DOCSLIB.ORG

IBM Grid Technical Strategy & Related Research Projects

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
IBM Grid Technical Strategy & Related Research Projects Grid Computing IBM Grid Technical Strategy & Related Research Projects Jean-Pierre Prost Senior Technical Staff Member IBM Montpellier, France © 2005 IBM Corporation Grid Computing Outline Grid and Virtualization Examples of Commercial Grids Grid Components and IBM Grid Strategic Directions IBM Grid Research Projects 2 IBM Grid Technical Strategy & Research Projects © 2005 IBM Corporation Grid Computing Grid Computing: Enabling an On Demand Infrastructure BeforeAfter GridGrid ApplicationApplication ApplicationApplication ApplicationApplication ApplicationApplication ...... ApplicationApplication 1 1 2 2 3 3 4 4 N N Grid Middleware “Virtualized”“Siloed” infrastructure: architecture: • Creates• Higher a virtual costs application (capital + operational)operating, storthrouagegh limited& collaboration pooling of environmentIT assets across silos • Virtualizes• Challenging application cross services organization execution collaboration • Dynamically• Limited fulfills responsiveness requests over due a virtualto more pool manual of system scheduling resources and provisioning • Offers an adaptive, self-managed operating environment that offers high availability 3 IBM Grid Technical Strategy & Research Projects © 2005 IBM Corporation Grid Computing Grid and Virtualization Virtualize Outside the Enterprise IBM is focused on solutions that help clients realize value from the full spectrum of grid computing solutions Virtualize the Enterprise Suppliers, partners, customers and external resources Virtualize Unlike Resources Enterprise wide Grids, Information Insight, and Virtualize Like Global Fabrics Resources Heterogeneous systems, storage, and networks; Application-based Grids Cluster Single System (Partitioning) Simple Sophisticated (2-4) (4+) Homogenous systems, storage, and networks Homogenous Heterogeneous Single Organization Multiple Organizations Tightly Coupled Loosely Coupled 4 Grid Computing Realizing Increasing Levels of Business Value Virtualize Outside the Enterprise Virtualize the Enterprise IT Simplification for Virtualize Unlike Enterprise Optimization Resources Asset Utilization Workload Prioritization Virtualize Like Resources Infrastructure Flexibility Enabling Business Resiliency Cluster Aggregating Information Business Insight Single System (Partitioning) and Collaboration Simple Sophisticated (2-4) (4+) Application Acceleration Time to Results and Higher Quality 5 Grid Computing Characteristics of Grid Evolution Virtualize Outside the Enterprise Virtualize the Enterprise Virtualize Unlike Resources • Multiple Organizations Virtualize Like • Meta-scheduling Resources • Workload Prioritization • SLA Attainment • Advanced Scheduling • Orchestration • Multiple Apps/Resources • Information On Demand Cluster • Integrated Security Single System • Provisioning (Partitioning) • Workload Management Simple Sophisticated (2-4) (4+) • Information Virtualization • Cluster Scheduling • Single Application • Homogenous • Storage/File Virtualization 6 Grid Computing Across the Spectrum: IBM Customer References Virtualize Outside the –Range of technologies based on best fit for task Enterprise –Heterogeneous implementations address customer requirements Virtualize the –Open approach using standards-based tools, Enterprise IBM products and partner products ZJ Transportation Virtualize Unlike Resources China Grid Virtualize Like Resources Cluster Single System (Partitioning) Simple Sophisticated (2-4) (4+) National Digital Mammography Archive IBM Northeast US Financial Services 7 Grid Computing EngineeringEngineering andand DesignDesign MAGNA STEYR AutomotiveAutomotive Challenge Too much time required to effectively run clash tests between complex sub- assemblies which impacts quality of the end product and getting the product to market on time. Benefits: Significant performance improvement (72 – 4hrs) Too much administrative time required Risk and Error Reduction from design engineers Cost Reduction Solution Increased accuracy of data improves quality and reduces late changes) Grid enabled clash environment. IBM developed code and services using Improved Time to Market Platform Computing LSF w/ Dassault Faster evaluation of design alternatives Systemes CATIA & ENOVIA DMU More accurate and timely product applications providing clash detection development analysis. “Grid“Grid technology technology from from IBM IBM and and Pl Platformatform Computing Computing reduced reduced the the ti timeme required required for for our our clash clash testing testing fromfrom 72 72 – – 4 4 hours hours and and contributed contributed significantly significantly to to enhancing enhancing our our design design quality, quality,”” said said Dr. Dr. Heinz Heinz Mayer,Mayer, MAGNA MAGNA STEYR. STEYR. 