Powered by custom optimized chips that are joined together in the Microsoft servers to translate the whole of Wikipedia in literally less than a second. In the Microsoft’s Ignite conference in Orlando, the company tapped what is known as the global hyper-scale cloud to translate 3 billion words through 5 million articles in less than a tenth of a second. The FPGA (Field Programmable Gate Array) or programmable logic component has been custom designed by the company and added to every computing nodes.
More computing power will be required by the smarter and more intensive computational technology at the back end. These can be utilized in automobiles for recognizing a distracted driver, in Cortana to intelligently reschedule your workout etc. in order to perform a particular task faster an FPGA can be optimized or tuned according to that task. According to the company while the FPGA boards were compared to the four high-end CPU the FPGA board consumed much less power about five times. The new computing infrastructure is designed to empower the new AI initiative including the company’s bot framework and of course Cortana. The approach to democratizing and create AI is been applied just as the same approach used to create access to data.