This document discusses the benefits of upgrading from 1GbE networking to 10GbE networking. It notes that server virtualization has led to more virtual machines being deployed than physical servers, straining 1GbE networks. 10GbE enables higher performance by allowing all network, storage and LAN traffic to flow over a single Ethernet connection. It also simplifies networking. The document promotes Intel's 10GbE solutions, noting they are validated on Intel Xeon processors and support features important for virtualized environments like VMQ and SR-IOV. It positions 10GbE as essential for the future of server density, cloud computing and big data workloads.
This document discusses new hardware features including CAT, COD, and Haswell, as well as network platforms. It provides an overview of run-to-completion and pipeline software models for network processing. Run-to-completion allows I/O and application work to be handled on a single core, while pipeline distributes packets to other cores for application work. Lockless queues are used to share data between cores and threads. Rings are the primary mechanism to move data between software units and I/O sources in DPDK.
This document discusses the DPDK roadmap. Version 2.1 is scheduled for release at the end of July 2015 and will include features like dynamic memory management, cuckoo hashing, packet framework enhancements, and driver support for additional NICs. Version 2.2 is scheduled for November 2015 and aims to support bifurcated drivers assuming their availability in the Linux kernel. The document also provides information on getting involved with the DPDK community through the dpdk.org website.
The document discusses Intel's Network Builders program, which aims to accelerate software-defined infrastructure adoption through open standards and platforms. It does this by investing in strong ecosystems, committing to open source, and leveraging Intel's technology leadership. The program enables partners through technical resources, matchmaking opportunities, and marketing support. It also works with network operators on proofs of concept and trials. The goal is to move the industry from early SDN/NFV trials to commercial deployments through this ecosystem collaboration.
E5 Intel Xeon Processor E5 Family Making the Business Case Intel IT Center
This presentation highlights cloud computing advantages of the Intel® Xeon® processor E5 family and helps you make the business case for investing. Includes access to an ROI calculator.
Cradlepoint PS6S9000UE - USB to Ethernet Cradle (quantum-wireless.com)Ari Zoldan
Rugged, Industrial-class, Single-slot Ethernet Cradle
• Connect Symbol mobile computers into any Ethernet network
• Tailored design allows one-handed insertion and removal of the mobile computer
• Supports mobile computers deployed in harsh, dirty or extreme environments
“Wired” Ethernet Connection
• Increase efficiency and speed with 10 mbps or 10/100 mbps Ethernet connection
• Provides simple and reliable integration between Symbol mobile computers and the most widely deployed networking technology
• Performance advantage over wireless and modem
• Preferred solution for large data transfers
• Reduced security concerns as compared to wireless connectivity options
Built on Industry Standards
• Supports mobile computers operating serially, as USB “Host” or USB “Peripheral”
• Simple installation and seamless integration with existing network infrastructure
• Compatible with DHCP or Static IP Address assignment
• Minimize support, integration and hardware requirements
Improve Workflow & Productivity
• Users can access and synchronize their data while charging at the same time
• Access files, Local Area Networks (LANs) and the Internet
• Supports server synchronization software
Quantum-Wireless.com
1-877-259-4629
- Intel dominates the TOP500 supercomputer list, powering 427 of the top 500 systems and 111 of the newest systems using Intel Xeon and Xeon Phi processors.
- HPC performance has improved over 15,000x in the past 20 years, with innovations like clusters now enabling top performance to waterfall down to single-socket systems within 6-8 years.
- Knights Landing, the next generation Intel Xeon Phi product, will provide over 3 teraflops of performance in a single package in 2015, using enhanced Intel Atom cores and on-package memory.
The document discusses Intel's HPC portfolio and roadmap update. It provides an overview of the new Intel Xeon E5-2600 v2 processor family, highlighting its efficiency, performance, and security features. The Xeon E5-2600 v2 is expected to deliver up to 30% more performance using the same or less power compared to the previous generation. It offers up to 12 cores, 30MB of cache, and support for the latest I/O and memory technologies to provide powerful and efficient processing for modern data centers.
Noile solutii Intel pentru afaceri eficiente-tm-20mai2010Agora Group
This document discusses Intel technologies for efficient IT infrastructures and provides the following key points:
1. Intel has fabrication facilities around the world producing microchips on advanced technology nodes.
2. Intel's Core processor family delivers scalable performance across devices from netbooks to servers through Intel architecture.
3. Intel's new 2010 Core processors feature technologies like Turbo Boost for intelligent performance and integrated graphics.
This document discusses new hardware features including CAT, COD, and Haswell, as well as network platforms. It provides an overview of run-to-completion and pipeline software models for network processing. Run-to-completion allows I/O and application work to be handled on a single core, while pipeline distributes packets to other cores for application work. Lockless queues are used to share data between cores and threads. Rings are the primary mechanism to move data between software units and I/O sources in DPDK.
This document discusses the DPDK roadmap. Version 2.1 is scheduled for release at the end of July 2015 and will include features like dynamic memory management, cuckoo hashing, packet framework enhancements, and driver support for additional NICs. Version 2.2 is scheduled for November 2015 and aims to support bifurcated drivers assuming their availability in the Linux kernel. The document also provides information on getting involved with the DPDK community through the dpdk.org website.
The document discusses Intel's Network Builders program, which aims to accelerate software-defined infrastructure adoption through open standards and platforms. It does this by investing in strong ecosystems, committing to open source, and leveraging Intel's technology leadership. The program enables partners through technical resources, matchmaking opportunities, and marketing support. It also works with network operators on proofs of concept and trials. The goal is to move the industry from early SDN/NFV trials to commercial deployments through this ecosystem collaboration.
E5 Intel Xeon Processor E5 Family Making the Business Case Intel IT Center
This presentation highlights cloud computing advantages of the Intel® Xeon® processor E5 family and helps you make the business case for investing. Includes access to an ROI calculator.
Cradlepoint PS6S9000UE - USB to Ethernet Cradle (quantum-wireless.com)Ari Zoldan
Rugged, Industrial-class, Single-slot Ethernet Cradle
• Connect Symbol mobile computers into any Ethernet network
• Tailored design allows one-handed insertion and removal of the mobile computer
• Supports mobile computers deployed in harsh, dirty or extreme environments
“Wired” Ethernet Connection
• Increase efficiency and speed with 10 mbps or 10/100 mbps Ethernet connection
• Provides simple and reliable integration between Symbol mobile computers and the most widely deployed networking technology
• Performance advantage over wireless and modem
• Preferred solution for large data transfers
• Reduced security concerns as compared to wireless connectivity options
Built on Industry Standards
• Supports mobile computers operating serially, as USB “Host” or USB “Peripheral”
• Simple installation and seamless integration with existing network infrastructure
• Compatible with DHCP or Static IP Address assignment
• Minimize support, integration and hardware requirements
Improve Workflow & Productivity
• Users can access and synchronize their data while charging at the same time
• Access files, Local Area Networks (LANs) and the Internet
• Supports server synchronization software
Quantum-Wireless.com
1-877-259-4629
- Intel dominates the TOP500 supercomputer list, powering 427 of the top 500 systems and 111 of the newest systems using Intel Xeon and Xeon Phi processors.
- HPC performance has improved over 15,000x in the past 20 years, with innovations like clusters now enabling top performance to waterfall down to single-socket systems within 6-8 years.
- Knights Landing, the next generation Intel Xeon Phi product, will provide over 3 teraflops of performance in a single package in 2015, using enhanced Intel Atom cores and on-package memory.
The document discusses Intel's HPC portfolio and roadmap update. It provides an overview of the new Intel Xeon E5-2600 v2 processor family, highlighting its efficiency, performance, and security features. The Xeon E5-2600 v2 is expected to deliver up to 30% more performance using the same or less power compared to the previous generation. It offers up to 12 cores, 30MB of cache, and support for the latest I/O and memory technologies to provide powerful and efficient processing for modern data centers.
Noile solutii Intel pentru afaceri eficiente-tm-20mai2010Agora Group
This document discusses Intel technologies for efficient IT infrastructures and provides the following key points:
1. Intel has fabrication facilities around the world producing microchips on advanced technology nodes.
2. Intel's Core processor family delivers scalable performance across devices from netbooks to servers through Intel architecture.
3. Intel's new 2010 Core processors feature technologies like Turbo Boost for intelligent performance and integrated graphics.
Noile tehnologii INTEL pentru infrastructuri IT eficiente-19mar2010Agora Group
This document discusses Intel technologies for efficient IT infrastructures and provides the following key points:
1. Intel has fabrication facilities around the world producing microchips on advanced technology nodes.
2. Intel's Core processor family delivers scalable performance across devices from netbooks to servers through Intel architecture.
3. Intel's new 2010 Core processors feature technologies like Turbo Boost for intelligent performance and integrated graphics.