8 Grid Computing Enterprise Optimization Audi/MSC/IBM Proof of Concept Automotive Distributed Simulation Data and Process Management Challenge Reuse defined simulation schemes to enhance engineer productivity and innovation. Improvement of integrated, secure and collaborative design environment across the distributed enterprise Technology Benefits: Shared access for engineers from all enterprise sites to all A robust, scalable multi site Simulation Data and Process simulation data from cars. Central coordination point for data Management (SDM) infrastructure and content Usage of standard applications to design a multi site collaboration environment Data sharing and data exchange capability for multi site collaboration to be implemented between Audi and SEAT Business Opportunities: Solution Flexibility to handle more projects, scale up engineer and design work An information grid enabling collaborative engineering Global view of all CAE data to assess entire vehicle at any Multi site engineering data sharing based on: given time in the design process – MSC.SimManager Reuse design work, maintain historical data on all analysis – IBM eServer and TotalStorage generated – IBM Data Management replication and federation More accurate and timely product development technology based on WebSphere Information Integrator Improved productivity of the Engineering Analysis Office and WebSphere MQ Improved Time to Market and faster evaluation of design – IBM integration services alternatives in collaborative tasks “Collaborating with our engineering team at SEAT will become more and more important due to the common activities inside the Audi “Collaborating with our engineering team at SEAT will become more and more important due to the common activities inside the Audi brand group. The new technology we are developing with IBM and MSC.Software will help us collaborate more effectively and build a real brand group. The new technology we are developing with IBM and MSC.Software will help us collaborate more effectively and build a real global engineering collaboration, reducing travel and taking both time and cost out of the product development process” global engineering collaboration, reducing travel and taking both time and cost out of the product development process” -- Dr.-Ing. Ulrich Widmann, head of functional design crash, occupant safety, NVH analysis AUDI AG -- Dr.-Ing. Ulrich Widmann, head of functional design crash, occupant safety, NVH analysis AUDI AG 9 Grid Computing Business Analytics Sal. Oppenheim Financial Challenge: ► Sal. Oppenheim is one of the leading independent private banks in Europe. Sal. Oppenheim is Technology Benefit upgrading its pricing system for equity derivative trading. High performance grid cluster based on latest ► Traditional dedicated compute resources handle Platform Symphony grid software and IBM the dynamic structure of the large amount of risk cluster technologies. and pricing calculations. ► Unbalanced peak performance sized solutions Solution is integrated into the IT management would be too inefficient in terms of price concept of Sal. Oppenheim performance. ► The grid has to deliver a service level based risk Business Benefit and price calculation solution. Product development for equity derivatives Solution: Flexible high speed grid solution which can ► one of a kind flexible grid system developed by Sal. Oppenheim, Platform grow with the demands of trading volumes. Computing, the IBM business partner Morse and the IBM Grid computing team. The service level based risk and pricing solution guarantees performance and availability of the – IBM x336 critical business application – IBM Cluster Management Software – Platform Symphony “„Our financial trading teams are working in a very volatile and extremely demanding market. Processing of large amounts of data and numerical intensive simulations in highest quality and speed decide between win and loss. With the introduction of the grid solution we believe to be able to fulfill the demand we are getting from the business side. With the grid solution, we will have in the mid-term an IT infrastructure that will allow us to react flexibly and grow as demand grows with acceptable cost“, said Arno Radermacher, Head of IT Investment10 Banking at Sal. Oppenheim. Grid Computing Business Analytics Hewitt Associates LLC Financial Challenge Create Grid Computing environment to: – Contain expenses for CalcEngine valuations – Maintain or improve availability, response time & scalability Benefits: – Insure personal-data security • Efficiently uses of the combined – Capitalize on existing application code processing power of their – Cooperate with z/OS Sysplex CICS Calling heterogeneous environment Environment • Experienced an immediate 10% faster – Enable smooth
Load more

Recommended publications
  • Platform LSF Foundations Chapter 1
    Platform LSF Version 9 Release 1.3 Foundations SC27-5304-03 Platform LSF Version 9 Release 1.3 Foundations SC27-5304-03 Note Before using this information and the product it supports, read the information in “Notices” on page 21. First edition This edition applies to version 9, release 1 of IBM Platform LSF (product number 5725G82) and to all subsequent releases and modifications until otherwise indicated in new editions. Significant changes or additions to the text and illustrations are indicated by a vertical line (|) to the left of the change. If you find an error in any Platform Computing documentation, or you have a suggestion for improving it, please let us know. In the IBM Knowledge Center, add your comments and feedback to any topic. You can also send your suggestions, comments and questions to the following email address: [email protected] Be sure include the publication title and order number, and, if applicable, the specific location of the information about which you have comments (for example, a page number or a browser URL). When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you. © Copyright IBM Corporation 1992, 2014. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Chapter 1. IBM Platform LSF: An Job submission .............12 Overview ..............1 Job scheduling and dispatch .........13 Introduction to IBM Platform LSF .......1 Host selection .............15 LSF cluster components...........2 Job execution environment .........15 Chapter 2.
