Intel Corporation Hardware Accelerator Research Program (HARP) Xeon+FPGA machine
This program provides access to a prototype 10 node machine located at the Paderborn Center for Parallel Computing (PC²). Each node contains a second generation Intel Xeon+Arria 10 FPGA.
Intel and Xilinx FPGA development boards
CEL has a number of Intel and Xilinx development boards for embedded and server applications.
Software defined radio
Research in applying machine learning to radio frequency signal is being undertaken using software defined radios, such as the Ettus X310.
Nvidia graphics processing units
This is a machine for training deep neural networks, powered by NVIDIA M40 Maxwell architecture graphics processing units.