This document discusses how DeviceNet networks commonly experience downtime from "bus off" errors, costing companies tens of thousands per year. It proposes that companies can reduce this downtime and cost by replacing DeviceNet components with an IO-Link solution, which uses point-to-point communication between IO-Link masters and slaves in a star configuration, making networks easier to troubleshoot. The document provides an example where converting to IO-Link eliminated 8-10 hours of downtime per month on a single machine.
The document discusses Intel technologies for high performance computing. It provides an overview of Intel's product families targeted at HPC workloads, including the Xeon E5-2600 v3 and E7-8800 v3 processor families. It also reviews some basics of HPC, including factors that impact performance such as memory bandwidth and latency. The document emphasizes that data movement between the CPU and memory hierarchy can often be a bottleneck, and that optimizing for high floating point operations per memory access is important.
NodeGrid Bold™ is the ultimate fully loaded IoT, POD, Retail, and Remote Office IT infrastructure management solution. NodeGrid Bold provides secure access and control for managing remote devices at the EDGE of your network.
For the full video of this presentation, please visit:
https://www.edge-ai-vision.com/2020/11/smarter-manufacturing-with-intels-deep-learning-based-machine-vision-a-presentation-from-intel/
For more information about edge AI and computer vision, please visit:
https://www.edge-ai-vision.com
Tara K. Thimmanaik, Solutions Architect at Intel, presents the “Smarter Manufacturing with Intel’s Deep Learning-Based Machine Vision” tutorial at the September 2020 Embedded Vision Summit.
As demand for smarter and more efficient manufacturing is growing, IoT technologies—including sensors, edge devices, gateways, servers and the cloud—are being used throughout the factory to compute deep learning analytics workloads at the appropriate location. Efficient data-driven manufacturing can help to reduce labor costs, increase quality and maximize profit. The biggest hindrance to achieving these outcomes is the difficulty in extracting data from vendor-locked and proprietary systems for analytics downstream.
In this presentation, Thimmanaik covers Intel’s approach to developing open, flexible and scalable solutions, including:
• Intel’s technologies such as OpenVINO, Movidius Vision Processor Units, Edge Insights Software (EIS) and deep learning algorithms
• How Intel’s offerings come together in the industrial marketplace with partnerships forged to address the constraints of manufacturing infrastructure
• Real-world examples highlighting defect detection in textile printing (where 90% accuracy at 50 fps was achieved) and smartphone screen production (where false negatives were only 0.6%)
Accelerate Ceph performance via SPDK related techniques Ceph Community
This document discusses techniques to accelerate Ceph performance using SPDK-related methods. It introduces DPDK for storage which uses DPDK and UNS technologies to optimize iSCSI front-end targets and provide higher system performance for iSCSI. A middle cache tiering solution is proposed to provide local caching and write logging between applications and Ceph for legacy protocol support, high performance, and high availability. The document also briefly mentions other building block techniques, I/O optimization, data processing acceleration, and ISA-L.
NodeGrid Flex™ is the ultimate IoT, POD, Retail, and Remote Office IT infrastructure management solution. NodeGrid Flex provides secure access and control with flexible types of ports to “mix and match” your needs for managing remote devices at the EDGE of your network.
Flexible Add-on Ports for IoT, POD, Retail & Remote Office EDGE Devices
EpiGuard is a hardware security appliance that secures data transmitted to and from devices like POS systems, mobile devices, printers and scanners. It is located between protected devices and the network to enforce security policies. As a small plug-and-play appliance, EpiGuard offers encryption and network segmentation for single or multiple devices. It is centrally managed through EpiForce and provides the same protections as EpiForce software agents through a hardware alternative.
The document compares the Motorola WiNG5 wireless LAN architecture to traditional hub-and-spoke WLAN solutions. It finds that the Motorola solution provides significantly greater distributed functionality by keeping authentication, bridging, roaming, firewalling and RF interference handling local to each access point. It also demonstrates strong site and service survivability when the link to the main controller is removed. Testing showed the Motorola solution was dramatically more efficient with WAN bandwidth usage and exhibited greater scalability, supporting over 10,000 access points across 4,000 domains serving 100,000 clients from a single controller.
This document summarizes two networking devices:
1) The TP-Link TL-ER5120 Router which has 1 gigabit WAN port and 3 interchangeable WAN/LAN ports. It offers security features like DoS defense and content filtering as well as load balancing and professional lightning protection.
2) The Brocade VDX 6730 Data Center Switch which has 10 gigabit Ethernet and Fibre Channel ports. It provides Ethernet storage connectivity and supports VLANs, MAC address filtering and port mirroring with ultra-low latency of 600 nanoseconds. It bridges Fibre Channel SANs and Ethernet fabrics.
Capstone Presentation For Five Rivers Medical Centersdjackson134
This here is the final presentation of our Capstone project that team members Cheyenna Carson, Susan Worcester, and myself Dawonne D. Jackson completed to acquire our Associate Degree of Applied Science in Computer Networking Systems.
The document discusses the responsibilities and skills required of IT professionals. It covers the importance of communication skills in troubleshooting, as well as ethics, legal issues, and how to handle difficult customers. It also describes call center environments and the roles of level 1 and 2 technicians. Key points are that technical expertise must be paired with strong communication, and IT professionals must maintain high ethical and professional standards.
The document proposes a network solution for a healthcare organization with 5 offices. It includes network designs, hardware selections, and costs. Key aspects include:
- Cabling and network designs for each office following industry standards
- Firewall, server, and desktop selections to meet current and future needs
- Virtualization, backups, and disaster recovery plans for data protection and business continuity
- Estimated total cost of $329,182 including hardware, labor, and optional service agreements
This document provides a summary of Jeff Zaleha's computer experience and employment history. It highlights his experience with various hardware, software, networking skills and certifications. It also lists his employment history from 2017 to present as a Prebill Auditor, Dispatcher, Help Desk Technician and previous roles he held as a technician supporting Windows migrations, POS installations and help desk roles.
Informix IWA data life cycle mgmt & Performance on Intel.Keshav Murthy
IWA works with a snapshot of the data in Informix data mart. Once you have defined the data mart on IWA and loaded the data, you need to periodically refresh the data. You can choose to refresh all of the data or just the partitions that were added, dropped or modified. Whether you have hundreds of gigabytes or many terabytes, faster refresh will help you analyze recent data more rapidly and get closer to real-time business. This session will explain the options for data refresh, review their performance and explain how to correctly implement a refresh plan. IBM and Intel will demonstrate live data refresh on the Intel Xeon platform and examine the impact on performance.
This document is a resume for Darnell Rias detailing his experience and qualifications for a network management role. It summarizes his certifications in IT, over 8 years of experience in roles such as system administrator and senior field service technician supporting the military, expertise in areas such as network engineering and software operating systems, and history of effectively managing help desk support and infrastructure projects. His professional development includes courses in network security, firewalls, networking essentials, and fiber optic termination.
Hands-on Lab: How to Unleash Your Storage Performance by Using NVM Express™ B...Odinot Stanislas
(FR)
Voici un excellent document qui explique étape après étape comment installer, monitorer et surtout correctement benchmarker ses SSD PCIe/NVMe (pas si simple que ça). Autre élément clé : comment analyser la charge I/O de véritables applications? Combien d'IOPS, en read, en write, quelle bande passante et surtout quel impact sur la durée de vie des SSD? Bref à mettre en toute les mains, et un merci à mon collègue Andrey Kudryavtsev.
(EN)
An excellent content which describe step by step how to install, monitor and benchmark PCIe/NVMe SSD (many trick not so simple). Another key learning: how to measure real I/O activities on a real workload? How many R/W IOPS, block size, throughtput, and finally what's the impact on SSD endurance and (real)life? A must read, and a huge thanks to my colleague Andrey Kudryavtsev.
Auteurs/Authors:
Andrey Kudryavtsev, SSD Solution Architect, Intel Corporation
Zhdan Bybin, Application Engineer, Intel Corporation
October 8, 2014 Webcast Identifying PCI Express 3.0 Dynamic Equalization Problems
Dynamic equalization training is a unique capability in modern day serial data communication. Through optimization of transmitter and receiver equalization parameters we can push the limits of serial data rates. Troubleshooting dynamic equalization problems at the Physical and Protocol layers requires testing for proper functionality and compliance.
Join Teledyne LeCroy's Stephen Mueller for this webinar to understand how to address this measurement challenge. The presentation will include real measurement examples and debugging techniques to pinpoint problems in the physical and protocol layers.
Scale-out Storage on Intel® Architecture Based Platforms: Characterizing and ...Odinot Stanislas
Issue du salon orienté développeurs d'Intel (l'IDF) voici une présentation plutôt sympa sur le stockage dit "scale out" avec une présentation des différents fournisseurs de solutions (slide 6) comprenant ceux qui font du mode fichier, bloc et objet. Puis du benchmark sur certains d'entre eux dont Swift, Ceph et GlusterFS.
