ZDNet’s Mary Jo Foley has published a new article detailing Microsoft’s plans to make field-programmable gate array (FPGA) hardware available to users of its Azure cloud computing service. FPGA chips can be custom-configured after manufacture. At the Build developer conference, Azure CTO Mark Russinovich said described the FPGA technology as “the core of an AI cloud” and “a major step toward democratizing AI.”
This week, Microsoft officials said they plan to make hardware microservices and FPGAs widely available to cloud customers in 2018. The FPGA technology is also playing a role in the company’s Brainwave deep neural net project.