Bhumika Mittal

Currently, I am a software engineer at Amuse Labs. I recently graduated from Ashoka University, where I studied Computer Science and Mathematics. My current research interests are broadly within theoretical computer science, and span several topics including foundational cryptography (both classical and quantum), complexity theory, and formal logic.

My undergraduate thesis, titled “Ring Trapdoor Functions”, focused on the theory of lattice-based cryptographic primitives. As part of this and related work, I have primarily worked on lattice-based post-quantum public-key encryption and signature schemes, as well as the fine-grained complexity of lattice problems.

Previously, I have worked with the Data Systems Group at the Max Planck Institute for Software Systems, the Centre for Artificial Intelligence and Robotics (CAIR) at DRDO, IIT Delhi, and the nanoDC Lab at IIT Gandhinagar.

updates

Jun, 2025	I will be attending the Programming Languages Mentoring Workshop at PLDI 2025 in Seoul.
Dec, 2024	I will attending the ACM India Winter School on Modern Cryptography 2024, hosted by the CYSTAR Lab at IIT Madras.
Aug, 2024	I will be in Budapest attending the Mathematics for post-quantum cryptanalysis conference. Excited to meet new and old friends and anyone interested in PQC!
Jul, 2024	Attending The Cornell, Maryland, Max Planck Pre-doctoral Research School in Computer Science (CMMRS) 2024.
Jun, 2024	I will be attending the Undergrad Architecture Mentoring Workshop (uArch) at ISCA 2024 in Buenos Aires. Excited to meet people interested in computer architecture and systems!

selected work

AxLaM: Energy-Efficient Accelerator Design for Language Models for Edge Computing

Tom Glint, Bhumika Mittal, Santripta Sharma, and 6 more authors

In Philosophical Transactions A, 2025

Abstract WEB PDF

Modern language models like BERT have revolutionized natural language processing tasks but are computationally intensive, limiting their deployment on edge devices. This paper presents an energy-efficient accelerator design tailored for encoder-based language models, enabling their integration into mobile and edge computing environments. A dataflow-aware hardware accelerator design for language models inspired by SIMBA, use of Approximate Fixed-Point POSIT (AFPOS) based multipliers, and use of high bandwidth memory achieves significant improvements in computational efficiency, power consumption, area, and latency compared to the hardware-realized scalable accelerator Simba. Compared to SIMBA, AxLaM achieves 9x energy reduction, 58% area reduction and 1.2x improved latency, making it suitable for deployment in edge devices. The energy efficiency of AxLaN is 1.8 TOPS/W, 65% higher than FACT which needs pre-processing of language model before implementing it on the hardware.
On the Existence of Balanced Generalized de Bruijn Sequences

Bhumika Mittal, Haran Mouli, Eric Tang, and 1 more author

In Discrete Mathematics, 2023

Abstract WEB PDF

A balanced generalized de Bruijn sequence with parameters (n,l,k) is a cyclic sequence of n bits such that (a) the number of 0’s equals the number of 1’s, and (b) each substring of length l occurs at most k times. We determine necessary and sufficient conditions on n,l, and k for the existence of such a sequence.