For this month’s career feature, we interviewed Henry Duwe who co-authored “Bespoke Processors for Applications with Ultra-Low Area and Power Constraints,” which appears in the May/June 2018 issue of IEEE Micro. Duwe is an assistant professor at Iowa State University. His research interests include computer architecture and the design and architecture of ultra-low-power computer systems suitable for Internet of Things applications. Henry’s research has been recognized by two Best Paper in Session Awards at SRC TECHCON (2014 and 2015), a Best of SELSE award (2016), and a Best of IEEE Computer Architecture Letters (2013). His other honors include the Mavis Future Faculty Fellowship and the E.A. Reid Fellowship Award. We asked Duwe about careers in microprocessor technology.
Computing Now: What types of tech advances in the field of microprocessor technology will see the most growth in the next several years?
Duwe: Given the current state of microprocessor technology and our society’s current appetite for pervasive technological solutions, I see the likelihood for a large growth in two areas—specialized hardware and secure hardware.
Specialized hardware design has been around for a long time in the microprocessor industry (e.g., DSP processors and original GPUs). However, the slow-down or ending of CMOS scaling with no clear next technology and the increasingly diverse set of applications that demand efficiency lead to a need for specialized architectures. We’ve already started to see this begin commercially with architectures such as Google’s Tensor Processing Unit (TPU) that targets accelerating neural network inference. To support an increase in the design of specialized hardware, design tools will also have to improve in order to support the rapid development of new hardware designs.
A lot of work and attention has been invested in software security over the last several decades. However, fundamentally insecure hardware designs (see Meltdown and Spectre vulnerabilities) have been able to hide underneath the covers of insecure software. As software security is being better understood and improved, I see that there will be an increased growth in providing secure hardware constructs and in verifying the security of microprocessor designs. Additionally, as more facets of our lives have microprocessors integrated into them and the negative effects of security vulnerabilities become more real, security will become a stronger differentiator between microprocessor products.
Computing Now: What advice would you give college students to give them an advantage over the competition and why?
Duwe: During your college years, learn how to learn. As one example, those interested in microprocessor design will undoubtedly encounter numerous tools required to do their work. These will range from programming and hardware description languages to logical simulation, synthesis, and place and route tools. While one needs to have learned some set of tools to accomplish design tasks, it isn’t critical to have mastered tool X or tool Y, but rather to understand how to learn a new tool when required—either because your organization uses that tool or because it is the best tool for the job you need to accomplish.
As you go through college, keep track of your accomplishments and then plan to put your time/energy into building a select set of accomplishments. By accomplishments, I mean technical or leadership projects completed (either course-, independent-, or work/internship-related) rather than simply courses completed or tools learned. By focusing on completing tangible projects during college, you demonstrate that you are someone who can get things done with some degree of independence. If you don’t keep track of your accomplishments while they are happening you are likely to forget about them when it comes time for interviews and so forth, and you may not notice a lack of accomplishments in one area.
Computing Now: If a graduate must begin work as an intern, freelancer, or independent contractor in the field of microprocessor technology what are some tips for building a strong portfolio for presentation in possible future interviews and why?
Duwe: Stay interested in and interact with the context of your work beyond the specific piece of a project you are responsible for. Make sure you know why your project is important and why major design decisions were made. By understanding how your project fits into the whole you will be able to connect what you do to a wider number of people who may be interviewing you (and an even wider set of companies when you are looking for jobs).
In the vein of building a strong portfolio, if you have a choice between jobs or projects, work on projects—where you have specific unique contributions to make—and then complete those projects well. If you do this, you’ll be able to talk very concretely about the type of unique value you can bring to a future job in addition to demonstrating that you are capable of completing independent work.
Computing Now: Name one critical mistake for young graduates to avoid when starting their careers?
Duwe: Doing just enough to get by. To some it is seemingly easy to treat work like they treated school and do only enough to get the target grade. While this may initially work at a job, in the long term it is a recipe for disaster, in particular as new problems arise, technology changes, automation increases, and jobs become scarce. Those who work in the microprocessor field will need to be ready to quickly redirect their skills to these new technologies and problems which can be a real challenge to those who have been complacent in their jobs and education.
Computing Now: Do you have any learning experiences you could share that could benefit those just starting out in their careers?
Duwe: One of the most influential learning experiences of my career was an eight-month internship I had during the last year of my undergraduate studies. Not only did I get the opportunity to observe the inner workings of a microprocessor design company (at a time when it was making major structural changes to its operations), but I got to interact with workers across all points in their careers and get their perspectives on what made them more or less successful. Often when I think about how I should approach a career decision (such as which job offer to accept or which projects to take on), I end up thinking back to observations and discussions I had during that internship.
About Lori Cameron
Computing Now’s Lori Cameron interviewed Duwe for this article. Contact her at l.cameron@computer.org if you would like to contribute to a future computing careers article. Contact Duwe at duwe@iastate.edu.
