OPen TransPREcision COMPuting (OPRECOMP) European project
OPRECOMP is one of the EU research projects funded under the EU Framework Horizon 2020 – Re-Program Future and Emerging Technologies (FET). It aims to replace ultra-conservative computing technologies with a more flexible and efficient one, namely transprecision computing. The objective is to develop a complete framework for transprecision computing, covering devices, circuits, software tools, and algorithms, along with the mathematical theory and physical foundations. The application scenarios will cover the domains of IoT, Big Data Analytics, Deep Learning, and HPC simulations: from the sub-milliWatt to the MegaWatt range.
GreenWaves Technologies contributes to OPRECOMP project by designing disruptive ultra-low power embedded solutions for image, sound and vibration AI processing in sensing devices. GreenWaves was founded in late 2014 with the mission of revolutionizing the market for intelligent sensors and devices with ultra-low energy and cost-efficient solutions. GreenWaves’ pioneering GAP8 IoT application processor is uniquely optimized to execute a large spectrum of image and audio algorithms including convolutional neural network (CNN), with extreme energy efficiency, thanks to an integrated 8-core computational cluster combined with a convolution hardware accelerator. Vector units, intrinsics and proprietary instruction set extensions help the 8 cluster cores execute faster than traditional parallel processor architecture without writing a single line of assembly code. Fully programmable, RISC-V based PULP parallel architecture enables to cope with the rapidly evolving state-of-the-art in machine learning algorithms. Next-generation of GreenWaves’ processors will include a transprecision unit and will enable even more power efficiency.