The Parallel Systems Group carries out research to facilitate the use of extreme scale computers for scientific discovery. We are especially focused on tools research to maximize the effectiveness of applications running on today’s largest parallel computers. Our expertise includes performance measurement, analysis, and optimization in addition to debugging and power optimization. We also have expertise in existing and new programming models, power-aware supercomputing, fault tolerant computing, I/O, and numerical kernel optimization.

Group Lead

Ignacio Laguna: program analysis, software correctness, compiler analysis, resilience, fault tolerance, debugging

Research Staff

Kshitij Bhardwaj: heterogeneous systems-on-chip, hardware accelerator design, system optimization using machine learning, reconfigurable computing, and on-chip interconnection networks

Giorgis Georgakoudis: program analysis, compiler optimization (Clang/LLVM), parallel programming models and runtimes, performance analysis and tuning, fault tolerance

Maya Gokhale: data intensive computing, reconfigurable computing, co-processor accelerators, HPC architectures

E Green

Hari Hariharan Devarajan: high performance storage and I/O, storage for AI, heterogeneous storage architectures, composable storage architecture, and performance measurement and analysis tools

Aniruddha Marathe: power-aware and power-constrained supercomputing, performance analysis and optimization, HPC in cloud

Harshitha Menon: approximate computing, mixed-precision computing, performance analysis, fault tolerance, run-time system, load balancing, charm++

Dan Milroy: converged HPC and cloud resource and job management, software correctness, program analysis

Kathryn Mohror: performance measurement and analysis, fault-tolerance, scalable tools, high performance storage and I/O

Konstantinos Parasyris: programming models, fault tolerance, low-energy and low-power technologies, undervolting

Tapasya Patki: power-aware supercomputing, HPC scheduling and resource management, performance modeling and optimization, multi-constraint performance optimization, low-level heterogeneous architecture interfaces

Barry Rountree: power telemetry and control, performance optimization under power bounds, machine learning for climate simulation performance optimization, system programming tools

Chen Wang