CEVT is an innovation and development company at the forefront of mobility. Modular development, ground-breaking virtual engineering, software development and continuous innovation enable us to deliver world-class technology. Today we deliver to different automotive brands in the Geely family such as Geely Auto, Volvo Cars, and Lynk & Co and are based at Lindholmen in Gothenburg.
Master Thesis Project
Title: Decentralized AI and machine learning on the edge, using Single Instruction Multiple Data (SIMD) architectures
This project from CEVT will be part of the joint master’s thesis projects at AI Sweden (National center of applied AI, and hosts the AI Sweden Edge Lab). They provide an arena where students and experts in the field meet, and where students get to explore their topics using state-of-the-art, industrial-grade hardware and tools with close support from AI Sweden staff.
Single Instruction Multiple Data (SIMD) architecture was the base for the early vector supercomputers. Small-scale SIMD has since then become popular in embedded systems. For graphics cards, SIMD has largely taken over the processing of 3D graphics from the CPU. There are also benefits to using SIMD in the area of decentralized AI and machine learning on the edge.
- The focus of this thesis project is to implement AI and machine learning on an edge device using an ARM Cortex Neon SIMD extension.
- As a base, the open-source libraries Ne10 can be used. On that basis, this thesis project will explore different ways to implement AI/ML functions.
- We can decide on exploring library functions or more specific physics functions later when planning the thesis project more in detail. Here the applicants have the possibility to express their own interests.
- Implementation can be done either using an emulator or real hardware.
- A correlation and comparison to other hardware and implementation methods shall be done by collaborating at AI Sweden and with other thesis projects.
- Literature study
- Set up tools and development environment
- Develop and implement AI/ML functions
- Perform study in the simulator or/and on the target environment
- Do a correlation and comparison study (benchmarking) towards other hardware and implementation methods
- Analyze results and evaluate benefits and disadvantages.
- Recommend implementation methods for AI/ML on the edge using ARM Cortex Neon SIMD extension
- Documentation and thesis writing.
- Master student in Computer Science, Electrical Engineering, Engineering Physics or similar.
- Good knowledge in AI/ML, Embedded systems SW/HW, Programming, Algorithms,
- Interest in AI and Machine Learning implementation, in low-level programming methodology and processor architecture.
- Ability to take initiative and work independently as well as in a team.
Number of students: 1 or 2, please apply with CV and cover letter.
Starting date: January 2022. The duration of the thesis work is 20 weeks
For more information please contact:
Henok Fessehatsion, Basic Software Developer, +46 729 88 84 96,
Thyagaraja Naidu, Senior Software Architect, +46 738 62 17 26, firstname.lastname@example.org
Stefan Carlsson, Principal Expert Embedded Systems, +46 721 84 36 76, Stefan.Carlsson@cevt.se
Last application date: 2021-11-15
Apply today. We will perform ongoing selection during the application period. We look forward to hearing from you!
Please note that due to GDPR regulations we can only accept applications sent through the recruitment system, not via email or other channels.