New open source platform VerSys allows early software tests via virtual chips

Open source is on the rise: Alongside open software, new hardware architectures offer an independent and cost-effective alternative to large chip manufacturers. In order to be able to write and test software for the processors of the future even before their production, the German Research Institute for Artificial Intelligence (DFKI) is working on a new verification platform. This is good news not only for the German industry, but also for start-ups. The project VerSys is funded by the Federal Ministry for Education and Research (BMBF) with roughly 1.6 million Euros.

Smart homes, smartphones, self-driving cars and the industry of tomorrow – all of these future technologies require powerful computer chips for their control. The systems running on these chips are usually developed and sold as closed systems on chip (SoC). However, the wind is changing: Next to open software like the operating system Linux, open source architectures for hardware are also gaining in popularity. One of the forerunners is the RISC-V processor (Reduced Instruction Set Computer), which can be designed, produced and sold freely via an open source licence.

The open instruction set architecture offers many advantages: Apart from the non-existent patent costs, RISC-V also persuades with its standardisation, expandability and producer independency. This is why the processors are assumed to be particularly future-proof. But the question is: How can the development of software for the chips of the future start without having to wait for their production? As an answer to this, the research department for Cyber-Physical Systems of the DFKI led by Prof. Dr. Rolf Drechsler started working on an open verification platform for the testing of software via a virtual RISC-V chip on 1 August 2019.

Open source platform aims at speeding up software development

The verification of software via virtual prototypes for computer chips is a standard in the industry. Nonetheless, this prototype is still missing for the RISC-V architecture. This is where the project VerSys (short for “Consistent Verification Platform for Early Software Development for RISC-V Based Systems”) is creating a remedy. The goal is an open source verification platform that is scalable, modular and correct and can easily be adapted to the requirements of the users.

The project VerSys is funded by the Federal Ministry of Education and Research (BMBF) with roughly 1.6 million Euros for a runtime of three years. The resulting platform will subsequently be enhanced and made industrially implementable with the help of partners from the industry. After the project runtime, the DFKI research department for Cyber-Physical Systems hopes for an ecosystem of smaller companies that develop around the platform and offer different kinds of services relating to the open source architecture.

Industry and start-ups profit from early software verification

RISC-V chips and their software are used in the areas of automotive, communications and production as well as in smart homes and the internet of things (IoT). The need for a verification platform for early software tests such as the project VerSys has already been proven by lively interest shown by the industry. Furthermore, start-ups can advance their software development faster, more independently and with less risk by using the VerSys platform.

The hardware architecture RISC-V was developed and introduced in 2010 at the University of California in Berkley, US. In order to establish and publish software standards, the RISC-V foundation was created. The DFKI research department for Cyber-Physical Systems plans on becoming a member of the foundation as part of the project VerSys.

Photo

A visualisation for the project can be downloaded under https://cloud.dfki.de/owncloud/index.php/s/rn2Qq5reGZkpcrz Please name the source “DFKI GmbH. Illustration: Lisa Jungmann”.

DFKI Contact:

Dr.-Ing. Daniel Große
German Research Center for Artificial Intelligence (DFKI)
Cyber-Physical Systems

Phone: +49 421 421 218 63935

Prof. Dr. Christoph Lüth
German Research Center for Artificial Intelligence (DFKI)
Cyber-Physical Systems

Phone: +49 421 218 59830

DFKI Press Contact

German Research Center for Artificial Intelligence (DFKI)
Corporate Communications Bremen

Phone: +49 421 178 45 4180

Share this post:

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz