Sahit Kavukuntla

Logo

View My GitHub Profile

Hello there! I’m Sahit Kavukuntla. I’m a GPU Architect at Nvidia, under the SM Core team. I graduated from the Georgia Institute of Technology with my Master’s degree in Computer Science in May 2024, with specializations in Machine Learning and Computing Systems. I completed my Bachelor’s in Computer Science from Georgia Tech in May 2023, with threads in Systems & Architecture and Artificial Intelligence.

I supported students in Dr. Joesph Jaeger’s CS 4510 Automata and Complexity course as a Graduate Teaching Assistant in Fall 2023. As part of my role, I assisted students in comprehending different models of computation and determining the appropriate situations in which to apply them.

In Summer 2023 I worked at Apple in Cupertino as a Platform Architecture Intern under John Dorsey, Bryan Hinch, Ronit Banerjee, and Karthic Palaniappan. I developed analysis within the performance controller to flag and correct threads that are inefficiently using Apple’s asymmetric multiproccesors, reducing power consumption by 1% within the SoC complex while maintaining performance.

In Summer 2022 I worked at Nvidia under Hardy Doelfel, Christian Zander, and Lily Huang in the Linux Graphics Drivers team as a System Software Intern. As part of Nvidia’s initiative to open source their Linux GPU kernel drivers, I helped debug drivers and test performance analytics to support users seeking to use this code. I also worked in the Graphics System Processor Performance team to create an interrupt profiler tool to assist engineers in improving performance and execution.

As a student advisor of Team Phoenix under Rich Vuduc, I studied and taught high performance computing (AKA supercomputing), along with software/hardware optimizations and modularizations. I competed in Nvidia’s GPU Hackathon and the International Supercomputing Conference, ISC22, and I was our team’s advisor for our students competing in the Student Cluster Competition, IndySCC22.

I was one of the Master Peer Instructors for 3D Printing and Prototyping at the HIVE Makerspace at Georgia Tech, where I helped students and faculty use our tools, maintained our machines and space, and trained new Peer Instructors every semester on how to use our printers and assist others with any of their printing needs.

In my spare time I organized events for the Georgia Tech Honors Program as part of the Community Development Committee on the Honors Leadership Council. I’ve planned, organized, and hosted events ranging from smores night and succulent painting to paintball and Honors Prom.

In Summer 2021 I worked at Nvidia in RM Core Microcode Security as a System Software Intern under James Xu and Jerry Zhou, implementing a screen capture prototype for an anti-cheat neural algorithm onto a secure RISCV processor. My prototype captured a screenshot of the screen every frame during gameplay and passed it through the algorithm to detect cheating, bypassing any hacks occurring at the application or kernel level.

You can read more about me at my LinkedIn, look at my Emotion Aware Music Generation project, or view my resume.

I can be contacted through my email. Thanks for visiting my website!