This document discusses userspace storage systems as an alternative to kernel-based storage for petascale workloads. It outlines several userspace filesystems, block storage systems, and object storage systems used in practice. Common languages used include C, Python, Java, and Golang. Interfaces to the kernel include FUSE, UIO, DPDK and libvma. Challenges include balancing performance, scalability, and complexity across unified, self-managing systems. Specific examples covered are NFS-Ganesha, GlusterFS, HDFS, NBD, tgt, and caching systems like Tachyon and Redis.
This document provides an overview of the PCI Express physical layer technology. It discusses the lane counts and data rates supported by different PCIe versions. It describes the three logical layers of PCIe and divides the physical layer into logical and electrical sub-blocks. It explains several key physical layer technologies including 8b/10b encoding, data scrambling, de-emphasis, and link training. Link training negotiates link parameters and ensures synchronization between devices on the link.
Noile tehnologii INTEL pentru infrastructuri IT eficiente-19mar2010Agora Group
This document discusses Intel technologies for efficient IT infrastructures and provides the following key points:
1. Intel has fabrication facilities around the world producing microchips on advanced technology nodes.
2. Intel's Core processor family delivers scalable performance across devices from netbooks to servers through Intel architecture.
3. Intel's new 2010 Core processors feature technologies like Turbo Boost for intelligent performance and integrated graphics.
This document discusses how DeviceNet networks commonly experience downtime from "bus off" errors, costing companies tens of thousands per year. It proposes that companies can reduce this downtime and cost by replacing DeviceNet components with an IO-Link solution, which uses point-to-point communication between IO-Link masters and slaves in a star configuration, making networks easier to troubleshoot. The document provides an example where converting to IO-Link eliminated 8-10 hours of downtime per month on a single machine.
The document discusses Intel technologies for high performance computing. It provides an overview of Intel's product families targeted at HPC workloads, including the Xeon E5-2600 v3 and E7-8800 v3 processor families. It also reviews some basics of HPC, including factors that impact performance such as memory bandwidth and latency. The document emphasizes that data movement between the CPU and memory hierarchy can often be a bottleneck, and that optimizing for high floating point operations per memory access is important.
NodeGrid Bold™ is the ultimate fully loaded IoT, POD, Retail, and Remote Office IT infrastructure management solution. NodeGrid Bold provides secure access and control for managing remote devices at the EDGE of your network.
For the full video of this presentation, please visit:
https://www.edge-ai-vision.com/2020/11/smarter-manufacturing-with-intels-deep-learning-based-machine-vision-a-presentation-from-intel/
For more information about edge AI and computer vision, please visit:
https://www.edge-ai-vision.com
Tara K. Thimmanaik, Solutions Architect at Intel, presents the “Smarter Manufacturing with Intel’s Deep Learning-Based Machine Vision” tutorial at the September 2020 Embedded Vision Summit.
As demand for smarter and more efficient manufacturing is growing, IoT technologies—including sensors, edge devices, gateways, servers and the cloud—are being used throughout the factory to compute deep learning analytics workloads at the appropriate location. Efficient data-driven manufacturing can help to reduce labor costs, increase quality and maximize profit. The biggest hindrance to achieving these outcomes is the difficulty in extracting data from vendor-locked and proprietary systems for analytics downstream.
In this presentation, Thimmanaik covers Intel’s approach to developing open, flexible and scalable solutions, including:
• Intel’s technologies such as OpenVINO, Movidius Vision Processor Units, Edge Insights Software (EIS) and deep learning algorithms
• How Intel’s offerings come together in the industrial marketplace with partnerships forged to address the constraints of manufacturing infrastructure
• Real-world examples highlighting defect detection in textile printing (where 90% accuracy at 50 fps was achieved) and smartphone screen production (where false negatives were only 0.6%)
Accelerate Ceph performance via SPDK related techniques Ceph Community
This document discusses techniques to accelerate Ceph performance using SPDK-related methods. It introduces DPDK for storage which uses DPDK and UNS technologies to optimize iSCSI front-end targets and provide higher system performance for iSCSI. A middle cache tiering solution is proposed to provide local caching and write logging between applications and Ceph for legacy protocol support, high performance, and high availability. The document also briefly mentions other building block techniques, I/O optimization, data processing acceleration, and ISA-L.
NodeGrid Flex™ is the ultimate IoT, POD, Retail, and Remote Office IT infrastructure management solution. NodeGrid Flex provides secure access and control with flexible types of ports to “mix and match” your needs for managing remote devices at the EDGE of your network.
Flexible Add-on Ports for IoT, POD, Retail & Remote Office EDGE Devices
EpiGuard is a hardware security appliance that secures data transmitted to and from devices like POS systems, mobile devices, printers and scanners. It is located between protected devices and the network to enforce security policies. As a small plug-and-play appliance, EpiGuard offers encryption and network segmentation for single or multiple devices. It is centrally managed through EpiForce and provides the same protections as EpiForce software agents through a hardware alternative.
The document compares the Motorola WiNG5 wireless LAN architecture to traditional hub-and-spoke WLAN solutions. It finds that the Motorola solution provides significantly greater distributed functionality by keeping authentication, bridging, roaming, firewalling and RF interference handling local to each access point. It also demonstrates strong site and service survivability when the link to the main controller is removed. Testing showed the Motorola solution was dramatically more efficient with WAN bandwidth usage and exhibited greater scalability, supporting over 10,000 access points across 4,000 domains serving 100,000 clients from a single controller.
This document summarizes two networking devices:
1) The TP-Link TL-ER5120 Router which has 1 gigabit WAN port and 3 interchangeable WAN/LAN ports. It offers security features like DoS defense and content filtering as well as load balancing and professional lightning protection.
2) The Brocade VDX 6730 Data Center Switch which has 10 gigabit Ethernet and Fibre Channel ports. It provides Ethernet storage connectivity and supports VLANs, MAC address filtering and port mirroring with ultra-low latency of 600 nanoseconds. It bridges Fibre Channel SANs and Ethernet fabrics.
Capstone Presentation For Five Rivers Medical Centersdjackson134
This here is the final presentation of our Capstone project that team members Cheyenna Carson, Susan Worcester, and myself Dawonne D. Jackson completed to acquire our Associate Degree of Applied Science in Computer Networking Systems.
The document discusses the responsibilities and skills required of IT professionals. It covers the importance of communication skills in troubleshooting, as well as ethics, legal issues, and how to handle difficult customers. It also describes call center environments and the roles of level 1 and 2 technicians. Key points are that technical expertise must be paired with strong communication, and IT professionals must maintain high ethical and professional standards.
The document proposes a network solution for a healthcare organization with 5 offices. It includes network designs, hardware selections, and costs. Key aspects include:
- Cabling and network designs for each office following industry standards
- Firewall, server, and desktop selections to meet current and future needs
- Virtualization, backups, and disaster recovery plans for data protection and business continuity
- Estimated total cost of $329,182 including hardware, labor, and optional service agreements
This document provides a summary of Jeff Zaleha's computer experience and employment history. It highlights his experience with various hardware, software, networking skills and certifications. It also lists his employment history from 2017 to present as a Prebill Auditor, Dispatcher, Help Desk Technician and previous roles he held as a technician supporting Windows migrations, POS installations and help desk roles.
Informix IWA data life cycle mgmt & Performance on Intel.Keshav Murthy
IWA works with a snapshot of the data in Informix data mart. Once you have defined the data mart on IWA and loaded the data, you need to periodically refresh the data. You can choose to refresh all of the data or just the partitions that were added, dropped or modified. Whether you have hundreds of gigabytes or many terabytes, faster refresh will help you analyze recent data more rapidly and get closer to real-time business. This session will explain the options for data refresh, review their performance and explain how to correctly implement a refresh plan. IBM and Intel will demonstrate live data refresh on the Intel Xeon platform and examine the impact on performance.
This document is a resume for Darnell Rias detailing his experience and qualifications for a network management role. It summarizes his certifications in IT, over 8 years of experience in roles such as system administrator and senior field service technician supporting the military, expertise in areas such as network engineering and software operating systems, and history of effectively managing help desk support and infrastructure projects. His professional development includes courses in network security, firewalls, networking essentials, and fiber optic termination.
Hands-on Lab: How to Unleash Your Storage Performance by Using NVM Express™ B...Odinot Stanislas
(FR)
Voici un excellent document qui explique étape après étape comment installer, monitorer et surtout correctement benchmarker ses SSD PCIe/NVMe (pas si simple que ça). Autre élément clé : comment analyser la charge I/O de véritables applications? Combien d'IOPS, en read, en write, quelle bande passante et surtout quel impact sur la durée de vie des SSD? Bref à mettre en toute les mains, et un merci à mon collègue Andrey Kudryavtsev.
(EN)
An excellent content which describe step by step how to install, monitor and benchmark PCIe/NVMe SSD (many trick not so simple). Another key learning: how to measure real I/O activities on a real workload? How many R/W IOPS, block size, throughtput, and finally what's the impact on SSD endurance and (real)life? A must read, and a huge thanks to my colleague Andrey Kudryavtsev.
