Evolved Machines has activated a 10,136-core parallel computing facility dedicated to the development of artificial neural circuits. The system incorporates 42 NVIDIA Tesla GPU processors, controlled by 14 quad-core AMD Phenom microprocessors, and has a nominal computation capacity of over 40 TFLOPS.
"The complex mechanisms embedded in biological neural circuitry enable the extraordinary capabilities of the brain," said Paul A. Rhodes, CEO of Evolved Machines. "The simulation of these mechanisms requires an enormous amount of parallel computing power and we get that from NVIDIA's GPU Computing technology. With this new GPU-enabled facility, we expect our work to be able to guide the development of artificial neural circuits which will lie at the heart of a new generation of devices in artificial olfaction and vision."
"Evolved Machines was one of the first companies to recognize the enormous potential of the GPU to solve large complex problems," said NVIDIA"s Andy Keane. "There is perhaps no greater challenge than creating a machine that behaves like a human brain, so we are excited to see what advances in neural science Evolved Machines can make with their new facility."