School of Engineering and Applied Science Honors Dave Garrett’s Achievementsmkw3a@virginia.edu
Dave Garrett loves engineering. “My whole career has been solving puzzles, inventing things, building new systems,” Garrett said. “I can’t imagine doing anything else.”
The University of Virginia School of Engineering and Applied Science honored Garrett’s passion and research impact, awarding him the 2021 School of Engineering Achievement Award.
Garrett is a Double Hoo in electrical engineering. He earned his Bachelor of Science Degree in 1994 and his Ph.D. in 2000. Garrett’s vitae has the depth and breadth of corporate R&D. He joined Lockheed Martin’s engineering leadership development program between degrees, conducted research at Bell Labs and later joined Broadcom as associate technical director.
Garrett also embraces the freedom offered by start-ups. Beceem, a semiconductor company based in Santa Clara, Ca., hired Garrett as director of engineering; he now works as vice president of hardware engineering at Syntiant, a semiconductor firm based in Irvine, Ca.
Garrett is best known for low-power chip design. During his tenure with Broadcom, Garrett enabled well over 50 different chip products, designing circuits, architectures, systems-on-chip and algorithms for wired and wireless communications and influencing the development and adoption of wireless standards, for which he was elected fellow of the Institute of Electrical and Electronics Engineers.
Garrett is most proud of a game-changing WiFi technology called beamforming. Beamforming is a feature added to wireless device chipsets that strengthens WiFi connections, performing billions of calculations per second to steer energy to a particular user in real time.
“This was around 2010 or 2011, right in the middle of the internet of things revolution when products and services were just beginning to come on line,” Garrett said.
Integrating the beamforming feature in Broadcom’s chip sets gave the company a competitive advantage at a critical moment in the market’s development, shipping billions of those devices. “Still to this day, every high-end smart phone ships with this technology embedded.”
Garrett is again in the thick of a technology revolution, building machine learning architectures from the ground up for Syntiant, which specializes in very low-power devices for edge computing.
“We’re putting machine learning at the edge, computing where the user is,” Garrett said. Garrett and his Syntiant teammates have built two generations of the firm’s neural decision processors, adopted into a multitude of device designs from ear buds to automobiles. The team’s most recent chip earned best product of the year at the tinyML summit.
“My Ph.D. in low-power chip design is fundamental to what we do at Syntiant,” Garrett said. “We build ultra-low-power devices that can run neural networks on sensor data locally rather than connecting to the cloud or a remote server. Low power is the key component, ensuring the devices can be battery powered.”
Garrett credits James H. Aylor, UVA Engineering dean emeritus and interim chair of the Charles L. Brown Department of Electrical and Computer Engineering, for pointing him toward a career in the semiconductor industry.
“The first really big moment in my career happened when I was a fourth-year electrical engineering student,” Garrett recalled. He took a class with Aylor, where the task was to build a microprocessor from just the instruction set and map across three field-programmable gate arrays; a gate array is a hardware circuit that a user can program to carry out one or more logical operations. “That was the first moment I could really say, ‘I can build something new from scratch.’ All of a sudden, I could see this puzzle that I wanted to solve for the rest of my life.”