Heads of the Research Group: Dr. Zoltán NAGY, Prof. Péter SZOLGAY
Contact: nagy.zoltan@itk.ppke.hu , szolgay.peter@itk.ppke.hu
In our laboratory, we work on the design and measurement of analog and digital circuits. Currently, the main focus is the design of digital circuits on FPGA devices. The applications cover a wide range, from the simulation of complex physical systems through image processing to the acceleration of search in bioinformatics databases. We have developed a compact, low-power image processing architecture with the UAV Vision Lab. We have implemented a hybrid FPGA–CPU-based architecture for object recognition, which is the visual analysis subsystem of a prosthetic arm controller. The tasks of the system were partitioned between the FPGA fabric and the processors, based on performance measure-ments of the building blocks.
We accelerated a charged alpha-helix detection algorithm on FPGA with the Structural Biology and Proteomics Lab. We accelerated the simulation of micromagnetic arrays on FPGA in cooperation with the Optics, Magnonics and New Gen-eration Hardware Laboratory. During circuit design, in addition to traditional hardware de-scription languages (VHDL, Verilog), high-level synthesis plays an increasingly important role, as it enables the architecture to be described in C/C++.
Accelerating Computational Fluid Dynamics (CFD) simulation on FPGA: shock waves in a 2D cross-section of a scramjet engine.
Future research directions, collaboration opportunities
Our lab contributes to the acceleration of computationally intensive tasks on programmable logic devices. After analyzing the performance of the algorithms, we identify bottlenecks and partition the tasks between different architectures (CPU/GPU/FPGA).
Key publications
- Fejér, A., Nagy, Z., Benois-Pineau, J., Szolgay, P., de Rugy, A. and Domenger, J-P. (2022). Hybrid FPGA-CPU-Based Architecture for Object Recognition in Visual Servoing of Arm Prosthesis JOURNAL OF IMAGING 8:2 p. 44.
- Zarandy, A., Horvath, A. and Szolgay, P. (2018) . CNN Technology-Tools and Applications IEEE CIRCUITS AND SYSTEMS MAGAZINE 18:2 pp. 77-89.
- Nagy, Z., Nemes, Cs., Hiba, A., Csík, Á., Kiss, A., Ruszinkó, M. and Szolgay, P. (2014). Accel-erating unstructured finite volume computations on field-programmable gate arrays CONCURRENCY AND COMPUTATION: PRACTICE AND EXPERIENCE 26(3) pp. 615-643.
