Architecting the Cloud Infrastructure for the Future with Intel
2 likes1,610 views
The document discusses the future of cloud infrastructure, emphasizing the transformation toward workload-optimized technologies and software-defined infrastructure. It highlights the collaboration between Intel and Nuance to enhance speech recognition and user experiences through powerful computing capabilities in the cloud. Additionally, it outlines the need for composable resources and intelligent orchestration to meet evolving demands in data centers and optimize performance.
Architecting the Cloud Infrastructure for the Future with Intel
- 1. Architecting Cloud Infrastructure for the Future Jason Waxman VP and General Manager Cloud Platforms Group Data Center and Connected Systems Group July 22, 2013
- 2. Experiences Today and Tomorrow Predictive Analytics Improve healthcare Reduce car/aircraft parts failure Video/Media Content Delivery Video Search Voice & Gestures Personal assistant Natural Interaction 20X growth in speech driven mobile network traffic1 >22X increase in smartphones with gesture recognition features1 16X increase in mobile video traffic2 4X increase in servers for media / graphics3 1. ABI Research Aug 2012 2. Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2012–2017 3. Intel internal analysis and forecast, 2013 4. IDC Storage Solutions, 2013, doc #241515, May 2013. 43% CAGR for infrastructure supporting Big Data & Analytics4 ‘12-’17 ‘12-’17 ‘11-’16 ‘13-’17
- 3. Sean Brown Senior Manager, Innovation
- 4. © 2002-2013 Nuance Communications, Inc. All rights reserved. Page 4 6 billion mobile handsets 70+ languages 65+ countries An Inflection Point in Speech Revolutionary shift from task-centric to user-centric Intelligent Systems Speech Recognition Understand Inform Be Aware Ubiquitous Command & Control Dictation Search Natural Language
- 5. © 2002-2013 Nuance Communications, Inc. All rights reserved. Page 5 Delivering an Optimal UX Globally Only the cloud can deliver the compute power required New FeaturesSpeedAccuracy Languages Users Transactions Data Platforms Models
- 6. © 2002-2013 Nuance Communications, Inc. All rights reserved. Page 6 Intel & Nuance Collaboration The Nuance Cloud, powered by Intel Nuance Data Centers Powered by Intel Xeon Processors Hardware + Software Accelerators Balanced Computing Device and cloud data centers leveraged for optimal performance Joint Optimizations on Intel-based Platforms Scale Natural Language Processing
- 7. Delivering Great User Experiences What’s Needed?
- 8. Architecting Cloud Infrastructure for the Future Addressing Requirements Workload optimized technologies Composable Resources Software defined infrastructure
- 9. Architecting Cloud Infrastructure for the Future Addressing Requirements Workload optimized technologies Composable Resources Software defined infrastructure
- 10. Architecting Cloud Infrastructure for the Future Workload Optimized Systems DIVERSITY OF REQUIREMENTS NetworkingStorage Cold Warm Hot Cost of storage$ $$$$ Access Frequent Infrequent Servers I/O Intensive CPU & Memory Intensive I/O Intensive CPU Intensive Wireless Base Station Media Processing Enterprise Security & Routing Wireless Core Branch Office Router E- Commerce Dedicated Hosting Enterprise Applications High Performance Computing Big DataContent Delivery and Gaming Graphics Rendering Cold Storage Low End Networking Edge Routin g Storage De-dupe Cloud RAN Small Cell SMB Security Appliance
- 11. Architecting Cloud Infrastructure for the Future Workload Optimized Systems: Intel Developments CUSTOMIZED SILICON ACCELERATORS EXTREME LOW POWER *Other names and brands may be claimed as the property of others.
- 12. SLM Core PCIe Gen2 16 Lanes , 4 Controllers DDR3/3L-1600 Crypto Accelerator 1MB L2 Cache SLM Core SLM Core 1MB L2 Cache SLM Core SLM Core 1MB L2 Cache SLM Core SLM Core 1MB L2 Cache SLM Core 4x GbE 4x USB 2.0 Legacy I/O 2x SATA 3.0 4x SATA 2.0 Highly Scalable Up to 8 cores with integrated I/O Higher Efficiency Higher Performance Up to 7x faster1, 2 Up to 4x higher performance per watt1, 3 Intel® Atom™ Processor C2000 Product Family 2nd Generation 64 bit Workload Optimized SoCs “Avoton” & “Rangeley” Workload Optimized 8x (64GB) Memory capacity1 Intel® QuickAssist Technology IA Software Compatibility 2.5X increase in system designs for microservers, comms and storage Datacenter Class Features 64-bit, ECC memory, Intel® Virtualization Tech 2H’13 1vs Intel® AtomTM S2100. Intel Atom C2000 pre-production silicon measurements SLM= Next Generation Atom® micro architecture “Silvermont” Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products.. For more information go to http://www.intel.com/performance. Copyright © 2013, Intel Corporation. Configuration: 2Dynamic Web Benchmark: Atom S1260(8GB,SSD,1GbE), Score=1522. Atom C2xxx(32GB, SSD,10GbE), Score=11109. 3SPECint*_rate2006 for Atom S1260(8GB, HDD) 18.7, est pwr=20W, Atom C2xxx(16GB, HDD), 69, est node pwr=19 Intel Internal measurements as of July 2013. Refer to backup for additional details. * Other names and brands may be claimed as the property of others.
