Petr Ulvr, Business Development Manager Intel Corp
[email protected]
Digitální služby transformují trhy
2
Ekonomika digitálních služeb přichází s novými požadavky na datová centra Škálovatelnost
Efektivní, levný provoz Konzistentní, spolehlivý provoz 3
Průzkum ukázal, že v 51 % v dotazovaných firmách průměrné stáří serverů překračuje 4 roky, z toho více než v pětině firem spravují jejich data servery starší než pět let. Intel doporučuje zachovat tříletý cyklus obměny firemního hardwaru. Pevně nastavený cyklus v intervalu tři roky a méně má jen 8 % firem. Ve srovnání s výsledky před dvěma roky, kdy to byla pouhá 2 %, je to jen mírný posun k lepšímu.
4
Aktuální výzvy pro IT Implementace privátního cloudu
Zvýšení výpočetního výkonu
Zvýšení využití serverů
Analýza dat
Energetická efektivita
Provozní náklady 5
5
Cloud – ANO vs. NE
Impact of Cloud on IT Consumption Models A study, Cisco® Consulting Services, in partnership with Intel
6
Vize otevřeného cloudu Požadavky IT & Otevřené standardy
Optimalizované platformy
Intel® Cloud Builders & Cloud Finder
Definování a prioritizace požadavků IT & Urychlení přijetí otevřených industry standardů
Optimalizované stavební prvky pro bezpečnou, efektivní, automatizovanou platformu postavenou na jednotném základě
Prověřená řešení usnaňující adopci cloudu & výběr cloud service providera 7
What’s Inside Matters
70% Of Enterprises Specifying HW In Service Selection 1
Top Reasons:
PERFORMANCE
End User Benefits:
Revenue Acceleration: Easy to create new apps and services
SECURITY
OpEx Reduction: Automation of the data center
WORKLOAD OPTIMIZATION
CapEx Reduction: Maximize use of system resources
Source: 1) IMR Research: Base: Total using Public Cloud US n=207; PRC n=241, Germany n=90, Brazil n=81 Q. What percent of these applications have you specified the hardware and/or other components to be used by the Cloud service provider as opposed to only specifying the level of service you expect? 2) Intellitrends Survey 2013 - (Comuware Report Measuring the Business Impact of Technology Performance) 3) HotCloud 2012 White Paper 8
8
Široké portfolio řešení na Intel® technologiích Optimální efektivita na standardních průmyslných stavebních základech
Compute
Network
Storage
Intel® Xeon® processor E5v2 product family
Intel® Ethernet CNA X520/X540 family
Intel® architecture based storage and solid state drives DC P3500 / DC P3700
Intel® Xeon® processors
Security
Intel® architecture
Network acceleration
Intel® TXT
Intel® VT-c
Hyper-Threading
Intel® AES-NI
Intel® VT-d
Communications chipset
Turbo boost
Intel® Secure Key
Next gen NVM
Storage accelerators & SoCs
Software
Ecosystem enabling
Storage software
QuickAssist APIs
HP
Storage acceleration libraries (ISA-L)
10Gb ethernet Ethernet switch silicon
Intel® AVX Intel® VT-x
Ecosystem enabling
Software
Open Data Center Alliance
Open Attestation SDK Intel® DCM: KVM gateway Intel® DCM: Node Manager Intel® DCM: Service Assurance Administrator
Intel® Cloud Builders
*Other names and brands may be claimed as the property of others.
Ecosystem enabling
Red Hat Nexenta Intel® Open Network
Plus others…
Cache acceleration software
Intel® Network Builders
9
“Tick-Tock” Model
Sustained Microprocessor Leadership
Current Platform Span
Tick
Tock
45nm Penryn
Nehalem
Tick
Tock
Tick
22nm
32nm Westmere
Sandybridge
Tock
Ivybridge
Haswell
Tick
Tock 14nm
Broadwell
TBD
Tick: Shrink – Introduction of new Semiconductor Process Tock: Innovation in Microarchitecture
Intel Confidential
PUBLIC
Intel® Xeon® Processor E5-2600 v2 Product Family Historical 2S Energy Efficient Performance 12 000
Up to 27x
SPECpower*_ssj2008 (Overall ssj_ops/watt)
Higher is better 10 000 8 000 6 000 4 000 2 000 0 2006 Xeon 5160 (2C, 3GHz, 80W)
Up to
2007 X5272 (2C, 3.4GHz, 80W)
2007 L5335 (4C, 2GHz, 50W)
2008 L5430 (4C, 2.66GHz, 50W)
2009 X5570 (4C, 2.93GHz, 95W)
2010 X5675 (6C, 3.06GHz, 95W)
27x increase in energy efficiency on Intel
®
2012 E5-2660 (8C, 2.2GHz, 95W)
2013 E5-2660 v2 (10C, 2.2GHz, 95W)
2014 E5-2699 v3 (18C, 2.3GHz, 145W)
Xeon® since 2006
Source as of Sept 8 2014: results published or submitted to www.spec.org. 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 Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. *Other names and brands may be claimed as the property of others
11
Intel® Trusted Execution Technologie Trusted Compute Pools
Any Non-Sensitive Workloads
Hypervisor/VMM
Intel TXT for Sensitive Workloads
Hypervisor/VMM
Geo Fencing + Intel TXT for Sensitive Workloads
Hypervisor/VMM
Při každém startu serveru se provede kontrolní součet zaváděného kódu před jeho spuštěním a hodnoty se uloží bezpečně do TPM modulu. Atestační služba následně porovná tyto hodnoty s hodnotami, které byly označeny za důvěryhodné (trusted) při první čisté instalaci serveru a hypervisoru Provozujte důvěrná data na důvěryhodných záklladech
Example: HyTrust policy enforcement with Intel® TXT Virtual administrator
Virtual infrastructure Trusted Intel® TXT
Hosts automatically labeled based on Intel® TXT measurement compared to Known Good Host
Visa’s private cloud deployment with HyTrust
Rules enforce VMs are only allowed on trusted hosts
Achievements:
Eliminates possibility of using VMs on untrusted hosts through policy enforcement
• Meet compliance efforts and able to satisfy audits more effectively. Enhanced reporting of virtualization security controls
*Other names and brands may be claimed as the property of others.
