Amir Raoofy

Researcher at LRZ Future Computing

Research Interests

System Software and System Programming
High Performance and Smart Networks in HPC
Programming and Usage Models
Analysis of large datasets on HPC systems
Machine Learning and Data Mining on HPC systems
Data mining on using HPC systems
Parallelization and code optimization on accelerators and multi-core HPC systems

Chair of Computer Architecture and Parallel Systems, Technical University of Munich

Thesis: Time Series Mining on High Performance Computing Systems (2024 - Supervisor: Prof. Dr. Martin Schulz)

Thesis: Semi-Implicit Solver for Over-land Flows on HPC Systems (2017 - Supervisor: Prof. Dr. Hans-Joachim Bungartz)

IPDPS, Supercomputing (SC), ISC, Computing Frontiers (CF), ScalPerf, MLSys, OBDP, HIPS, ICPE, etc

ScalNEXT: BMBF project link
DEEP-SEA: EU-wide project link
MLAB: collaboration with Airbus, TUM, Orora Technologies sponsored by ESA, link
TurbO and SensE: collaboration with IfTA GmbH sponsored by Bayerische Forschungsstiftung link

Practical Course: Experimental Evaluation and Characterization of Quantum Computing Systems (TUM: WS24/25)
Practical Course Experimental Evaluation of Modern Computing Systems and Accelerators (TUM: SS25, WS23/24, SS23, WS22/23, WS21/22, WS20/21)
Seminar: Cloud Computing: Internet of Things Technologies (TUM: SS20)
Seminar: History of Computer Architecture (TUM: SS19)
Practical Course: Advanced Topics in Computer Architecture and Parallel Systems (TUM: WS18/19)
Recitation for Lecture Course Parallel Programming (TUM: SS18)

Progressing Non-blocking Two-Sided Communication Using BlueField DPUs, ARCS'25

AutoBench: A Holistic Platform for Automated and Reproducible Benchmarking in HPC Testbeds, ICPE'25

Performance Comparison of GPU Programming Models Using HeCBench Benchmarks, CF'25

Phase-aware System-Side Sampling for HPC, CF'23

Modeling and Simulation of MPI Communication Dynamics in InfiniBand Networks Using SimGrid, PARS'25

Exploiting reduced precision for GPU-based time series mining, IPDPS'22

Benchmarking Machine Learning Inference in FPGA-based Accelerated Space Applications, MLBench'21

Time Series Mining at Petascale Performance, ISC'20 (Hans Meuer Best Paper Award)

Enabling Malleability for Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics using LAIK, PARS'19

Contribution to the supervision of around 30+ Bachelor and Master student theses