implementation alternatives, and I can balance the cost, performance and power consumed by my system before implementing my design. The RTL design itself may be a commodity, but defining the hardware that should be built and making the architectural decisions that define what RTL should be built absolutely are not commoditized skills.

In my opinion, RTL hardware designers need to begin this transition to become architectural designers. Someone who does not understand hardware concepts is not going to be effective at making the tradeoffs required to create a successful system design. As a hardware designer, if I expand my capabilities to embrace an understanding of the system level, embrace the concept that hardware design does begin above RTL, then I will be able to leverage my hardware understanding into the future. I will be a valuable contributor to the emerging and challenging portion of the hardware design task by creating unique architectural solutions to meet systems cost, performance and power optimization problems. And I will “future-proof” my value as a hardware designer.

–Jon McDonald