PARS Research Group at IZTECH
PARS (Parallel Systems) Research Group is led by Assistant Professor Işıl Öz. We are part of Computer Engineering Department at IZTECH.
Our group focuses on various issues:
- Multicore/GPU architectures
- Performance and reliability analysis of parallel systems
- Parallel software development
- Parallel programming models
We are always looking for motivated students interested in related topics. Please feel free to email Işıl!
Recent News
[25/09/24] Emre presented our paper, titled “Community Detection for Large Graphs on GPUs with Unified Memory” in IEEE High Performance Extreme Computing Conference (HPEC).
[11/09/24] Işıl attended ROAR-NET (Randomised Optimisation Algorithms Research Network) Working Group-1 meeting in Wuppertal, Germany.
[01/09/24] Furkan attended CERCIRAS Training School in Klagenfurt, Austria.
[10/07/24] Işıl attended PUMPS+AI ACM Europe Summer School in Barcelona, Spain, in the context of Erasmus Staff Training Mobility. She attended the lectures and presented our project poster titled “RAPPROX - Resource-Aware Compiler Design for Approximate Computing Techniques in GPGPUs” in the poster session.
[04/07/24] Işıl attended CERCIRAS Closing Conference in Maribor, Slovenia. In the 3rd workshop on Resource AWareness of Systems and Society (RAW 2024), she presented our paper titled “Evaluating CUDA-Aware Approximate Computing Techniques”.
[22/05/24] Our project titled “Efficient Graph Neural Networks” has been accepted within the scope of the European High Performance Computing Joint Undertaking (EuroHPC JU) Benchmark And Development Access Calls. Our team will have access to GPU resources of LUMI supercomputer.
[17/04/24] Yağız’s paper, titled “Performance-Reliability Tradeoff Analysis for Safety-Critical Embedded Systems with GPUs” has been accepted for presentation in BAŞARIM conference.
[28/03/24] Işıl has been certified to teach Scaling CUDA C++ Applications to Multiple Nodes as NVIDIA Deep Learning Institute (DLI) Certified Instructor.
[07/03/24] Ercüment’s paper, titled “Compiler-Managed Replication of CUDA Kernels for Reliable Execution of GPGPU Applications,” has been accepted for publication in Journal of Circuits, Systems, and Computers.
[06/02/24] Işıl joined CA22137 - Randomised Optimisation Algorithms Research Network (ROAR-NET) COST Action as a working group member.
[20/01/24] Işıl attended HiPEAC conference in Munich, Germany. She gave a talk at Reproducibility Student Challenge and chaired Code Generation session in the paper track.
[08/12/23] Yağız and Emre defended their MSc thesis successfully.
[20/09/23] We are organizing HiPEAC Reproducibility Student Challenge, which will take place in Munich at the HiPEAC conference on 18 January 2024. Undergraduate or graduate students are strongly encouraged to join.
[18/09/23] Our project titled “Heterogeneous Parallel Programming Course with Hands-on Experience” has been accepted within the scope of the NVIDIA-BTF Disaster Response Innovation and Education Grant. NVIDIA donates Jetson Nano Developer kits and financial support to develop a real-time object detection application for a disaster relief situation in our Heterogeneous Parallel Programming course.
[10/09/23] Serhat, Furkan and Emre attended CERCIRAS (Connecting Education and Research Communities for an Innovative Resource Aware Society) Training School in Riga, Latvia.
[01/08/23] Işıl has been invited to the Technical Program Committee (TPC) of the Design, Automation and Test in Europe Conference (DATE 2024) for topic E3 “Machine Learning Solutions for Embedded and Cyber-Physical Systems”.
[10/07/23] Burak defended his MSc thesis successfully.
[25/05/23] Işıl has been invited to the Program Committee (PC) of EduHPC23 workshop co-located with SC23.
[16/01/23] Işıl attended HiPEAC conference in Toulouse, France.
[06/11/22] Burak’s paper, titled “Soft error vulnerability prediction of GPGPU applications,” has been accepted for publication in The Journal of Supercomputing.
[17/10/22] Our TÜBİTAK 2519 project titled “RAPPROX – Resource-Aware Compiler Design for Approximate Computing Techniques in General Purpose Graphics Processing Units (GPGPU) Applications” has been accepted within the scope of CERCIRAS COST Action.
[26/09/22] Burak and Emre attended CERCIRAS (Connecting Education and Research Communities for an Innovative Resource Aware Society) Training School in Split, Croatia.
[08/09/22] Işıl gave a talk titled “Trusted Execution Inside Secure Enclaves” in the ELISA summit about her Linux Foundation mentorship.
[23/08/22] Our project titled “Efficient Community Detection in Biological Networks” has been accepted within the scope of the European High Performance Computing Joint Undertaking (EuroHPC JU) Benchmark And Development Access Calls. Our team will have access to GPU resources of HPC Vega supercomputer.
[25/07/22] Işıl has been invited to the Technical Program Committee (TPC) of the Design, Automation and Test in Europe Conference (DATE 2023) for topic E3 “Machine Learning Solutions for Embedded and Cyber-Physical Systems”.
[23/06/22] Ercüment defended his MSc thesis successfully.
[10/03/22] Burak presented our paper titled “Predicting the Soft Error Vulnerability of GPGPU Applications” in the 30th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP 2022). Link
[20/12/21] Işıl has been certified as NVIDIA Deep Learning Institute (DLI) Instructor and University Ambassador. She will bring free accelerated computing and deep learning workshops for students and lecturers.
[26/10/21] We attended HiPEAC CSW Student Challenge in Lyon and presented our work.