Sanchari Sen
Department of Computer Science and Automation (CSA)
Indian Institute of Science (IISc) Bangalore
Research Overview
I am a faculty member in the Department of Computer Science and Automation (CSA) at Indian Institute of Science (IISc) Bangalore. My research focuses on building efficient, scalable, and sustainable AI systems through different techniques across the hardware–software stack. My work spans accelerator, compiler and algorithm design with the goal of enabling high-performance and energy-efficient AI.
Research Interests
- Efficient systems for machine learning and AI
- Hardware–software co-design for AI applications
- Hardware architecture and compiler design for AI accelerators
- Performance and energy modeling of AI systems
Publications
Full list on Google Scholar | DBLP
Selected Papers
-
RaPiD: AI Accelerator for Ultra-Low Precision Training and Inference
IEEE 48th Annual International Symposium on Computer Architecture (ISCA), 2021
-
DNNDaSher: A Compiler Framework for Dataflow-Compatible End-to-End Acceleration on IBM AIU
IEEE Micro, 2024
-
EMPIR: Ensembles of Mixed Precision Deep Networks for Increased Robustness Against Adversarial Attacks
International Conference on Learning Representations (ICLR), 2020
-
Efficacy of Pruning in Ultra-Low Precision DNNs
IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), 2021
-
Approximate Computing for Long Short Term Memory (LSTM) Neural Networks
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD) , 2018
-
SparCE: Sparsity-Aware General-Purpose Core Extensions to Accelerate Deep Neural Networks
IEEE Transactions on Computers (TC) , 2018
-
Approximate Computing for Spiking Neural Networks
IEEE Design, Automation & Test in Europe Conference & Exhibition (DATE), pages , 2017
-
Deep Compression of Pre-Trained Transformer Models
Advances in Neural Information Processing Systems (NeuRIPS) , 2022
-
14.1 A Software-Assisted Peak Current Regulation Scheme to Improve Power-Limited Inference Performance in a 5nm AI SoC
IEEE International Solid-State Circuits Conference (ISSCC) , 2024
-
Sparsity Turns Adversarial: Energy and Latency Attacks on Deep Neural Networks
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD) , 2020
-
Efficiency Attacks on Spiking Neural Networks
ACM/IEEE Design Automation Conference (DAC) , 2022
-
A Case for Generalizable DNN Cost Models for Mobile Devices
IEEE International Symposium on Workload Characterization (IISWC) , 2020
-
Dynamic Spike Bundling for Energy-Efficient Spiking Neural Networks
IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED) , 2019
-
Axformer: Accuracy-Driven Approximation of Transformers for Faster, Smaller and More Accurate NLP Models
International Joint Conference on Neural Networks (IJCNN) , 2022
Patents
-
Family Of Lossy Sparse Load SIMD Instructions
US Patent 11,663,001, May 30 2023
-
Exploiting Fine-Grained Structured Weight Sparsity In Systolic Arrays
US Patent 11,941,111, March 26 2024
-
Sparse Systolic Array Design
US Patent 11,669,489, June 6 2023
-
Stickification using anywhere padding to accelerate data manipulation
US Patent 12,468,947, Nov 11 2025
-
Generating program binaries for a transformer to process sequence concatenations having different input lengths
US Patent App. 18/344,487
-
Reusing Weights and Biases in an Artificial Intelligence Accelerator for a Neural Network for Different Minibatch Sizes of Inferences
US Patent App. 18/344,491, Nov 11 2025
-
Deep Learning Optimization Through Zero Tile Manipulation
US Patent App. 18/448,390
Biography
Sanchari Sen is an Assistant Professor in the Department of Computer Science and Automation (CSA) at Indian Institute of Science (IISc) Bangalore. Her research focuses on building efficient, scalable, and sustainable AI systems through different techniques across the hardware–software stack.
Education
-
Ph.D. in Electrical and Computer Engineering, Purdue University, USA, 2020
Dissertation: Efficient and Robust Deep Learning through Approximate Computing - B.Tech in Electronics and Electrical Communication Engineering, Indian Institute of Technology, Kharagpur, India, 2015
Professional Experience
- Assistant Professor, Department of Computer Science and Automation (CSA), Indian Institute of Science (IISc) Bangalore, India. 2026–Present
- Senior Research Scientist, AI Accelerator Compiler and Architecture Design Team, IBM T. J. Watson Research Center, Yorktown Heights, USA. 2025
- Staff Research Scientist, AI Accelerator Compiler and Architecture Design Team, IBM T. J. Watson Research Center, Yorktown Heights, USA. 2020-2024
Awards and Honors
- Outstanding (O-level) Research Division Accomplishment Award, IBM Research. 2022, 2023
- Outstanding (O-level) Technical Achievement Award, IBM Research. 2023
- Bilsland Dissertation Fellowship, Purdue University, 2019-2020
- Ross Fellowship, Purdue University, 2015-2016
- Institute Silver Medal and Swapan Kumar Saha Memorial Prize (for securing a departmental rank of 1 in the undergraduate program), IIT Kharagpur, 2015
- Bigyan Sinha Memorial Prize (for securing an institute rank of 2 in the undergraduate program), IIT Kharagpur, 2015
- Sarat Memorial Prize (for securing an institute rank of 1 in the undergraduate program among all graduating women students), IIT Kharagpur, 2015
Professional Activities
- Technical Program Committee Member: HPCA 2026, ASPLOS 2025, ISCA Industry Track 2024, ICCD 2024, DAC 2022-2024, CF 2023, IPDPS 2022
- External Review Committee: ISCA 2024, HPCA 2023, MICRO 2022
- Journal Reviewer: IEEE TC, IEEE TVLSI, IEEE TCAD, IEEE Embedded Systems Letters, IEEE Signal Processing Letters
- IBM liaison, Semiconductor Research Corporation (SRC)'s ACE (Evolvable Computing for Next-Generation Distributed Computing Systems) Center under the JUMP 2.0 program
Teaching
- UMC 102: Introduction to Computer Systems, Spring 2026
(Co-teaching with Prof. Vinod Ganapathy)
Contact
Email: sancharisen[at]iisc.ac.in
Office:
CSA 234
Computer Science and Automation
Indian Institute of Science
Bangalore 560012
India
Links:
Google Scholar |
DBLP |
LinkedIn
