Research

My research is centered on power and performance optimization of modern computing systems. I investigate how processor resources can be managed dynamically to improve performance per watt while respecting system-level power constraints.

During my Ph.D. at IIIT-Delhi, I worked on runtime-level power management for multicore and multiprocessor systems running co-scheduled workloads. My work combines hardware performance monitoring, workload characterization, processor frequency control, and power capping.

Research interests

High-Performance Computing (HPC), CPU/GPU performance optimization, power management, DVFS, uncore frequency scaling, power capping, runtime systems, PMU/PMC profiling, workload characterization, performance/watt, and resource management.

Research focus

I study workloads with different compute, memory, and I/O characteristics and analyze how core frequency, uncore frequency, power budgets, and application parallelism affect performance and energy efficiency.

My recent work explores dynamic power redistribution among co-running applications, frequency sensitivity, throughput optimization, and fairness under a global power budget. The resulting systems include Cuttlefish, KarmaPM, and Fulcrum.

Education

Ph.D., Computer Science and Engineering
Indraprastha Institute of Information Technology, Delhi

B.Tech., Computer Science and Engineering
Indraprastha Institute of Information Technology, Delhi