Amir Raoofy

Researcher at LRZ Future Computing & CAPS
Technische Universität München, Informatik 10
Lehrstuhl für Rechnerarchitektur & Parallele Systeme (Prof. Schulz)

Address:
Room I.1.030
Boltzmannstr. 1
85748 Garching b. München

Contact:
Email: amir.raoofy@tum.de
Tel.: +49 (89) 35831 - 8871

Research interests

Programming and Usage Models
System Software and Programming
Analysis of large datasets on HPC systems
Machine Learning and Data Mining on HPC systems
Time-series analysis on industrial datasets using HPC systems
Parallelization and code optimization on accelerators and multi-core HPC systems
Porting applications to HPC systems

Projects

Dynamical Exascale Entry Platform - Software for Exascale Architectures (DEEP-SEA) ( www.deep-projects.eu ): Research project sponsored by European Commission and EuroHPC Programmes.
Commercial off-the-shelf Inference Processor ML Benchmark (MLAB): Research project in collaboration with Airbus Defence and Space GmbH, TUM chair of Big Geospatial Data Management, and Orora Technologies GmbH.
From the Edge to the Cloud and Back: Scalable and Adaptive Sensor Data Processing (SensE): Research project funded by Bayerische Forschungsstiftung in cooperation with IfTA Ingenieurbüro für Thermoakustik GmbH
Gas Turbine Optimization using Big Data and Machine Learning (TurbO): Research project funded by Bayerische Forschungsstiftung in cooperation with IfTA Ingenieurbüro für Thermoakustik GmbH
Porting Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics (LULESH) to LAIK (A Library for Automatic Data Migration in Parallel Applications) Code

Teaching

WS 22/23: Praktikum - Evaluation of Modern HPC-Architectures and - Accelerators
WS 21/22: Praktikum - Evaluation of Modern HPC-Architectures and - Accelerators
SS 21: Praktikum - Evaluation of Modern HPC-Architectures and - Accelerators
SS 20: Seminar - Cloud Computing: Internet of Things Technologies
SS 19: Seminar - Geschichte der Rechnerarchitektur
WS 18/19: Praktikum - Advanced Topics in Computer Architecture and Parallel Systems
SS 18: Lecture - Parallel Programming (Central Tutorial)

Publications

Amir Raoofy, Roman Karlstetter, Martin Schreiber, Carsten Trinitis, Martin Schulz: Overcoming Weak Scaling Challenges in Tree-based Nearest Neighbor Time Series Mining, ISC 2023.
Yi Ju, Amir Raoofy, Dai Yang, Erwin Laure, Martin Schulz: Exploiting Reduced Precision for GPU-based Time Series Mining, IPDPS 2022.
Max Ghiglione, Amir Raoofy ,Gabriel Dax, Gianluca Furano, Richard Wiest, Carsten Trinitis, Martin Werner, Martin Schulz, Martin Langer: Machine Learning Application Benchmark for In-Orbit On-Board Data Processing, OBDP 2021.
Amir Raoofy, Gabriel Dax, Max Ghiglione, Martin Langer, Carsten Trinitis, Martin Werner, and Martin Schulz: Benchmarking Machine Learning Inference in FPGA-based Accelerated Space Applications, MLBench 2021.
Roman Karlstetter, Amir Raoofy, Martin Radev, Carsten Trinitis, Jakob Hermann, and Martin Schulz: Living on the Edge: Efficient Handling of Large Scale Sensor Data, CCGrid 2021.
Amir Raoofy, Roman Karlstetter, Dai Yang, Carsten Trinitis, Martin Schulz: Time Series Mining at Petascale Performance: ISC High Performance 2020 (Winner of the Hans Meuer Best Paper Award).
Roman Karlstetter, Robert Widhopf-Fenk, Jakob Hermann, Driek Rouwenhorst, Amir Raoofy, Carsten Trinitis, Martin Schulz: Turning dynamic sensor measurements from gas turbines into insights: a big data approach. ASME Turbo-Expo conference 2019.
Amir Raoofy, Dai Yang, Josef Weidendorfer, Carsten Trinitis and Martin Schulz: Enabling Malleability for Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics using LAIK. PARS Workshop 2019
Dai Yang, Moritz Dötterl, Sebastian Rückerl and Amir Raoofy: Hardening the Linux Kernel agains Soft Errors. Poster for The 13th International School on the Effects of Radiation on Embedded Systems for Space
Arash Bakhtiari, Dhairya Malhotra, Amir Raoofy, Miriam Mehl, Hans-Joachim Bungartz, George Biros. A parallel arbitrary-order accurate AMR algorithm for the scalar advection-diffusion equation. SC '16.