Caroline Trippel

Ph.D. 2019, Computer Science

Mishawaka, Indiana

B.S., Purdue University; M.A. Princeton University

"I am a Princeton engineer because I want to discover solutions to challenges that impact our modern society."

Research interests

Her research models different patterns in computer hardware that can be exploited to breach security.

Image from Trippel et al. (2018).

Caroline’s research lies at the intersection of security and hardware design. Her work verifies whether hardware designs will run in insecure ways or leak information. Much prior work in this area has involved ad hoc, manual testing, so she applies rigorous, formal approaches to systematically evaluate the security of hardware designs. Such approaches specify or describe a hardware design in mathematical terms, allowing automated analysis to find security bugs.

Her most recent project, her dissertation, describes how hardware designers can test how securely and correctly programs run on specific hardware. Modern hardware reorders instructions to execute software more efficiently and increase performance, but this practice can introduce security risks. Caroline’s work guarantees the correct ordering of the program to optimize correctness, security, and performance.

Career goals

Another formative part of Caroline’s experience at Princeton was lecturing in COS375/ELE375. “I liked being able to help students arrive at particular conclusions on their own,” she said, adding that she learns best when the material is presented in a way that incrementally builds up intuition for a concept so she essentially teaches the material to herself. She enjoyed “breaking a problem down into more digestible pieces for students so that, hopefully, they can understand the basic concepts and then almost arrive at the idea by themselves.” This opportunity to teach helped solidify her interest in pursuing an academic career. She is currently a faculty member at Stanford, after completing a year as a research scientist at Facebook earlier last academic year.

Path to Princeton

Caroline found time to participate in running throughout her time in the graduate program. She is seen here with Greg Darnell graduate student in Quantative and Cellular Biology, advised by Professor Barbara Engelhardt in CS, and Themis Melissaris, grad student in CS in Professor Margaret Martonosi’s research lab.

Image courtesy the researcher.

Caroline came to Princeton primarily because of her advisor Margaret Martonosi, whose work strongly interested her. From visits and discussions, it “felt like a good student-advisor match,” she said. “More so than picking a school, graduate school is about picking an advisor who you think will be a great mentor and great person to work with.” 

At Princeton, the computer science department was compact and manageable enough that she knew the experts to ask questions outside of her area. Princeton’s culture of collaboration also made this really easy and was an important feature of the academic environment for Caroline.

There were also many events for graduate students to interact and get to know each other, which she also said was useful. “I always like places where I feel like there’s a good sense of community, and I felt like there was definitely that at Princeton and the computer science department, as well,” she said.

When Caroline started applying to graduate programs she was unsure about the differences between master’s and doctoral programs until Martonosi explained that a Ph.D. program is basically more courses, but with a research component at the end. Originally, she had just wanted to get more depth in some particular area before finding an industry job, but ended up pursuing a doctorate because she liked research process so much. So for her, the graduate degree was an exploration of ideas in which her interest built over time.

Early interest in STEM

Caroline’s earliest engineering memory was playing with BRIO blocks and trains. “You can build all types of neat, intricate track designs,” she said, adding that design appealed to her “from both an artistic perspective and a construction perspective.” She went on to participate in a lot of science fairs and science fair projects but didn’t really know what she wanted to do when she went to college at Purdue. She chose engineering because of the program’s strong reputation at Purdue and specialized in computer engineering with encouragement of her mother who earned degrees in math and computer science.

Extracurricular activities

Outside Small World Coffee in downtown Princeton was a great place to relax and unwind for Caroline and her fiancé, Greg Darnell.

Image courtesy the researcher.

“I guess for me,” Caroline said, “Princeton was a really perfect place to do a PhD.” She knew many of the students in her cohort, as well as in the department as a whole, because of the manageable program size. She also liked the town of Princeton. “I was able to walk to my office. I think there’s a enough going on for the amount of time you have pursuing a Ph.D.”

She enjoyed going to the McCarter Theatre and Richardson Auditorium to see plays or concerts. She took advantage of the ODUS Passport to the Arts program to see concerts and attend events. “It was fun to be able to take advantage of those things in Princeton. I felt like it provided a good environment for thinking, but also, there were fun things to do in my free time,” she said.

Awards

CheckMate chosen as an IEEE MICRO Top Pick of 2018 (top 12 computer architecture papers of 2018)
Selected for 2018 MIT Rising Stars in EECS Workshop
Selected for 2018 ACM Heidelberg Laureate Forum
TriCheck chosen as an IEEE MICRO Top Pick of 2017 (top 12 computer architecture papers of 2017)
NVIDIA Graduate Fellowship Recipient, Fall 2017–Spring 2018
NVIDIA Graduate Fellowship Finalist, Fall 2016–Spring 2017

Selected publications and presentations

Caroline Trippel, Daniel Lustig, and Margaret Martonosi. “Security Verification through Automatic Hardware-Aware Exploit Synthesis: The CheckMate Approach”. IEEE Micro, 39 (3), May-June 2018.

Caroline Trippel, Daniel Lustig, and Margaret Martonosi. “CheckMate: Automated Synthesis of Hardware Exploits and Security Litmus Tests”. In Proceedings of the 51st IEEE/ACM International Symposium on Microarchitecture (MICRO), Fukuoka, Japan. October 2018.

Caroline Trippel, Yatin A. Manerkar, Daniel Lustig, Michael Pellauer, and Margaret Martonosi. “Full-Stack Memory Consistency Model Verification with TriCheck”. IEEE Micro, 38 (3), May-June 2018.