Auteurs/Authors:
Andrey Kudryavtsev, SSD Solution Architect, Intel Corporation
Zhdan Bybin, Application Engineer, Intel Corporation
October 8, 2014 Webcast Identifying PCI Express 3.0 Dynamic Equalization Problems
Dynamic equalization training is a unique capability in modern day serial data communication. Through optimization of transmitter and receiver equalization parameters we can push the limits of serial data rates. Troubleshooting dynamic equalization problems at the Physical and Protocol layers requires testing for proper functionality and compliance.
Join Teledyne LeCroy's Stephen Mueller for this webinar to understand how to address this measurement challenge. The presentation will include real measurement examples and debugging techniques to pinpoint problems in the physical and protocol layers.
Scale-out Storage on Intel® Architecture Based Platforms: Characterizing and ...Odinot Stanislas
Issue du salon orienté développeurs d'Intel (l'IDF) voici une présentation plutôt sympa sur le stockage dit "scale out" avec une présentation des différents fournisseurs de solutions (slide 6) comprenant ceux qui font du mode fichier, bloc et objet. Puis du benchmark sur certains d'entre eux dont Swift, Ceph et GlusterFS.
This document discusses userspace storage systems as an alternative to kernel-based storage for petascale workloads. It outlines several userspace filesystems, block storage systems, and object storage systems used in practice. Common languages used include C, Python, Java, and Golang. Interfaces to the kernel include FUSE, UIO, DPDK and libvma. Challenges include balancing performance, scalability, and complexity across unified, self-managing systems. Specific examples covered are NFS-Ganesha, GlusterFS, HDFS, NBD, tgt, and caching systems like Tachyon and Redis.
This document provides an overview of the PCI Express physical layer technology. It discusses the lane counts and data rates supported by different PCIe versions. It describes the three logical layers of PCIe and divides the physical layer into logical and electrical sub-blocks. It explains several key physical layer technologies including 8b/10b encoding, data scrambling, de-emphasis, and link training. Link training negotiates link parameters and ensures synchronization between devices on the link.
The webinar discusses how NVMe will change flash storage by providing faster performance compared to SAS/SATA and proprietary PCIe flash. It will explain what NVMe is, its advantages over other interconnects, and how to get started with NVMe. Attendees will learn about NVMe from representatives of OCZ Storage Solutions and analyst firm Storage Switzerland. OCZ will also showcase its new Z-Drive 6000 SSD series designed for the NVMe standard.
Hardware accelerated virtio networking for nfv linux consprdd
This document discusses network function virtualization (NFV) and its benefits over traditional hardware-based network architectures. It describes how NFV allows network services to be deployed virtually through software rather than requiring new physical hardware. This reduces costs and improves flexibility compared to traditional networks which require increased capital expenditure for new infrastructure. The document also covers virtualization mechanisms like virtio and vhost-net that help enable the virtualization of network services.
PCI Express is a serial computer expansion bus standard designed to replace older standards like PCI and AGP. It uses point-to-point connections between two devices using serial communication over one or more lanes. PCIe protocol has three layers - the transaction layer which interacts with software, the data link layer which provides reliable packet exchange, and the physical layer which isolates the other layers from signaling technology.
PCIe Gen 3.0 Presentation @ 4th FPGA CampFPGA Central
PCIe Gen3 presentation by PLDA at 4th FPGA Camp in Santa Clara, CA. For more details visit http://www.fpgacentral.com/fpgacamp or http://www.fpgacentral.com
NVMe PCIe and TLC V-NAND It’s about TimeDell World
With an explosion in data and the relentless growth in demand for information, identifying a much more efficient means of storage has become extremely important. In this session, we will cover the key drivers behind the need for faster and more efficient storage. NVMe, a standardized protocol for PCIe-based storage, is giving users the huge leap in bandwidth required for demanding applications. Samsung, who makes the fastest NVMe SSDs on the market, will cover the benefits enabled by such technology, in areas such as fraud prevention and surgical procedures.
The technology behind flash drives – NAND memory – will be spotlighted in this presentation. Memory manufacturers have improved NAND’s value by migrating from single-level-cell to multi-level-cell designs, but the most significant evolution will be a marriage of triple-level-cell and V-NAND flash manufacturing technologies. Samsung will also provide an overview of the prospects for TLC V-NAND with mobile device manufacturers, while examining the strong potential for a much wider TLC V-NAND market in data centers.
DPDK в виртуальном коммутаторе Open vSwitch / Александр Джуринский (Selectel)Ontico
Intel DPDK (Data Plane Development Kit) — набор драйверов и библиотек, позволяющих приложениям взаимодействовать с сетевым устройством напрямую, минуя сетевой стек Linux. Это значительно увеличивает скорость обработки пакетов. DPDK интегрируется с рядом популярных программных решений, например, c виртуальным коммутатором Open vSwitch.
Возможностям и перспективам использования связи Open vSwitch + DPDK в облачных проектах и будет посвящен наш доклад. Мы подробно остановимся на проведённых тестах производительности и интерпретируем их результаты. Отдельное внимание будет уделено анализу трудностей и ограничений, с которыми пришлось столкнуться в ходе экспериментов.
PCI Express* based Storage: Data Center NVM Express* Platform TopologiesOdinot Stanislas
This document discusses PCI Express based solid state drives (SSDs) for data centers. It covers the growth opportunity for PCIe SSDs, topology options using various form factors like SFF-8639 and M.2, and validation tools. It also discusses hot plug support on Intel Xeon processor based servers and upcoming industry workshops to advance the PCIe SSD ecosystem.
This document summarizes and compares the function call flow of the Linux NVMe driver on legacy versions and newer versions that use blk-mq.
In legacy versions before 3.19, bios bypass the block layer and are directly submitted to the NVMe driver. In versions 3.19 and later, bios go through blk-mq and are converted to requests that are queued to the NVMe driver.
The document outlines the key steps in the function call flow for legacy versions, which include converting bios to iods and submitting them, as well as setting up callback information. It then describes the similar but modified process for blk-mq versions, which includes building requests from bios, flushing requests between software queues, and
PCIe and PCIe driver in WEC7 (Windows Embedded compact 7)gnkeshava
This document provides an overview of PCI and PCI Express buses. It defines PCI as a parallel bus standard introduced in 1993 to connect peripherals to the computer. PCI Express is described as a newer serial standard introduced in 2004 that offers improvements over PCI like higher throughput and hot plugging support. The document outlines key differences between the two standards and provides details on PCI Express architecture including root complexes, endpoints, switches, and bridges. It also covers topics like PCI Express lanes, connectors, configuration space, and memory mapping. In the end it provides a brief introduction to PCI bus drivers in the Windows environment.
The document provides an overview of the PCI Express system architecture. It discusses the architectural perspective of PCI Express including how it maintains backwards compatibility with PCI/PCI-X while improving performance through serial point-to-point connectivity and packet-based transactions. It also covers the PCI Express transaction model and types, including memory, I/O, configuration and message transactions, as well as posted and non-posted transaction types.
Red Hat Storage Day New York - Intel Unlocking Big Data Infrastructure Effici...Red_Hat_Storage
This document discusses using Ceph storage with Apache Hadoop to provide a scalable and efficient storage solution for big data workloads. It outlines the challenges of scaling Hadoop storage independently from compute resources using the native Hadoop Distributed File System. The solution presented is to use the open source Ceph storage system instead of direct-attached storage. This allows Hadoop compute and storage resources to scale independently and provides a centralized storage platform for all enterprise data workloads. Performance tests showed the Ceph and Hadoop configuration providing up to a 60% improvement in I/O performance when using Intel caching software and SSDs.
This document provides a summary of a book about PCI Express technology. It includes an introduction to the book, a table of contents listing the topics covered, biographies of the authors, and endorsements of the book. The summary is as follows:
The document introduces a book that provides a comprehensive guide to PCI Express generations 1.x, 2.x, and 3.0. It includes biographies of the authors and an endorsement quoting that the book is essential for understanding PCI Express. The table of contents indicates it will cover topics such as the origins and architecture of PCI Express, various technical specifications and features, and considerations for high-speed signaling.
Red Hat Storage Day New York -Performance Intensive Workloads with Samsung NV...Red_Hat_Storage
This document discusses using Samsung NVMe SSDs and Red Hat Ceph storage to create a high performance storage tier for OpenStack environments. It presents a reference architecture using a 3-node Ceph cluster with Samsung NVMe SSDs that achieved over 28GB/s for sequential reads. This architecture provides scalable, open source storage optimized for performance-intensive workloads like databases, analytics, and networking. Future work is discussed to develop a similar architecture using GlusterFS storage.
Devconf2017 - Can VMs networking benefit from DPDKMaxime Coquelin
DPDK brings high-performance/low-latency virtualization networking capabilities thanks to its Vhost/Virtio support. The session will first introduce DPDK and its Vhost/Virtio implementations, exposing to the audience examples of possible uses, and challenges that need to be addressed to achieve high-performance, functionality and reliability. Then, Vhost/Virtio improvements introduced in last DPDK release will be covered, such as receive path optimizations, Virtio's indirect descriptors support, or transmit zero copy to name a few. The speakers will explain which problems they aim to address, how they address them, mentioning their limitations.
