Jakob Winkler

PhD Student at JKU Linz

Informations

For this public online CV contact details have been omitted. For more information, please email me at . Thank you.

Profile

I am a PhD student at the Institute of Signal Processing at Johannes Kepler University Linz, Austria. My current work focuses on algorithms and hardware architectures for embedded machine learning where I am working on ensemble learning methods for resource-constrained systems.

Due to my strong computer engineering and computer science background my current interests also include interfaces, drivers and verification of SoC systems and peripherals.

Professional Experience

since 11/2022

Johannes Kepler University Linz, Institute of Signal Processing, Linz, Austria

PhD Student

Algorithms and hardware architectures for embedded machine learning focusing on digital implementation of ensemble methods such as gradient boosted decision trees.

10/2020 – 10/2022

Danube Dynamics, Linz, Austria

Lead Software Architect

System design and implementation for an NVIDIA Jetson based embedded system serving as a platform for embedded computer vision applications, including Linux kernel builds, custom drivers, and root filesystems.

API design and system architecture for a distributed system for embedded computer vision applications.

12/2019 – 09/2020

Silicon Austria Labs, Linz, Austria

Junior Scientist, RF Architecture & Design

Design and implementation of measurement systems for 77 GHz mmWave/RF based on the Xilinx RFSoC platform. Implemented signal generation and capture as well as a programming interface for automated measurement setups.

07/2019 – 11/2019

University of Applied Sciences Upper Austria, Hagenberg, Austria

Research Assistant

Design and implementation of prototypes for FPGA based digital video signal processing of multiple synchronized concurrent high-resolution streams.

02/2017 – 09/2019

Intel Mobile Communications, DMCE GmbH & Co KG, Linz, Austria

R&D Trainee

Software implementation for pre- and post-silicon verification tooling. Worked on optimization and bugfixing of a custom scripting environment and bytecode compiler.

Education

09/2017 – 09/2019

University of Applied Sciences Upper Austria, MSc Embedded Systems Design, Hagenberg, Austria

with distinction

Thesis “A Framework for FPGA-based Video Signal Processing”

Teaching assistant for concurrent software development, Linux drivers, and VLSI design courses.

10/2014 – 09/2017

University of Applied Sciences Upper Austria, BSc Hardware-Software-Design, Hagenberg, Austria

with distinction

Thesis topic: Usability of scripting environments in EDA tools

09/2009 – 06/2014
HTL Krems, Matura (high school diploma), with distinction

Awards and Honors

HITB+ PRO CTF 2021
Invited finalist at HITB+ PRO CTF in Abu Dhabi, UAE with HgbSec

Publications

J. Winkler and M. Lunglmayr, “A Resource-Efficient SoC Accelerator for Boosted Decision Trees” 2024 22nd IEEE Interregional NEWCAS Conference (NEWCAS), Sherbrooke, Canada, 2024 (accepted for publication)

J. Winkler, D. Hackl and M. Lunglmayr, “Efficient Hardware Architecture for Random Forest Training” Computer Aided Systems Theory – EUROCAST 2024. EUROCAST 2024. Lecture Notes in Computer Science, Springer, Cham. (accepted for publication)

J. Winkler and M. Lunglmayr, “FPGA Processing of Decision Tree Ensembles Stored in External DRAM,” 2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), Tenerife, Canary Islands, Spain, 2023, pp. 1-6, doi: 10.1109/ICECCME57830.2023.10252353.