Jason P. Waxman, Corporate Vice President in the Data Center Group and Data Centric Chief Strategy Officer (CSO)
As a founding member of the Open Compute Project (OCP) since its beginning 8 years ago, I’ve witnessed the tremendous growth in the community. Today, the OCP ecosystem has reached over $2.56B1 in hardware revenue annually (110% growth rate vs. 2017) by addressing the need for power efficient hardware, scalable systems and racks, and standards-based designs. Across OCP, there is a broad array of open designs for compute, network and storage. Building upon our history of over 80 OCP-inspired products with partners, I’m excited to share the newest innovations using Intel technology for the OCP community during my keynote at OCP’s Global Summit.
Together with our partners, Intel is designing platforms optimized for cloud workloads. As cloud service providers strive for greater density, the OCP community will share a new four-socket High-Density Cloud-Optimized system design with next-generation Intel® Xeon® Scalable processors (code-named Cascade Lake) providing up to 112 high-performance cores in a single 2U platform. New platform features include Intel® Optane™ DC persistent memory, which delivers memory-like performance with the permanent data storage of disks, and Intel® Deep Learning Boost with Vector Neural Network Instructions (VNNI).
Faster Network and Interconnects
The OCP NIC 3.0 specification enables the industry to deliver improvements in serviceability, power delivery, and management using Intel network adapter products ranging from 1GbE to 100GbE. Intel’s new OCP NIC 3.0 adapter products will start shipping in Q3’19. 400G pluggable optics will support cloud and AI growth drivers and in the future, co-packaging of the optical interfaces with the switch ASIC will help reduce power and total cost of ownership.
In addition, Intel and a consortium of eight other technology leaders this week announced the first-generation of Compute Express Link (CXL), a new CPU-to-device interconnect standard for high-performance accelerators and platform enhancements that will work side-by-side with microprocessors. This new interconnect enables a coherent memory space between the CPU and the device, high-performance data transfer, and uses the well-understood physical and electrical infrastructure of PCI Express. Intel believes that there is an opportunity to optimize form factors with OCP and we expect platform innovations, harnessing the capabilities of CXL, as early as 2021.
Open Hardware Management with Intel® Rack Scale Design Rack Management Module
Customers need software and fully-integrated solutions to ease cloud infrastructure management. This year, we’re taking further steps to address these needs. This starts with open standards for firmware, telemetry, virtualization, orchestration, and application development. To fuel development, I’m pleased to announce we are contributing our rack management module code to OCP. This represents a significant body of work, over a year in the making, which extends software defined infrastructure and hardware resource pooling from power and cooling to NVMe over fabric pooling with the Storage Networking Industry Association (SNIA) Swordfish* standard-based storage management. We aligned efforts to drive simple, common standards for BIOS, baseboard management controller (BMC), and rack management software through our engagement in the OCP system, OpenBMC, and OpenRMC firmware projects. Intel’s has a long-standing commitment to open source that extends to other communities such as Openstack, Cloud Native Computing Foundation, Kubernetes, Linux Foundation, Distributed Management Task Force (DMTF), SNIA, and others.
Security has come even more to the forefront, and over the past year, we’ve worked closely with Project Cerberus*, but we need go even further. That is why I am pleased to announce that Intel, Microsoft, Facebook and other partners are teaming up on an open security specification built upon Cerberus-v1 and Intel® PFR (Platform Firmware Resilience). Future root of trust (RoT) capabilities will include in-silicon RoT, secure key measurement/storage, and advanced key management. Microsoft’s next generation of the Mount Olympus platform, powered by next-generation Intel Xeon Scalable processors, will introduce in-silicon support of Cerberus Root of Trust. Early ship deployments have begun, and we will have OCP platforms for broad production launching soon. This collaborative effort is open for other partners to contribute and join in through the OCP security incubation project at https://www.opencompute.org/projects/security.
The rising demands of cloud, network, connectivity, and AI opens new needs for hardware innovation. We need to be more collaborative and open than ever before. Join us in accelerating these efforts within the OCP community, together we can build a better data center for all.
Find details on our latest OCP contributions at https://www.opencompute.org/contributions and stop by the Intel booth if you’re attending the OCP Global Summit in San Jose. For all the latest news and announcements from the summit follow #OCPSummit19 on Twitter.
Statements in this document that refer to Intel’s plans and expectations for the quarter, the year, and the future, are forward-looking statements that involve a number of risks and uncertainties. A detailed discussion of the factors that could affect Intel’s results and plans is included in Intel’s SEC filings, including the annual report on Form 10-K.
Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No product or component can be absolutely secure.
*Other names and brands may be claimed as the property of others.