Finally, the speakers, who are active DPDK's Virtio/Vhost contributors, will expose what new developments are in the pipe to tackle the remaining challenges.
The session will be presented so that DPDK developers and users find useful information on current developments and status. People not familiar with DPDK may find a overview, get and share ideas with other projects.
Red Hat Storage Day New York - Penguin Computing Spotlight: Delivering Open S...Red_Hat_Storage
This document discusses Penguin Computing's open solutions utilizing Red Hat Storage. It describes Penguin Computing as providing compute, storage, and networking solutions using open technologies. It then discusses various Penguin Computing solutions like the Tundra Extreme Scale open compute platform, Arc5ca Ethernet switches, and FrostByte HS storage appliances. The document also summarizes Red Hat Gluster Storage benefits for financial data analytics like deeper analysis, lower costs, and better performance compared to traditional storage solutions.
DPDK Summit - 08 Sept 2014 - Intel - Networking Workloads on Intel ArchitectureJim St. Leger
Venky Venkatesan presents information on the Data Plane Development Kit (DPDK) including an overview, background, methodology, and future direction and developments.
A Path to NFV/SDN - Intel. Michael Brennan, INTELWalton Institute
This document discusses Intel's approach to accelerating the adoption of Network Functions Virtualization (NFV) and Software-Defined Networking (SDN). It outlines Intel's open strategy of advancing open source software and standards, delivering open reference designs for Intel platforms, and enabling a broad ecosystem of partners. The goal is to help networking platforms based on Intel architecture replace proprietary, dedicated networking appliances. The document also references Intel's Open Network Platform (ONP) server and switch software reference designs, and examples of trials and deployments Intel is collaborating on with telecom, cloud, and enterprise customers.
Omni path-fabric-software-architecture-overviewDESMOND YUEN
Overview of the Omni Path Fabric Software Architecture. How to take advantage of the new features and optimize for the maximum performance for the different software applications out there.
If you like what you read be sure you ♥ it below. Thank you!
An overview of Intel® Omni-Path Architecture (Intel® OPA), its main APIs and software capabilities, which includes high-performance computing (HPC) applications and cluster management features.
Introduction to container networking in K8s - SDN/NFV London meetupHaidee McMahon
This document discusses Intel's work on container networking technologies for network functions virtualization (NFV). It outlines three deployment models for containers in NFV environments - bare metal, unified infrastructure, and hybrid. It also addresses key challenges for using containers in bare metal environments, such as providing multiple network interfaces and high-performance data planes. Intel is working to help solve these challenges through open source solutions and experience kits that provide best practices.
Accelerating Insights in the Technical Computing TransformationIntel IT Center
- Intel dominates the TOP500 supercomputer list, powering 427 of the top 500 systems and 111 of the newest systems using Intel Xeon and Xeon Phi processors.
- HPC performance has improved over 15,000x in the past 20 years, with innovations like clusters now bringing high performance computing to single sockets.
- Intel's next-generation Knights Landing Xeon Phi will provide over 3 teraflops of performance in a single package while improving power efficiency 5x compared to previous generations.
Embedded Chief River Design-In Presentation_30442998.pdfOemTest
This document provides an overview of the Ivy Bridge-Mobile and Ivy Bridge-Mobile+ECC processors and the Panther Point-Mobile platform controller hub. Key details include:
- Ivy Bridge is Intel's 22nm processor featuring improved performance over Sandy Bridge through architecture enhancements and a new manufacturing process.
- It includes integrated graphics with support for up to 3 displays, as well as new instruction sets like AVX.
- The Panther Point chipset provides additional connectivity options like USB 3.0 and improved PCIe support.
- Memory support is expanded to include DDR3 up to 1600MHz and DDR3L 1333MHz.
- Power optimizations aim to maintain or reduce power consumption compared to
The document discusses reimagining the datacenter through software defined infrastructure. This allows datacenters to become more dynamic, automated and efficient by treating compute, storage and networking resources as composable blocks that can be allocated on demand. This approach breaks down traditional silos and allows simpler deployment and maintenance while improving agility, automation and efficiency. The software defined approach is compared to the traditional rigid infrastructure model and examples are given of how it can improve provisioning times, utilization rates and flexibility.
Spring Hill (NNP-I 1000): Intel's Data Center Inference Chipinside-BigData.com
- SpringHill (NNP-I1000) is Intel's new data center inference chip that provides best-in-class performance per watt for major data center inference workloads.
- It delivers 4.8 TOPs/watt of performance and can scale from 10-50 watts to boost performance.
- The chip features 12 inference compute engines, 24MB of shared cache, and Intel architecture cores to drive AI innovation while maintaining high performance and efficiency.
The document discusses Intel's efforts to promote network function virtualization (NFV) and software-defined networking (SDN) for 5G networks. It provides details on Intel's Network Builders program and Developer Zone community which aim to enable NFV/SDN solutions using Intel architecture. The rest of the document outlines Intel's investments in technologies like FlexRAN and MEC to deliver a "network in a box" approach for 5G, as well as Intel's engagement with mobile operators on their NFV strategies and requirements.
OSN Bay Area Feb 2019 Meetup: Intel, Dynamic Device Personalization - Journey...Lumina Networks
Profiles for Intel® Ethernet 700 Series enables run-time updating and configuration of parse graph to expand or modify protocol support allowing early adoption of new technologies.
Lynn Comp - Intel Big Data & Cloud Summit 2013 (2)IntelAPAC
This document discusses architecting cloud infrastructure for the future. It addresses requirements through workload optimized technologies, composable resources, and software defined infrastructure. Workload optimized technologies match different workloads with optimized server, storage, and network technologies. Composable resources allow flexible, efficient data centers through modular compute, memory, storage, and fabric resources that can be pooled and shared. Software defined infrastructure involves re-architecting the network through software defined networking and re-architecting storage through software defined storage.
Intel® Select Solutions for the Network provide a faster means to address these challenges as we transition to 5G with pre-validated, optimized building blocks to help drive scale. Hear the what, why, when and where around Intel® Select Solutions for the Network.
This document provides an introduction to the Intel Data Plane Development Kit (DPDK) and discusses:
- DPDK addresses the challenges of high-speed packet processing on Intel architectures by eliminating kernel and interrupt overheads through a userspace polling model.
- DPDK is open source under a BSD license, allowing free use and modification of the code.
- DPDK optimizes packet processing performance through techniques like huge pages, prefetching, and affinity of threads to CPU cores.
Edge Computing and 5G - SDN/NFV London meetupHaidee McMahon
Edge computing and 5G will enable one compute platform from edge to cloud. This will provide virtual, software-defined, and cloud-ready capabilities to support the 5G future. Key applications that will benefit include gaming/VR with high bandwidth and low latency requirements, as well as industrial and public safety uses involving real-time video analytics, surveillance, and facial recognition. Edge computing deployments will optimize performance for applications with strict latency constraints by placing computing resources closer to endpoints and users.
This document discusses the Internet of Things (IoT) and smart cities. It provides an overview of the growth of unconnected devices and the opportunity for data sharing and analytics. Intel offers a range of products from devices to the cloud to enable end-to-end IoT solutions. Example use cases are described for smart buildings, energy, retail, and transport. The document also outlines Intel's gateway solutions and a smart city project in Dublin using a network of gateways for environmental monitoring, participatory sensing, and data fusion to power citizen-centric services and apps.
The document discusses the future of storage technologies for cloud computing. It notes that cloud adoption is driving significant business opportunities but also increasing complexity. Intel's strategy is to build an open ecosystem, reduce complexity, and enable massive compute capabilities. New storage technologies like SSDs and NVMe can help optimize performance by providing much higher bandwidth and lower latency compared to hard disk drives. For example, using Intel SSDs with NVMe instead of HDDs can provide over 100x cost savings and 1400x power savings while also improving performance for database restart tasks by over 30 times.
Industrial IoT with Intel IoT Gateway & OctobluIntel® Software
This document discusses Intel and Octoblu's partnership to enable industrial IoT solutions. It describes Intel's IoT gateway platform and Octoblu's IoT cloud platform. Together, their platforms provide a full-stack solution for connecting devices, routing data, running automation workflows, and providing security and management tools. The document also notes the large market opportunity for industrial IoT and provides an overview of how customers can deploy the joint solution either on-premises or via various cloud options.