    [Show full text]
  • IBM Platform Computing Cloud Service Ready to Use Platform LSF & Symphony Clusters in the Softlayer Cloud
    Platform Computing IBM Platform Computing Cloud Service Ready to use Platform LSF & Symphony clusters in the SoftLayer cloud February 25, 2014 1 © 2014 IBM Corporation Platform Computing Agenda v Mapping clients needs to cloud technologies v Addressing your pain points v Introducing IBM Platform Computing Cloud Service v Product features and benefits v Use cases v Performance benchmarks 2 © 2014 IBM Corporation Platform Computing HPC cloud characteristics and economics are different than general-purpose computing • High-end hardware and special purpose devices (e.g. GPUs) are typically used to supply the needed processing, memory, network, and storage capabilities • The performance requirements of technical computing and service-oriented workloads means that performance may be impacted in a virtualized cloud environment, especially when latency or I/O is a constraint • HPC cluster/grid utilization is usually in the 70-90% range, removing a major potential advantage of a public cloud service provider for stable workload volumes HPC Workloads Recommended for Private Cloud HPC Workloads with Best Potential for Virtualized Public & Hybrid Cloud Primary HPC Workloads 3 © 2014 IBM Corporation Platform Computing IBM’s HPC cloud strategy provides a flexible approach to address a variety of client needs Private Hybrid Public Clouds Clouds Clouds Evolve existing infrastructure to Enable integrated HPC Cloud to enhance approach to improve Access additional responsiveness, HPC cost and HPC capacity with flexibility, and capability variable cost model
    [Show full text]
  • Platform LSF Quick Reference IBM Platform LSF 9.1.2 Quick Reference
    Platform LSF Version 9 Release 1.2 Quick Reference GC27-5309-02 Platform LSF Version 9 Release 1.2 Quick Reference GC27-5309-02 Note Before using this information and the product it supports, read the information in “Notices” on page 9. First edition This edition applies to version 9, release 1 of IBM Platform LSF (product number 5725G82) and to all subsequent releases and modifications until otherwise indicated in new editions. Significant changes or additions to the text and illustrations are indicated by a vertical line (|) to the left of the change. If you find an error in any Platform Computing documentation, or you have a suggestion for improving it, please let us know. Send your suggestions, comments and questions to the following email address: [email protected] Be sure include the publication title and order number, and, if applicable, the specific location of the information about which you have comments (for example, a page number or a browser URL). When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you. © Copyright IBM Corporation 1992, 2013. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Notices ...............9 Privacy policy considerations ........11 Trademarks ..............11 © Copyright IBM Corp. 1992, 2013 iii iv Platform LSF Quick Reference IBM Platform LSF 9.1.2 Quick Reference Sample UNIX installation directories Daemon error log files Daemon error log files are stored in the directory defined by LSF_LOGDIR in lsf.conf.
    [Show full text]
  • Administering IBM Platform LSF Chapter 1
    Platform LSF Version 9 Release 1.3 Administering Platform LSF SC27-5302-03 Platform LSF Version 9 Release 1.3 Administering Platform LSF SC27-5302-03 Note Before using this information and the product it supports, read the information in “Notices” on page 827. First edition This edition applies to version 9, release 1 of IBM Platform LSF (product number 5725G82) and to all subsequent releases and modifications until otherwise indicated in new editions. Significant changes or additions to the text and illustrations are indicated by a vertical line (|) to the left of the change. If you find an error in any Platform Computing documentation, or you have a suggestion for improving it, please let us know. In the IBM Knowledge Center, add your comments and feedback to any topic. You can also send your suggestions, comments and questions to the following email address: [email protected] Be sure include the publication title and order number, and, if applicable, the specific location of the information about which you have comments (for example, a page number or a browser URL). When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you. © Copyright IBM Corporation 1992, 2014. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Chapter 1. Managing Your Cluster . 1 Job Directories and Data..........441 Working with Your Cluster .........1 Resource
    [Show full text]
  • Experimental Study of Remote Job Submission and Execution on LRM Through Grid Computing Mechanisms
    Experimental Study of Remote Job Submission and Execution on LRM through Grid Computing Mechanisms Harshadkumar B. Prajapati Vipul A. Shah Information Technology Department Instrumentation & Control Engineering Dharmsinh Desai University Department Nadiad, INDIA Dharmsinh Desai University e-mail: [email protected], Nadiad, INDIA [email protected] e-mail: [email protected], [email protected] Abstract —Remote job submission and execution is The fundamental unit of work-done in Grid fundamental requirement of distributed computing done computing is successful execution of a job. A job in using Cluster computing. However, Cluster computing Grid is generally a batch-job, which does not interact limits usage within a single organization. Grid computing with user while it is running. Higher-level Grid environment can allow use of resources for remote job services and applications use job submission and execution that are available in other organizations. This execution facilities that are supported by a Grid paper discusses concepts of batch-job execution using LRM and using Grid. The paper discusses two ways of middleware. Therefore, it is very important to preparing test Grid computing environment that we use understand how a job is submitted, executed, and for experimental testing of concepts. This paper presents monitored in Grid computing environment. experimental testing of remote job submission and Researchers and scientists who are working in higher- execution mechanisms through LRM specific way and level Grid services and applications do not pay much Grid computing ways. Moreover, the paper also discusses attention to the underlying Grid infrastructure in their various problems faced while working with Grid published work.
