The aim of the project is the development of an ultra-fast and compact processor with supercomputer performance and optimized for pattern recognition, data reduction, and fast information extraction for high-quality image processing. The proposed hardware prototype provides sufficient flexibility for applications in a wide range of fields, e.g. triggering in High Energy Physics (HEP) experiments, simulating some human brain functions, as well as automating processing of imaging data in medical physics. In general, any process based on massive pattern recognition could be implemented in our device, provided that input data are adequately prepared and formatted.
Our first goal is to prove the processor can perform online track reconstruction at the Large Hadron Collider (LHC) experiments ATLAS and CMS. We are convinced this is possibile even at the highest instantaneous luminosity expected at super-LHC, in a condition of overwhelming confusion due to the large pile-up and track multiplicity. We are building a real-time tracker, named Fast TracKer (FTK) processor, for ATLAS. We expect FTK will improve the ATLAS trigger capability to select interesting events, for example events with heavy leptons or quarks. Since these events are very rare and hidden in an enormous level of background, this project is extremely ambitious. FTK makes a wide use of FPGAs and ASICs and performs a three-dimensional reconstruction of particle trajectories. On average FTK processes one LHC event in few tens of microseconds and provides online track reconstruction with offline-quality.
We are also studying new industrial applications within an industry-academia cooperation which is going to reinforce our team. This cooperation will increase the transfer of knowledge between academia and industry and increase the capability of the industrial partners to compete on the market.