Open Source Hardware

For the third research focus, expertise in the area of new processor architectures and open exchange formats is being built up. It was recently announced in the press that Intel will be setting up a special infrastructure near Magdeburg for the production of semiconductors in the 5nm structure size range. The company also has extensive expertise in the design of multi- and many-core systems, including from the successful completion of the DFG SFB/TRR 89 („Invasives Computing“), which investigates how 1000 and more processors can be integrated, managed and programmed on one chip at all levels of hardware and software.

Many industrial companies are very concerned that the dependence on hardware components in their products from primarily American and Asian companies will lead to strong dependencies. As a result, there is a growing demand for national and European processor and system developments that are not proprietary or subject to complex intellectual property rights. One example of this is the RISC-V processor initiative. In order for Germany and Europe as a whole to be able to develop hardware components, especially complete processor systems, themselves in future or to be able to use and market existing hardware more easily in their products, it is important to investigate promising alternatives in the direction of open source hardware. While open source is currently the norm rather than the exception in the software sector, even in the area of embedded systems, the opposite is true for hardware components.

Important research questions here concern formats for the reusable exchange of so-called IP (intellectual property) blocks, questions of the portability of such modules between tools, questions of licensing as well as the exploration of new business models and marketing despite or precisely because of open source hardware. Conversely, there are questions about the safe use of freely available hardware, particularly with regard to security. This concerns, for example, questions of protection against attacks if the hardware is no longer an anchor of trust per se, but how downloaded hardware modules can be checked for trustworthiness, freedom from side-channel attacks or Trojans. This also applies to checking corresponding synthesis tools for trustworthiness. It is also necessary to investigate how libraries of open source hardware can be set up and how suitable modules can be searched for and found in them. The question also arises as to how OSH modules can be automatically integrated into synthesis flows and design tools.