Asmita

I am a Memory Systems Performance Architect at AMD in the Data Center GPU organization, where I work on next-generation platforms for AI, HPC, and enterprise data centers. My role focuses on workload characterization, memory subsystem performance, and microarchitecture exploration across tiered memory, scalable interconnects, and high-performance AI accelerators. I develop tools and methodologies for workload analysis, identify bottlenecks, and propose architectural solutions to improve efficiency and quality of service in complex SoC designs.

Previously, I earned my PhD from the Department of Electrical and Computer Engineering, at University of Wisconsin - Madison, in the STACS Lab, advised by Professor Joshua San Miguel. My doctoral research spanned several topics in computer architecture and systems with a flavor of security-aware design space exploration. More specifically I have worked on approximate computing, privacy-preserving traces, fully homomorphic encryption and interconnect security. Through these projects, I have developed a deep understanding of workload behavior, memory access patterns, and security vulnerabilities in modern computing platforms. During my PhD, I also interned with AMD Research, where I worked on workload characterization to identify opportunities for acceleration and optimized last-level cache management policies. My research has been recognized by several awards - most recently by Google, as part of the N+1 Institute's Reverse Pitch Competition. I have also served on several Artifact Evaluation Committees for ASPLOS, MICRO, ISCA and MLSys.

I was born and raised in Kolkata, India, renowned for its colonial architecture, artistic galleries and cultural vibrancy. Often referred to as the cultural capital of India, it has been home to six Nobel laureates and numerous luminaries across various fields. I pursued my undergraduate studies in Electronics and Instrumentation Engineering, where I gained expertise in hardware design and signal processing. Following which I moved to the United States, to pursue a Master's degree at Utah State University, where I worked on GPU power-performance trade-offs at the Bridge Lab under the supervision of Professor Koushik Chakraborty and Professor Sanghamitra Roy. My research there deepened my interest in architectural optimizations, laying the foundation for my PhD work and fueling my drive to advance the future of computing.

Beyond academia, I love to stay active - currently, I'm training (or at least trying) for a marathon. I also play tennis, albeit with questionable technique. Traveling is my way of keeping an open perspective, much like reading - a lifelong habit shaped by Tagore's verses, Rumi's reflections, and the cinematic brilliance of Satyajit Ray. Additionally, I enjoy food and learning about different cultures. And above all, I adore art in all forms - whether it's paintings, theatre, or music. Every now and then I share glimpses of my journey on Instagram to stay relatable.