- 13. Architecting Cloud Infrastructure for the Future Addressing Requirements Workload optimized technologies Composable Resources Software defined infrastructure
- 14. Today FutureNext Physical Aggregation Shared Power Shared Cooling Rack Management Fabric Integration Rack Fabric Optical Interconnects Modular refresh Subsystem Aggregation Deliver efficient, adaptable data center infrastructure IO Compute Memory Storage Pooled compute Pooled storage Pooled memory Shared boot Architecting Cloud Infrastructure for the Future Flexible Rack Scale Infrastructure
- 15. Mark Roenigk Chief Operating Officer
- 16. RACKSPACE OPEN JOURNEY INTEL DATA CENTER DAY
- 17. 205,000+ CUSTOMERS 100,000+ SERVERS 26,000+ VM ≅70 PB STORED 10 WORLDWIDE DATA CENTERS GLOBAL FOOTPRINT CUSTOMERS IN 120+ COUNTRIES About Rackspace
- 18. We’re on the “Open Journey”
- 19. The Rackspace OpenRack Server Using Intel technologies for servers and networking
- 20. • Take advantage of greater VM density • Improved revenue per WATT • Great partnership • Accelerated time to market • A strong product roadmap • OCP deployment will include Intel 10G implementation Rackspace Innovating Faster with Intel® Xeon® family
- 21. Flexible & Cost Effective Data Tiering for optimized performance Extreme network bandwidth Platform flexibility Storage-PCIe-SSD & Caching Photonics & Switch Fabric Silicon: Intel® Atom™ & Xeon Open Network Platform CPU / Mem Modules Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or configuration may affect actual performance. 1. Improvement based on standard rack with 40 DP servers, 48 port ToR switch, 1GE downlink/server and 4 x10GE uplinks, Cables: 40 downlink and 4 uplink vs . rack with 42 DP servers, SiPh patch panel, 25Gb/s downlink, 100Gb/s uplink, , Cables: 14 optical downlink, and 1 optical uplink. Actual improvement will vary depending on configuration and actual implementation. 2. Improvement as compared to 20 Dell PowerEdge R720, N+1 redundant power, 705W PSU x2, peak power provisioned 30,000 Watts vs. same server, shared DC power using 1 power shelf of 7x 700W modules and 4200W (N+1) : power provisioned 4900 Watts http://www.opencompute.org/wp/wp-content/uploads/2013/01/Open_Compute_Project_Power_Shelf_v0.3.pdf, http://www.spec.org/power_ssj2008/results/res2012q4/power_ssj2008-20121030-00569.html .. Software Defined Network Orchestration Cable Reduction Network uplink Improvement Network downlink Improvement Density servers/rack Improvement Up to 1.5X 1 Power provisioning Reduction Up to 6X 2 Up to 3X 1 Up to 2.5X 1 Up to 25X 1 Architecting Cloud Infrastructure for the Future Intel Rack Scale Architecture Innovation BENEFITS
- 22. Architecting Cloud Infrastructure for the Future Addressing Requirements Workload optimized technologies Composable Resources Software defined infrastructure
- 23. Architecting Cloud Infrastructure for the Future Software Defined Infrastructure TODAY’S CHALLENGES • Pressure to meet SLAs • Overprovisioning of resources • Inefficient workload placement IO Compute Memory Storage ORCHESTRATION SOFTWARE FUTURE: INTELLIGENT ORCHESTRATION powerthermals performance utilization security
- 24. Architecting Cloud Infrastructure for the Future Evolving to Software Defined Networks Discrete Compute Storage Networking ControlControlControl Management Virtualized Control Control Management Control Resource Pools Software Defined Orchestration Control Software Defined Networks Centrally ManagedUnified Network Virtualized FunctionsFixed Function Vertically Managed
- 25. Architecting Cloud Infrastructure for the Future Intel: Accelerating Software Defined Networks Intel Open Network Platform Server Reference Design Deliver Open Building Blocks Deliver Open Software Open source Open network OS with open APIs Use Open Standards SDN Controller SW HW Intel NIC Intel Product 3rd PartyWind River Product OEM App SP App ISV App Open Network Platform Software Intel Communications Chipset Intel® Architecture
- 26. Architecting Cloud Infrastructure for the Future Growth Opportunities Public Cloud Servers 25% $15B 1 Distributed Storage 50% $21B2 SW-Defined Networking 175% $5.5B3 By 2016 CAGR 1. Intel forecast, 2013 2. IDC Storage Solutions, 2013, doc #241515, May 2013 3. SDN Market Sizing Plexxi, Lightspeed Venture Partners, SDN Central April 2013; Markets & Markets - 2013 ‘16 Revenues
- 27. Intel® CPUs & Chipsets SOCs & Accelerators Intel 10GbE Intel Switch Silicon Intel Silicon Photonics Flexible building blocks Intel NVM & SSD Network StorageServer Software Defined Networking Rack Scale Architecture Delivering Great User Experiences Cache Acceleration Software