Facebook has announced plans to open-source its artificial intelligence hardware design. The Open Rack-compatible hardware, codenamed Big Sur, is designed to handle AI computing at a large scale, according to the company.
“At Facebook, we’ve made great progress thus far with off-the-shelf infrastructure components and design,” wrote Facebook developers Kevin Lee and Serkan Piantino in a blog post. “We’ve developed software that can read stories, answer questions about scenes, play games and even learn unspecified tasks through observing some examples. But we realized that truly tackling these problems at scale would require us to design our own systems.”
(Related: How to take an intelligent approach to AI)
Facebook hopes open-sourcing the hardware will make it easier for AI researchers to share their techniques and technologies, and improve on the system. In addition to open-sourcing Big Sur, Facebook plans to submit its design materials to the Open Compute Project, a collaborative initiative to deliver “efficient server, storage and datacenter hardware designs for scalable computing,” according to the project’s mission statement.
“We believe that this open collaboration helps foster innovation for future designs, putting us all one step closer to building complex AI systems that bring this kind of innovation to our users and, ultimately, help us build a more open and connected world,” the company wrote.
Big Sur leverages NVIDIA’s Tesla Accelerated Computing Platform, and features eight high-performance GPUs, improved performance from the company’s previous generation, and optimized servers for thermal and power efficiency.