The OPRECOM project explores new computing strategies based on transprecision analytics. Transprecision analytics comprises of exploiting approximation in hardware and software, from both a statistical and a deterministic viewpoint. This is used namely in approximate computing, which nevertheless lacks precision in results accuracy. OPRECOM wants to overcome this shortcoming, and propose a method, called transprecision computing, which leads to signiﬁcant energy savings without sacriﬁcing overall quality of results.
This unique approach can broaden the range of application scenarios in the domains of Internet of things, Big Data Analytics, Deep Learning, and High-performance computing simulation. The project has its domain specifics, as the project coordinator, Dr. Cristiano Malossi, mentioned in the FETFX article:
“One challenge is that transprecision computing is not widely known – even among computer scientists. Automation is key to helping an open international community to grow, so that transprecision computing can be accessible to a broad audience, including engineers who do not have expertise in approximation and inexact computations. In this respect, OPRECOMP is developing a transprecision software development kit that enables developers to easily program and experiment with transprecision algorithms and small computing devices, such as the PULP. The roadmap to see this technology in everyday applications is still long, but with OPRECOMP we have made a big step forward.”
Within the last 4 years, OPRECOMP has been working on algorithms based on transprecision computing, new low-energy platforms for executing large workload, software environment libraries to enable use of transprecision computing, as well as emulation libraries that accelerate prototyping and development.
As the project ends in December 2020, many outcomes have been already delivered. One of them is FloatX (Float eXtended), an open-source C++ programming library for low-precision computing, which provides a framework to investigate the effect of leveraging customized floating-point formats in numerical applications.
- For more information about FloatX and its design principles, please read a research article by the OPRECOMP team.
FET-Open and FET Proactive are now part of the Enhanced European Innovation Council (EIC) Pilot (specifically the Pathfinder), the new home for deep-tech research and innovation in Horizon 2020, the EU funding programme for research and innovation.