Disclaimer: this is a guest blog post. The authors are Prof. Cristina Silvano & Prof. William Fornaciari
As our day-to-day expectations of technology increase so does the complexity of the processors required to meet those expectations. The current engineering approach is to increase the number of processing “cores”, individual compute engines, grouped inside a single processor chip.
Current generation smart phones have dual- or quad-core processors combined with graphics accelerators; tablet computers are often more complex; and laptops more still. The trend is for the number of cores on devices in the marketplace to double about every two years. In the near future some systems will have 64 or more cores in order to provide for ever-growing greater complexity. To master the design of such complex systems in such a way as to deliver the performance, using low power consumption and a smaller CO2 footprint is an enormous engineering challenge.
This growth path requires a global rethink about approaches to software and hardware design and it throws up new challenges in software management. The 2PARMA project was set up with support from the European Commission (under FP7-ICT Programme in the domain of Complex Systems and Advanced Computing) specifically in order to address how such systems are managed. Here “management” means balancing and optimising workloads across large numbers of cores and indeed across large numbers of processors as they run our daily lives. This is actually a very difficult problem and it has been the goal of many projects to find suitable strategies for doing so.
2PARMA developed technologies to handle the issues that arise, while maintaining uniform quality-of-service. They have been tested in a variety of industrial scenarios using a wide variety of hardware. While the outcomes do not solve all the issues that arise, 2PARMAs technologies represent a major breakthrough, enough of one to make the results compelling for its industrial participants.
Close cooperation between industry and academia has resulted in faster technology transfer, providing the industrial partner with faster times-to-market for new, more sophisticated domestic products, strengthening their global competitivity. It has also laid the base for sustainable privately-funded research infrastructures.
The public availability of the results of 2PARMA will impact both industrial and academic embedded systems communities. A start-up company (Rubber Duck Software P.C.) has been recently founded based on the needs that the 2PARMA project raises and others are being actively considered. Rubber Duck will exploit the 2PARMA project results and use them to create complex, real-life applications on innovative, commercial heterogeneous platforms. Other companies are already integrating the results of the project into real-world applications which will be available in the market soon.
The project has been presented as one of the success story during the Conference: "Cyber-Physical Systems: Uplifting Europe's innovation capacity", 29-30 October 2013 in Brussels.