At the 2017 Open Compute Project U.S. Summit on Wednesday, Microsoft discussed details regarding Project Olympus. This is the company’s upcoming hyperscale cloud hardware designer. Microsoft has created a hardware standard that can be adopted when buying equipment that will be compatible with Azure.
Microsoft has been increasingly embracing cloud computing as the idea of a connected world grows. By creating its own hardware standard through Project Olympus, the company wants to reduce the cost of creating platforms on Azure.
You may remember Microsoft discussed the project last year. Kushagra Vaid GM, Azure Hardware Infrastructure said the company will share its cloud hardware designs when they are around 50% complete.
Project Olympus consists of 1U and 2U server chassis, power supplies, power distribution, a universal motherboard and other components.
Microsoft worked with the Open Compute Project (OCP) to create the hardware standard. The company says Project Olympus brings quality server hardware designs to developers earlier in cycles. As an open source, dev’s can modify, fork, and change the platform.
“This has enabled bootstrapping a diverse and broad ecosystem for Project Olympus, making it the de facto open source cloud hardware design for the next generation of scale computing.,” Microsoft says in a blog post.
Silicon Partnerships for Project Olympus
Microsoft says the new standard is the first OCP solution to offer diverse microprocessor selections. Long-time partner Intel is contributing its Xeon (Skylake) Processors with Olympus support.
Yesterday, Microsoft also confirmed it will embrace ARM for the first time with Windows Server. The company will use Qualcomm’s chips to run ARM in its own datacenters. AMD has also committed its upcoming ‘Naples’ chips to the project.
Finally, Microsoft also announced a partnership with NVIDIA. The silicon maker will bring its HGX-1 hyperscale GPU accelerator to Project Olympus.