Personal Webpage of Zachary Susskind.

A bit about me.

I just recently finished my PhD in Electrical and Computer Engineering at The University of Texas at Austin, where I was a member of the Laboratory for Computer Architecture research group under Dr. Lizy John. I also did my Bachelor's at UT, from 2015 to 2019.

Primarily, I'm interested in ways to improve the efficiency of machine learning through algorithm-hardware co-design. My dissertation focused on weightless neural networks, which use lookup tables as their primary computational unit, rather than neurons based on linear scaling of activations through multiply-accumulate or XNOR-popcount operations. WNNs are theoretically really useful, since individual neurons can learn non-linear behaviors, which lets us do more with smaller and shallower models. This is particularly good for very low-power "extreme edge" devices such as embedded smart sensors, where DNNs and BNNs aren't necessarily sufficiently scalable. There are a lot of challenges in constructing and training effective WNNs, but that's why I did a PhD on them.

Personally, I'm interested in archery, cooking, woodworking, bouldering, and occasionally still find time for gaming. I also restored a 1946 Raleigh Tourist bicycle a while back.

Contact Me.

Email is definitely the best way. It's my first initial followed by my last name, and it's at gmail.com.

If you like, you can also reach out to me on LinkedIn.

I also have a GitHub, where you can view the code for the BTHOWeN and ULEEN projects and some other miscellaneous work.

Recent Updates.

08/24: I've finished my PhD and graduated! Ciao, UT!
07/24: I have successfully defended my dissertation, "Weightless Neural Networks for Fast, Low-Energy Inference."
05/24: Our paper, "LogicNets vs. ULEEN : Comparing Two Novel High Throughput Edge ML Inference Techniques on FPGA", was accepted to MWSCAS 2024 (co-first author).
05/24: Our paper, "Differentiable Weightless Neural Networks", was accepted to ICML 2024! (co-first author).
12/23: I presented a short paper, "Dendrite-inspired Computing to improve Resilience of Neural Networks to Faults in Emerging Memory Technologies", at ICRC 2023.
10/23: My first-author paper, "ULEEN: A Novel Architecture for Ultra Low-Energy Edge Neural Networks", was accepted to ACM TACO.
11/22: My first-author work, "An FPGA-Based Weightless Neural Network for Edge Network Intrusion Detection", was accepted to FPGA 2023 as a poster session.
07/22: My first-author paper, "Weightless Neural Networks for Efficient Edge Inference", was accepted to PACT 2022.
06/22: My first-author paper, "Pruning Weightless Neural Networks", was accepted to ESANN 2022.