Re-architecting the Datacenter to Deliver Better Experiences (Intel)COMPUTEX TAIPEI
The document discusses Intel's efforts to re-architect datacenters to better meet growing demands and enable new digital experiences. Key points include:
- Convergence of cloud, big data, and connected devices is driving new user experiences
- Intel is reducing cost, complexity and power consumption by re-architecting the datacenter at the rack and system level
- Intel's broad portfolio of compute, storage, networking and software technologies allow it to optimize workloads and deliver better performance
Re-architecting the Datacenter to Deliver Better Experiences (Intel)
HP_NextGEN_Training_Q4_2015
1. Placeholder Footer Copy / BU Logo or Name Goes Here
Chris Hopkins – Product Marketing Engineer
Networking Division
Q4’15
Intel® Ethernet
Benefits of Moving to
the Next Generation
3. Placeholder Footer Copy / BU Logo or Name Goes Here
• Ethernet Market Overview
• Evolution of the Network
• Moving to the next generation
• New technologies for virtualized networks
• Working Together for greater ROI
• Why Choose Intel® Ethernet Wrap up
4. 4
TRANSFORM THE NETWORKTRANSFORM THE NETWORK
4
Long Range Forecast For Ethernet
Intel Confidential
Source: Crehan Research, LRF , 2015
5. 5
TRANSFORM THE NETWORKTRANSFORM THE NETWORK
5
Thank You for Driving ND MSS!
(NIC, Mezzanine, & LOM)
Intel ConfidentialSource: Crehan Research, Q3'15
7. TRANSFORMING NETWORKING & STORAGE
7
Software Defined Infrastructure
Network Virtualization
Network Functions Virtualization
10/20/40+ Gb Ethernet
Unified Network
I/O Virtualization
10Gb Ethernet
Discrete Networks
Multi-port I/O
1Gb Ethernet
Discrete Data Center
Compute Storage Networking
ConsoleConsoleConsole
Manageability
Virtualized Data Center
Console
Console
Manageability
Console
Resource Pools
Cloud Data Center
Manageability
Console
…Evolution of the Network
Evolution of the Data Center…
8. TRANSFORMING NETWORKING & STORAGE
8
Intel® Ethernet Controller X710 &
XL710Family
Designed For the Datacenter of Tomorrow
Yesterday
Manual – Fixed – Hardware
Defined
Tomorrow
Automated – Flexible – Software Defined
9. TRANSFORMING NETWORKING & STORAGE
9
Intel® Xeon® processor
E5-2600 v3 Family
Intel® Data Directed I/O makes the
processor cache the primary destination and
source of I/O data rather than main memory
Intel® Ingredients for Workload Optimization
Storage
Intel® Solid State Drive
DC P3700 Family
PCI Express* brings extreme data throughput
directly to Intel® Xeon® processors
Intel® Ethernet Controller
XL710 Family
40GbE & 10GbE connectivity for
Enterprise, Cloud and Communications
Intel Ethernet Converged
Network Adapter
XL710 / X710 Family
Intel® QuickAssist
Technology
Offloads packet processing technology
thereby reserving processor cycles for
application and control processing
Intel® QuickAssist Adapter
8950-SCCP
Intel® Solid-State Drive
DC P3700 Series Family
Intel® Communications Chipset 89xx
Intel® C610 Series Chipset Chipset
Network
Acceleration
Software
Intel® Data Plane
Development Kit
Packet Processing Software create the
foundation for NFV / SDN, server
virtualization and vSwitch optimizations
Compute
10. TRANSFORMING NETWORKING & STORAGE
10
Intel® Ethernet Controller X710 Features
• Low power single chip design for PCI Express* 3.0
• Software configurable Ethernet Port Speed for up to 2x10GbE or up to 4x10GbE
• Interfaces for Converged Network Adapters, backplanes and LAN on MotherboardFlexibility
• Network virtualization Overlay stateless offloads for Geneve, VXLAN and NVGRE
• Intelligent load balance for high performance traffic flows of virtual machines
• Intelligent Offloads accelerate operating systems storage initiators to deliver high
performance for NAS (NFS, SMB), and SAN (iSCSI)
Acceleration
• Intel® Data Plane Development Kit (DPDK) optimized for efficient packet
processing to support Network Function Virtualization (NFV)
• Intel® Ethernet Flow Director for hardware based application traffic steering
• Intel® Data Direct I/O (Intel® DDIO) makes the processor cache the primary
destination and source of I/O data rather than main memory
Better
Together
11. TRANSFORMING NETWORKING & STORAGE
11
HP Portfolio of Intel® X710/XL710 NDC’s and Adapters
Intel® Ethernet XL710 (*40GbE) Targeted for Gen9, in 2H’2016
Hanson FlatGarden Key Henry Key Igloo
20GbE 20GbE 10GbE 10GbE
12. TRANSFORMING NETWORKING & STORAGE
12
Benefits of Moving to the Next Generation
Feature X520 X710/XL710 Benefit
Port Count 2 ports 2x10, 2x20 ports Scalable 10GbE
Host Interface PCIe Gen2 x8 PCIe Gen3 x8 Allows up to 2x20 GbE
Virtual Embedded Switch Up to 23Gb Up to 50Gb Double bandwidth for VM to VM traffic
across PCIe bus
Host Virtualization 64 Virtual Station Interfaces 384 Virtual Station Interfaces Increased consolidation by support more
than double virtual machines
Network Virtualization RSS for VXLAN outer
header
Full VXLAN, NVGRE, Geneve and
NSH offload support
Virtualized Network traffic at linerate with
low CPU usage for ESXi, Linux, KVM and
Windows
Intel® Ethernet Flow
Director
TCP/IP or SCTP/IP only Up to 64 different packet types
(L2/IPV4/IPv6/TCP/UDP/SCTP)
Increase HW base traffic steering for cloud
based applications
Multi-Function Port SR-IOV only Flex10 and SR-IOV Physical partitioning and virtual partitioning
Flexible Logic Driver NVM updates Support new protocols like Geneve and
NSH
13. Placeholder Footer Copy / BU Logo or Name Goes Here
Developing New Technologies
for the Virtualized Network
14. TRANSFORMING NETWORKING & STORAGE
14
Network Functions Virtualization (NFV)
Hypervisor
Virtual Switch
Physical
Hardware
Hypervisor
Virtual Switch
Physical
Hardware
Router VPN Firewall Load Balancer
Network Services
Switch
Current Model
• Services in dedicated hardware or physical boxes
that are Network Topology dependent
• Inflexible deployment model, requires changing
forwarding behavior NFV is about dynamic provisioning of services
• Virtualizing service functions on IA based servers in VMs
Today IT delivers a network service by utilizing ordered sets of cooperating network
applications known as Service Function Chain (SFC)
Intel Confidential
15. TRANSFORMING NETWORKING & STORAGE
15
Network Virtualization: Abstracts Physical Network
Server Virtualization Network Virtualization
Virtual Network 2
Hypervisor
Virtual Switch
Physical
Hardware
Physical IP Network
Virtual Network Abstraction using tunnel overlays e.g. VXLAN, Geneve and NVGRE
Open Virtual Switch
Open Virtual Switch
Open Virtual Switch
Open Virtual Switch
Network Virtualization Controller using VMware* NSX
Virtual Network 3Virtual Network 1
16. TRANSFORMING NETWORKING & STORAGE
16
Intel® Virtualization Technology
CPU utilization per core
Core 1 Core 2 Core 3 Core 4 Core5 Core N
CPU utilization per core
Core 1 Core 2 Core 3 Core 4 Core 5 Core N
VXLAN Network Virtualization Optimizations using Receive Side Scaling
VTEP / Virtual Switch
VTEP / Virtual Switch
Receive Side Scaling for VXLAN TrafficWithout Receive Side Scaling
Single
Rx Queue
Multiple
Rx Queues
Intel Confidential
Network Virtualization Offloads: RSS, VMDq, LSO, CRC
17. TRANSFORMING NETWORKING & STORAGE
17
Generic Network Virtualization Encapsulation (Geneve)
R –
Reserved
Geneve Option: Type, Length, Value (TLV) Format
Outer
Ethernet Header
Outer IP
Header
Outer UDP
Header
Geneve Base
Header
Geneve
Options
Inner
Payload
Outer
CRC
Geneve Header:
Co-authored byVersion | Option Length | OAM | Critical Options | Reserved | Protocol Type
Virtual Network Identifier (VNI) | Reserved
Variable Length Options
Option Class | Option Type | R | R | R | Length
Variable Length Options
Geneve overview:
• Geneve is UDP encapsulation for overlays
• Unifies VXLAN, NVGRE, STT formats
• Extensible to support future control planes
• Options infrastructure to carry metadata/ context
for network virtualization & service chaining
• Options use TLV format for flexibility
Motivation for Geneve:
• Metadata (system state, service context)
Example usage for metadata
• Service Chaining: Sharing service context between
service functions e.g. FW, LB, DPI, NAT, VPN
https://datatracker.ietf.org/doc/draft-gross-geneve/
Intel Confidential
19. Network Platforms Group
Step Up Your Server Game
Intel® Solid-State Drives
DC S3700 Series
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.
