BIGS001 - In-Memory Low Latency Analytics: Opportunities and Architecture Trends
The modern data driven enterprise pursues an analytics-at-your-fingertips ideal to sift through information quickly, resolve what-if questions interactively, and thus translate opportunities into decisions in real time. Only recently has this ideal become practical with the combination of ever-lower memory prices and large leaps in server memory capacity. These changes have made practical a new generation of hardware and software innovations that are introducing a new era of memory-based analytics and transforming businesses.
This session will:
• Review briefly the major design principles central to in-memory databases and object servers
• Bring out the role of new instructions, caches, and processor capabilities in accelerating in-memory analytics
• Describe joint work between Intel and major database providers in optimizing hardware and software together for ultra-high performance and scale
• Discuss the emerging trends that will further expand the use and reach of in-memory analytics
View webcast and audio recordings.
BIGS002 - Parallelizing Software for Intel® Xeon® and Intel® Xeon Phi™ Processor Based Platforms
Topics include:
• Parallelism Intel® Xeon® & Intel® Xeon Phi™ processor
• Exploiting task parallelism and task parallelism in software
• Fortran, C, and C++ techniques as well as using Message Passing Interface (MPI) in High Performance Computing (HPC) clusters
View webcast and audio recordings.
BIGS004 - Accelerating Hadoop* Performance on Intel® Architecture Based Platforms
This session will cover technical details and best known methods (BKMs) for optimizing big data clusters and Hadoop* workloads on Intel® Xeon® processor E5 v3 based platforms.
Topics include:
• How to deploy, tune and configure the hardware and operating systems for your big data cluster
• How to run an Industry standard benchmark to measure the cluster performance and TCO calculation
• BKMs for optimizing the Hadoop framework configuration and workloads including NoSQL (HBase*, Cassandra) and in-memory analytics frameworks (Spark* / Shark)
View webcast and audio recordings.
DATS002 - Virtualizing the Network to Enable a Software Defined Infrastructure
Intel is heavily investing in products and technologies for network overlays, network function virtualization (NFV) and software defined networking (SDN) to help drive the network hardware architectural transformation to enable Software Defined Infrastructure (SDI). This session will discuss industry trends and evolving standards that impact the datacenter network, provide details on deployment of Network Virtualization Overlays (NVO), and use live demonstrations of various types of applications and workloads. Specific focus on network hardware capabilities to optimize traffic management and performance of network overlays using VXLAN, NVGRE and Geneve.
Topics include:
• Network virtualization using network overlays
• Networking equipment and server configuration requirements for LACP LAGs, tunnel end-points and gateways
• What Intel is doing to enable network overlay technology like NVGRE, VXLAN and Geneve
• Configuration and benchmark testing examples
View webcast and audio recordings.
DATS003 - Software Application Development Using Non-volatile Memory DIMMs
Non-Volatile DIMMs (NVDIMMs) and similar technologies emerging in the industry offer compelling benefits to applications by allowing direct load/store access to Persistent Memory. However, that direct access requires application developers to think differently about how they allocate and manage memory. This session presents the Persistent Memory programming challenges, and APIs available to meet those challenges.
Topics include:
• Overview of Persistent Memory programming
• Walkthrough of key programming challenges
• Introduction to Persistent Memory APIs
• An in-depth example using Persistent Memory APIs
View webcast and audio recordings.
Topics in this session include:
• Update on 2014 Communication platforms including new capabilities
• Update on the Intel® Xeon® Processor E5-2600 v3 Product Family with the Intel® Communications Chipset 89xx Series as the latest Communication Infrastructure platform for Telco and Cloud Service Providers
• Highlight new communication platform usage models and capabilities including SSL and Compression
• Provide guidance on integrating and using technologies in the platform including Intel® QuickAssist and QoS capabilities for cloud applications that use SSL and compression
• Provide guidance on communication platform opportunities for 2014+
View webcast and audio recordings.
Ceph is an open-source, massively scalable, software-defined storage system which provides object, block and file system storage in a single platform. It runs on standard high volume servers. It saves cost while providing flexibility to expand capacity on demand. This session presents compute, network, storage and Ceph software optimizations on Intel® Architecture based platforms to deliver high performance scale out solution to address wide variety of cloud workloads including big data, virtual machine hosting and virtual desktop deployments.
Topics include:
• Overview of Ceph distributed storage system and usages
• Ceph end to end reference architecture on Intel® architecture platforms
• In-depth review of compute, networking, storage and Ceph software optimizations to deliver scalable platform including caching, erasure coding, hashing optimizations
• Best practices for deploying large scale Ceph deployments on Intel Architecture based platforms
View webcast and audio recordings.
Hear about the strategic pillars of the Intel Data Center Group’s initiatives to enhance the data center and network security. Focus will be on the powerful capabilities and use models of today and a look ahead to where Intel is investing to enhance enterprise and cloud security.
Topics include:
• Brief review of the types of security limitations in today’s cloud use models—controls, compliance, data protection challenges
• Showcase Intel Data Center Group and Cloud Security strategy directions
• Highlight the directions, technologies and solution spaces that Intel is pursuing for securing the data center and cloud, with a focus on data protection, chain of trust and enhanced network security capabilities
View webcast and audio recordings.
DATS011 - Dynamic Service Chaining with the Intel® Open Network Platform and Wind River
Topics in this session include:
• Explain the Service Provider’s challenges
• Why service chaining using Network Functions Virtualization (NFV) helps solve the challenge
• What challenges exist to dynamically provision a network application
• The Intel® Open Network Platform (Intel® ONP) hardware and software
• Intel ONP in action – Wind River ON Server and service chaining
• Wind River ON Server dynamic service chaining demo using OpenStack* and Open Daylight* technologies
View webcast and audio recordings.
SFTS005 - Oracle* Exalytics in the Speed of Thought: Extreme Scaling on Intel® Xeon® Processor E7
Oracle* Exalytics is an optimized system for in-memory business intelligence analytics and enterprise performance management. Enabling extreme performance at the “speed of thought” depends on highly optimized and scalable hardware & software foundation. Hear how Intel and Oracle are working together to deliver such optimized and scalable foundation on the Intel® Xeon® processor E7.
Topics include:
• Oracle Exalytics/Essbase overview and performance requirements
• Optimization methodology and technical insights for exploiting Intel Xeon processor E7 capabilities to improve performance and scalability
• Usage of Intel® VTune™ profiler to resolve scaling bottlenecks