Text

Heterogeneous systems - hardware software co-design

The group aims to boost exploitation of heterogeneous systems in terms of predictability, effective development and efficient software-hardware integration for next-generation intelligent embedded systems.

Contact

Senior Lecturer

Masoud Daneshtalab

+4621103111

masoud.daneshtalab@mdh.se

Senior Lecturer

Saad Mubeen

+4621103191

saad.mubeen@mdh.se

With the exploding need for high-performance computing, we are at the dawn of the heterogeneous era, where all future computing platforms are likely to embrace heterogeneity. In a heterogeneous system, there can be several different computational units such as multi-core central processing units (CPUs), graphics processing units (GPUs), field-programmable gate arrays (FPGAs), digital signal processing units (DSPs), and artificial intelligence (AI) accelerators/engines.

One major driving force for heterogeneous systems is the next generation intelligent, adaptive and autonomous systems that will form the base for coming products like autonomous vehicles and autonomous manufacturing.

With a diverse range of architectures (on a single chip or distributed), a main challenge is to make use of the enormous computational power in the best way, while still meeting several criteria like performance, energy efficiency, time predictability, and dependability.

 

Overall goals

The overall goal of this research group is to tackle the following scientific areas:

• Hardware/software co-design and integration.

• System architecture and specialization.

• AI and deep learning acceleration.

• Model-based development of predictable software architectures.

• Pre-runtime analysis of heterogeneous embedded systems.

Ongoing research projects

This project aims at developing innovative techniques to provide a full-fledged development environment for vehicular applications that use TSN as the backbone for on-board communication.


Project manager at MDH: Saad Mubeen

Main financing: Vinnova

DeepMaker aims to provide a framework to generate synthesizable accelerators of Deep Neural Networks (DNNs) that can be used for different FPGA fabrics.


Project manager at MDH: Masoud Daneshtalab

Main financing: The Knowledge foundation

In this joint project, we aim at decreasing the power consumption and computation load of the current image processing platform by employing the concept of computation reuse.


Project manager at MDH: Masoud Daneshtalab

Main financing: STINT - The Swedish Foundation for International Cooperation in Research and Higher Education

The overall goal of HERO is to provide a framework that enables development of optimized parallel software, automatic mapping of software to heterogeneous hardware platforms, and provision of automatic hardware acceleration for the developed software.


Project manager at MDH: Mikael Sjödin

Main financing: The Knowledge Foundation

This project addresses design methods for the use of DNNs in airborne safety-critical systems.


Project manager at MDH: Håkan Forsberg

Main financing: Vinnova