19
Intel® Xeon®
processor
E5-2697 v3
1Gb Ethernet
Server Adapter
Intel® Xeon®
processor
E5-2670 v2
67% more VMs
SAS 10K
Hard Drive
10Gb Intel® Ethernet
CNA Family
Intel®Xeon®E5-2699 V3
Intel®Xeon®E5-2670 V2
20. Network Platforms Group
Climb to 67% more virtual machines
for more performance—and GREATER ROI
Good
Windows Server* 2012
Intel® Xeon® processor E5-2670 v2
(running at optimal 57% utilization)
SAS hard disk drives
1GbE network connection
6,000
Microsoft
Exchange* users1
12
Microsoft
Exchange* VMs1
12
Database VMs1
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.
20
21. Network Platforms Group
Better
Windows Server* 2012 R2
Intel® Xeon® processor
E5-2697 v3
SAS hard disk drives
1GbE network connection
17%
more Microsoft
Exchange* users1
(7,000)
14
Microsoft
Exchange* VMs1
14
Database VMs1
+
…Upgrade Compute Power
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.
21
22. Network Platforms Group
Best
Windows Server* 2012 R2
Intel® Xeon® processor E5-2697 v3
Intel® Solid-State Drives
DC S3700 Series
10Gb Intel® Ethernet
Converged Network Adapter
Family
+
67%
more Microsoft
Exchange* users1
(10,000)
20
Microsoft
Exchange* VMs1
20
Database VMs1
…Upgrade Disk I/O &
Upgrade Network I/O
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.
+
22
23. Placeholder Footer Copy / BU Logo or Name Goes Here
Designed To Work Better
For HP Platforms
24. SKU: 71659-B21
SKU: 665249-B21 SKU: 727055-B21
Intel® Ethernet
• Uses existing, low-cost Cat 6/Cat/7 cables
• Cabling lengths from 1 m to 100 m
• Backwards compatible to1G network
infrastructures
• Allows for Auto negotiation between 1G to 10G:
“As-needed infrastructure upgrade”
• Pluggable solutions for Direct Attach (DA)
and LC Fiber Optic Cables ( SR / LR )
• SFP+ DA cables up to 7 m
• SR Fiber Optics up to 300 m
• LR Fiber Optics up to 10 km
• Cabling options include:
• Offers a low latency solution Suited
for high speed applications
Intel® Ethernet is validated and optimized on
Intel® Xeon® processors
Best choice for virtualized servers, providing
advanced features including Virtual Machine
Device Queues (VMDq) and Single Root I/O
Virtualization (SR-IOV) to deliver amazing
performance
Built to enable network virtualization and
network functions virtualization
Connecting Server, Storage, and Networking
10GBASE-T
Intel® Ethernet
10GbE SFP+
Intel® Ethernet – “It Just Works”
Why are customers upgrading from 1G to 10G Ethernet?
Server
Virtualization
Server
Performance
Simplified
Networking
Flash in the
Data Center
More virtual Machines
(VMs) are deployed
than
Physical servers
Many VMs now share
one Ethernet port
Even more powerful Intel®
Server processors support
more VMs and drive more
demanding requirements
on network I/O
More processor + more
VMs = Network
bottlenecks
10GbE enables running all
network storage and LAN
traffic over an Ethernet pipe
Simplified networking
Can lower total cost of
ownership (TCO)
Flash non-volatile
memory offers 2x the
read performance of
hard disk drives
(HDDs)
Old spinning HDDs are
often the critical bottleneck
in today’s data centers
Why choose Intel® 10G
Ethernet? Unparalleled supply line, customer support,
and warranty
Designed to support demanding workloads for
enterprise, cloud, and telecommunication service
providers and data centers.
Intel is the #1 Ethernet vendor- having shipped
over 1 billion Ethernet ports
10G Ethernet will be a required of the future
for SDI, Cloud, and Big Data
Component upgrades from Intel® - increase VM density and boost
performanceIntel®Xeon® processor E5-2699 v3
Intel® Solid-StateDrives
(SSD) DC S3700 Series
10Gb Intel® Ethernet
Converged Network Adapters
SAS Hard-Disk Drives
(HDDs) 1G Intel®Ethernet
1G Intel® Ethernet
SAS Hard-Disk Drives
(HDDs)
Best
Better
Good
10,000 Microsoft Exchange* Users
40 Total VMs supported
7,000 Microsoft Exchange* Users
28 Total VMs supported
6,000 Microsoft Exchange* Users
24 Total VMs supported
Intel® Xeon® processor E5-2699 v3
Intel® Xeon® processor E5-2670 v2
26. TRANSFORMING NETWORKING & STORAGE
26
Designed To Work Better for HP Platforms
Broad offering of FlexLOM, Adapters and Mezz Cards with
Flex10Flexibility
Network virtualization Overlay stateless offloads
Acceleration
DPDK optimized for Network Function Virtualization
(NFV)
Intel® Ethernet Flow Director for hardware based
application traffic steering
Better
Together
27. TRANSFORMING NETWORKING & STORAGE
27
Intel® Ethernet Controller XL710 Family (Fortville) Assets Asset Type
Platform Overview Animation
Platform Performance Animation
XL710 Launch Executive Video
XL710 Technical Video
Flow Director Technical Video
XL710 Technical Chit Chat Pod Cast
Network Virtualization IDF Technical
Session Presentation
XL710 Product Page
XL710 Adapter Product Page
Adapter Selector Guide
Controller Selector Guie
Product detail, ordering tool
XL710 Product Brief
XL710 Adapater Product Brief
XL710 Datasheet
Geneve Blog
Xeon E5v3 Launch Site
32. Network Platforms Group
Component Upgrades From Intel
Increase VM Density And Boost Performance
Scenario Base CPU + OS + SSDs +10GbE + SSDs +10GbE
Number of VMs 24 28 28 28 40
CPU Usage 57.0% 34.7% 37.3% 36.9% 50.5%
Network Usage 906.3 Mb/s 880.1 Mb/s 915.5 Mb/s 1,027.8 Mb/s 1,728.4 Mb/s
Total OPM (Orders per Minute) 331,987 360,848 381,943 381,663 532,637
OPM / VM 27,666 25,775 27,282 27,262 26,632
Exchange Latency 13.3 ms 20.9 ms 19.2 ms 12.8 ms 7.1 ms
1 http://www.principledtechnologies.com/Intel/R730_step-up_0415.pdf
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.
Limited Performance Increase
Bottle Neck Limits Performance
Better Together
32
33. Network Platforms Group
Matching Up The Right Technologies
9 Min 10 Sec
36 Sec
Time to Load
* Other names and brands may be claimed as the property of others.
Intel® SSD
DC P3700 Series
PCI Express* Interface
Intel® Ethernet CNA
X710-DA2
Intel® Ethernet CNA
XL710-QDA2
100GB file calculated time to load and transfer times
1 Min 22 Sec
22 Sec
14 Min 40 sec
Transfer Times
40GbE
10GbE
1GbE
Storage Network
15K SAS
Hard Disk Drive
Solid State Drive
NVMe Flash Memory
Spinning Disk
Serial Attached SCSI
Intel® Server Adapter
I350-T2
Source as of Aug 2014: Based on Intel SSD product data sheet Sequential Read and transfer time calculations of a 100GB file over the given Ethernet speed
3 Min 03 Sec
Solid State Drive
NAND Flash Memory
Intel® SSD
DC S3710 Series
SATA Interface
GoodBetterBest
33
34. Network Platforms Group
Benefits
34
Intel® Ethernet Controller and CNA X710 Family
Abstract the network for cloud flexibility with performant network overlays
Features
Intel® Ethernet
CNA X710
Intel® Ethernet
Controller X710
Intel® Ethernet Converged Network Adapter (CNA) X710
Broad Offering of Physical Interfaces
Server I/O Virtualization assistants and by-pass
Network Virtualization Overlay Accelerators and Offloads
Advanced Hardware Traffic Steering
Brings 10 Gigabit Ethernet to the broad market with reduced power,
performance improvements and deployment flexibility
Interfaces for Converged Network Adapters, backplanes and LAN on
Motherboard (LOM)
35. Network Platforms Group
Benefits
35
Intel® Ethernet Controller and CNA X540 Family
Industry first 10GBASE-T single chip solution with integrated MAC + PHY
Features
Intel® Ethernet
CNA X540
Intel® Ethernet
Controller X540
Intel® Ethernet Converged Network Adapter (CNA) X540
Low cost, low power, 10 Gigabit Ethernet (10GbE) performance
Backward compatible with existing 1000BASE-T networks
Optimized for server virtualization
Unified networking delivering LAN, iSCSI, and FCoE
Brings 10 Gigabit Ethernet to the broad market with reduced power,
performance improvements and deployment flexibility
Simplifies the transition to 10GbE through Backward Compatibility with existing
1GbE networks
36. Network Platforms Group
Benefits
36
Intel® Ethernet CNA X520 Family
The Industry Standard for 10GbE Server Connectivity
Features
Intel® Ethernet
CNA X520
Intel® 82599
10 Gigabit Ethernet
Controller
Intel® Ethernet Converged Network Adapter (CNA) X520
Intel® Ethernet CNAs are the industry’s top-selling 10GbE adapters
Low cost, low power, 10 Gigabit Ethernet (10GbE) performance
Optimized for server virtualization
Unified networking delivering LAN, iSCSI, and FCoE
10Gb Ethernet Blade, Direct Attach or Fiber Optic Connectivity
High-volume stable architecture with broad operating system support
Available as CNAs, LOM, and custom mezzanine adapters from most OEMs
37. Network Platforms Group 37
Intel® processors, SSDs, and server adapters are designed and
built to work together seamlessly. Reduce the time you spend
configuring, testing, and tuning your server components. Easily
integrate new systems with existing hardware. By selecting data
center solutions with Intel technologies inside, you can optimize
performance and maximize the value of your refresh.
