Microsoft Taps Intel Skylake Chips for Project Olympus
- By Kurt Mackie
- July 12, 2017
Microsoft this week said it plans to make Intel's newly released Xeon Scalable processors "a cornerstone" of Project Olympus, its open source datacenter hardware effort.
Generally available starting this week, Intel Xeon Scalable ("Skylake") processors are designed for use in datacenters, with support for up to 28 cores and 6TB of system memory. Intel sees these processors as being used for real-time analytics, high-performance computing (HPC), 5G networks and artificial intelligence kinds of applications, including automated driving.
AT&T is currently using the processors to manage communications traffic, according to an Intel announcement on Tuesday. Intel has already sold "more than 500,000 Intel Xeon Scalable processors" to enterprise, HPC and communication service providers, the company indicated.
Intel Xeon Scalable processors represent "the biggest datacenter platform advancement in a decade," according to Navin Shenoy, Intel's executive vice president and general manager of the Data Center Group, in video comments.
The new microprocessor, which was developed over five years, was completely re-architected for the datacenter and the hybrid cloud. It has scalable performance improvements (1.65x over the last-generation processor), pervasive security (2x data encryption protection) and better agility to offer new services and capabilities (4.2x virtual machine capacity), Shenoy added.
Intel also announced its Intel Select Solutions program, a partner effort based on Intel Xeon Scalable processors that aims to simplify and speed datacenter infrastructure deployments. Partners under the program include Canonical (Ubuntu), Microsoft (SQL Server 2016) and VMware (vSAN 6.6).
Microsoft also chimed in on the announcement, indicating that Intel Xeon Scalable processors are supported in Project Olympus, which is Microsoft's effort under the Open Compute Project to jumpstart the agile development of "open source hardware" for datacenters. In March, Microsoft had indicated it was working with Intel on integrating Xeon processors under Project Olympus, along with other collaborative efforts involving Nvidia, Ingrasys Technology, Cavium and Qualcomm.
Project Olympus hardware will support various "extremely demanding" datacenter applications, ranging from e-mail messaging to HPC and artificial intelligence, according to Microsoft. Intel Xeon Scalable processors will help meet those requirements.
"Whether it's high core counts and memory bandwidth for extreme multithreaded performance, IO scaling capabilities, or the new Intel AVX-512 instructions for HPC and AI workloads, Intel Xeon Scalable Processors, and Intel FPGAs provide a significant degree of flexibility and performance that allows us to meet the emerging demands of the cloud," Microsoft indicated.
Project Olympus hardware specifications get released as open source contributions to the Open Compute Project via a GitHub repository. The Open Compute Project is an industry coalition that fosters the use of open datacenter technologies. It was originally founded by Facebook, Intel and Rackspace.
Kurt Mackie is senior news producer for the 1105 Enterprise Computing Group.