• Consolidation of distributed applications with highly sensitive data
13
Standards – PCIe* SSD in the Data Center
Non-Volatile Memory Solutions Group *Other names and brands may be claimed as the property of others
Intel Confidential - Do Not Forward
Intel® SSDs for the Data Center Feed your Starving Apps Consistent performance providing high minimum IOP performance
Computex Launch Quotes
Intel® SSD Data Center Family Maximize your $/IOPs with consistent performance, data protection, and reliability New
DC P3700 Series PCIe
Stress Free Protection Enterprise class End-to-end data and power loss protection to help ensure data integrity
Fantastic Highest Performance High Endurance
New DC P3600 Series PCIe
for high performance, write intensive applications
New
Non-Volatile Memory Express (NVMe) Lowers memory latency
DC P3500
Higher Performance Mid Endurance for high performance, mixed workloads
Series SATA
DC S3610
Good
PCI Express Interface Designed to deliver amazing speeds
DC S3700
Great
Series PCIe
"The speed advantages of the NVMe interface are now indelibly clear, especially as demonstrated by Intel's new fire breathing Data Center product lines. This new interface removes many of the bottlenecks associated with SATA and AHCI…” The Register
High Performance Standard Endurance for read intensive applications
DC S3500 Series SATA
Series SATA
"Intel's SSD DC P3700 gives us our first look at an NVMe drive, and the results are impressive” AnandTech 16
Intel Confidential — Do Not Forward
Fortville 10GbE / 40GbE Designed for E5 v3 Fortville Differentiation • Next Gen 10 / 40GbE – Architected for next generation virtualized cloud environments • SW Defined Infrastructure - Network Virtualization Offload, Intel® Ethernet Flow director for traffic steering • Network Functions Virtualization – Excellent small packet performance, Intel® DPDK
Proven
Performance
Flexible
Platform validation Intel Drivers 30 years – Just Works
Virtualization: VMDq, SRIOV, Flow Director DPDK Optimized
Converged: LAN / SAN Broad Interfaces: 2x40, 1x40, 4x10, 2x10
1GbE
10GbE Intel Ethernet CNA X540
Intel Ethernet CNA X520
25GbE Intel Ethernet CNA X710
40GbE Intel Ethernet CNA XL710
17
100GbE RSA / Appliance 100GbE NIC Red Rock Canyon Red Rock Canyon
Jednodušší cesta k výběru nové cloud infrastruktury
HP CloudSystem
IBM PureFlex System
VCE vBlock
NetApp FlexPod
CloudSystem Foundation
SmartCloud Entry
vCloud Director Suite Advanced
UCS Director
CloudSystem Enterprise
SmartCloud Orchestration
vCloud Director Suite Enterprise
Cisco Intelligent Automation in Cloud (CIAC)
Up to 3.5X performance boost vs typical 4 year old server based on spec int1
Dell PowerEdge VRTX OpenManage w/ Chassis Management Controller
Up to 2x higher
Up to 4x more VMs/server
Up to 3x more I/O bandwidth
throughput in virtualized environment3
vs. typical 4 year old server2
vs. prior gen.4
1 Source: Typical 4-yr-old’ baseline configuration and score on SPECint*rate_base2006 Benchmark: BL265 using two Intel® Xeon® Processor X5570 (2.93 GHz, 4-core, 8MB L3 cache, 6.4 GT/s, 95W), 48GB memory (12x 4GB 2Rx4 PC3-10600R-9, ECC), 73 GB SAS 10 K RPM, SUSE Linux Enterprise Server 11 (x86_64) SP1, Kernel 2.6.32.12-0.7-default. Compiler version: Intel C++ Compiler XE for applications running on IA-32 Version 12.0.1.116 Build 20101116. Source: http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110215-14599.html as of Feb 2011. Score: 264. ‘New Generation’ new configuration and score on SPECint*rate_base2006 Benchmark: Intel® Server Board S2600CP platform with two Intel® Xeon® Processor E5-2697 v2 (2.7GHz, 12-core, 30MB L3 cache, 8.0GT/s, 130W, C0-stepping) , EIST Enabled, Turbo Boost enabled, Hyper-Threading Enabled, 128GB memory (8x16GB DDR3-1866), Red Hat* Enterprise Linux Server 6.3. CPU2006-1.2 with Intel compiler IC13.1 Source Intel TR#1270 as of 11 June 2013. Estimated Score: 906. 2 Source: Intel estimation as of September 2013. See next page for configuration data and assumptions. 3 Source: Intel Internal Estimates. IOPS comparison between Intel SSD and HDD. 4 Source: Intel internal measurements of average time for an I/O device read to local system memory under idle conditions comparing Intel® Xeon® processor E5-2600 product family (230 ns) vs. 5500 series (340 ns). 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
18
www.IntelCloudFinder.com 19
[email protected]