Learn More
For the full report from Principled Technologies, visit:
www.principledtechnologies.com/Intel/R730_stepup_0415.pdf
Learn more about the Intel technologies that can help you maximize the value of your refresh:
Intel Xeon processor E5-2600 v3 product family
www.intel.com/content/www/us/en/processors/xeon/xeon-processor-e5-family.html
Intel Solid-State Drives (SSDs) Data Center S3700 Series
www.intel.com/content/www/us/en/solid-state-drives/solid-state-drives-ssd.html
Intel Ethernet Converged Network Adapter X520 product family
www.intel.com/content/www/us/en/network-adapters/converged-network-adapters.html
Maximize the Benefits of a Hardware Refresh with
Upgraded Intel® Technologies
Editor's Notes
Geneve = Generic Network Virtualization Encapsulation
VXLAN = Virtual eXtensible LAN
NVGRE = Network Virtualization using Generic Routing Encapsulation
When is Gen9 Platform
256 VSI = VM
Others VSI’s are used for resources like storage etc…
Today IT delivers a network service by utilizing ordered sets of cooperating network applications
Services in dedicated hardware or physical boxes
Service Function Chain (SFC) is an ordered sets of services
Network Topology dependent
Inflexible deployment model, requires changing forwarding behavior
NFV is about dynamic provisioning of services
Virtualizing service functions on general purpose servers (VMs)
Services running in VMs on general purpose services
Geneve enables NFV with virtual networks
Independent of physical network topology
Metadata: to share context between service functions
Server virtualization
Multiple virtual servers on a physical server
Each VM has illusion it is running as a physical server
Network virtualization
Multiple virtual networks on a physical network
Each virtual network has illusion it is running as a physical network
For data center virtualization, two key issues must be addressed.
First, address space separation between tenants must be supported.
Second, it must be possible to place (and migrate) VMs anywhere in the data center, without restricting VM addressing to match the subnet boundaries of the underlying data center network.
Network Becomes the Bottleneck without the proper offloads
Intel® Technologies: Working Together for a Greater ROI
Intel® processors, SSDs, and server adapters are designed and built to work together seamlessly. Reduce the time you spend configuring, testing, and tuning your server components. Easily integrate new systems with existing hardware. By selecting data center solutions with Intel technologies inside, you can optimize performance and maximize the value of your refresh.
Meanwhile, the changes can also help your organization accommodate growth and reduce IT spending. For example, by running 67 percent more VMs on each physical server, you can consolidate your existing environment and launch
new applications while controlling capital expenditures and minimizing operating expenditures. By accommodating more mailbox users on each server, you can support organizational growth without having to purchase new hardware. Meanwhile, controlling the number of physical servers in your data center allows you to control management and software licensing costs as well as power, cooling, and real estate expenses.
Move Beyond Base Configurations
As the needs of your business grow, so must the power of your infrastructure. While purchasing new base-model hardware can give you some performance gains, upgrading key components can maximize gains while enabling you to
achieve a strong return on investment. Testing from Principled Technologies shows that upgrading Intel® components can help you hit the sweet spot for your IT and business goals while keeping costs under control.
Periodically refreshing your data center infrastructure is vital to maximizing the value of technology for your organization. Upgrading servers and storage systems to new models with the latest component technologies can offer the performance you need to support new applications, accommodate a growing number of users, and deliver better services.
If you simply replace your existing previous-generation servers with the latest base models, you will likely see a gain in performance. But selecting the right individual component upgrades as you refresh can produce significant performance improvements that deliver an even stronger return on investment.
Understanding which components to select can help you make the most of the upgrade. Recent tests from Principled Technology labs measure the impressive performance gains you can achieve by moving to systems with the new Intel® Xeon® processor E5 v3 family and running the Microsoft Windows Server 2012 R2* operating system.
Move to the Latest Processor Technology
The Principled Technology tests showed that refreshing servers and storage systems with the latest processor technology, and using the latest operating system version, can substantially increase database performance, enable you to support more e-mail users, and increase virtual machine (VM) density. Compared with previous generation infrastructure (see Figure 1), a refreshed processor and operating system infrastructure that uses the Intel Xeon processor E5-2699 v3 and the Microsoft Windows Server 2012 R2 operating system can deliver 9 percent more orders per minute (OPM) with Microsoft SQL Server 2012*, support 17 percent more Microsoft Exchange Server 2012* mailbox users, and support 16 percent more VMs per server.
Move Beyond Base Configurations
As the needs of your business grow, so must the power of your infrastructure. While purchasing new base-model hardware can give you some performance gains, upgrading key components can maximize gains while enabling you to
achieve a strong return on investment. Testing from Principled Technologies shows that upgrading Intel® components can help you hit the sweet spot for your IT and business goals while keeping costs under control.
Tests also show that replacing traditional hard disk drives (HDDs) with the Intel® Solid-State Drive (SSD) Data Center S3700 Series and implementing 10 gigabit Ethernet connectivity with the Intel® Ethernet Converged Network Adapter X520 product family can further enhance performance. With refreshed, optimized servers, you can provide a more responsive user experience and support growth while reducing costs.
Upgrade Drives and Network Adapters
Upgrading the storage drives and networking technology can provide additional performance gains above the 9 to 16 percent gain for upgrading the processor and operating system, especially for I/O-intensive workloads. Without those upgrades, you might experience latency from bottlenecks.
The Principled Technologies testing team replaced traditional SAS HDDs on the external storage array with the Intel SSD Data Center S3700 Series, and upgraded networking from a 1 GbE network connection to a 10 GbE connection using the Intel Ethernet Converged Network Adapter X520 product family on both the hypervisor server and storage server. With those changes in addition to the processor and OS upgrades, the new server delivered 60 percent better database performance, 67 percent more user mailboxes, and 67 percent more VMs compared with the previous-generation environment.
Geneve = Generic Network Virtualization Encapsulation
VXLAN = Virtual eXtensible LAN
NVGRE = Network Virtualization using Generic Routing Encapsulation
Investing in the right combination of upgrades is critical
Upgrading OS and the server only increased the total number of VMs supported from 24 to 28
BUT… upgrading to SSDs or 10GbE did not increase the total number of VMs due to the network bottlenecking the SSDs and the storage bottlenecking the 10GbE
Upgrading both SSDs and 10GbE increased the total number of VMs supported from 24 to 40
Achieve a Strong Return on Investment
Making these relatively easy upgrades can help you achieve a strong—and fast—return on investment. Your employees, partners, and customers benefit immediately from performance improvements, gaining a more responsive experience.
P3700 Data Sheet: Performance measured using Iometer* with 128 KB (131,072 bytes) of transfer size with Queue Depth 128. Power mode set at 25W.
Sequential Read (up to) 2,800MB/s
S3500 Data Sheet: Performance measured using Iometer* with 128 KB (131,072 bytes) of transfer size with Queue Depth 32.
Sequential Read (SATA 6Gb/s) 500MB/s
Need SAS Test Details:
http://i.dell.com/sites/doccontent/shared-content/data-sheets/en/Documents/Drive-Characteristics-and-metrics-for-Nov-2013.pdf
2.5” HDD SAS Mission Critical 15K 300GB
Estimated Sequential Read (SAS) 183MB/s
Estimated Sequential Write (SAS) 183MB/s
Why 10GBASE-T?
The industry transition to 10 Gigabit Ethernet (10GbE) is spurred on by 10GBASE-T, which delivers the following benefits:
• Reduced deployment costs to bring 10GbE to the broad market through integrated 10GbE, lower cable costs, and flexible cable lengths.
• Simplified transition to 10GbE from backward compatibility with the switching and cabling infrastructures of existing GbE networks.
• Reduced complexity through industry-standard twisted-pair copper cabling and support for advanced I/O virtualization and unified networking.
The Intel® Ethernet Controller X540 is Intel’s latest 10 Gigabit Ethernet controller and the industry’s first fully integrated 10GBASE-T controller, combining MAC and PHY into a single chip. Intel designed this controller for low-cost, low-power 10GBASE-T LAN on Motherboard (LOM) and converged network adapter (CNA) implementations. The Intel Ethernet Controller X540 includes advanced I/O virtualization (virtual machine device queues and support for PCI-SIG Single Root I/O Virtualization), storage over Ethernet (including NFS, iSCSI, and Fibre Channel over Ethernet), and support for I/O enhancements on the new Intel Xeon processor E5 family.