    [Show full text]
  • Turbo-Charging Open Source Hadoop for Faster, More Meaningful Insights
    Turbo-Charging Open Source Hadoop for Faster, more Meaningful Insights Gord Sissons Senior Manager, Technical Marketing IBM Platform Computing [email protected] Agenda • Some Context – IBM Platform Computing • Low-latency scheduling meets open-source • Breakthrough performance • Multi-tenancy (for real!) • Cluster-sprawl - The elephant in the room • Side step the looming challenges IBM Platform Computing De facto standard for commercial high- • Acquired by IBM in 2012 performance computing Powers financial • 20 year history in high-performance computing analytics grids for 60% of top investment banks • 2000+ global customers Over 5 million CPUs • 23 of 30 largest enterprises under management • High-performance, mission-critical, extreme scale Breakthrough performance in Big • Comprehensive capability Data analytics Technical Computing - HPC Platform LSF Scalable, comprehensive workload Family management for demanding heterogeneous environments Simplified, integrated HPC management Platform HPC software bundled with systems Analytics Infrastructure Software High-throughput, low-latency compute Platform and data intensive analytics Symphony Family • An SOA infrastructure for analytics • Extreme performance and scale • Complex Computations (i.e., risk) • Big Data Analytics via MapReduce Our worldview – shaped by time critical analytics • Financial firms compete on the their ability to maximize use of capital • Monte-Carlo simulation is a staple technique for simulating market outcomes • Underlying instruments are increasingly complex •
    [Show full text]
  • Platform LSF Command Reference Chapter 99
    Platform LSF Version 9 Release 1.3 Command Reference SC27-5305-03 Platform LSF Version 9 Release 1.3 Command Reference SC27-5305-03 Note Before using this information and the product it supports, read the information in “Notices” on page 521. First edition This edition applies to version 9, release 1 of IBM Platform LSF (product numbers 5725G82 and 5725L25) and to all subsequent releases and modifications until otherwise indicated in new editions. Significant changes or additions to the text and illustrations are indicated by a vertical line (|) to the left of the change. If you find an error in any Platform Computing documentation, or you have a suggestion for improving it, please let us know. In the IBM Knowledge Center, add your comments and feedback to any topic. You can also send your suggestions, comments and questions to the following email address: [email protected] Be sure include the publication title and order number, and, if applicable, the specific location of the information about which you have comments (for example, a page number or a browser URL). When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you. © Copyright IBM Corporation 1992, 2014. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Chapter 1. bacct ...........1 Chapter 23. blcollect ........165 Chapter 2. badmin ..........15 Chapter 24. blcstat .........167 Chapter 3. bapp ...........31 Chapter 25. blhosts .........169 Chapter 4.
    [Show full text]
  • Cabot Partners Cabot Group, Inc
    How IBM Platform LSF Architecture Accelerates Technical Computing Sponsored by IBM Srini Chari, Ph.D., MBA September, 2012 mailto:[email protected] Executive Summary Advances in High Performance Computing (HPC) have resulted in dramatic improvements in application processing performance across a wide range of disciplines ranging from engineering, manufacturing, finance - risk analysis and revenue management to geological, life and earth sciences. This mainstreaming of technical computing has driven solution providers towards innovative solutions that are faster, scalable, reliable, and secure. But these mission critical technical computing solutions are further challenged with reducing cost and managing complexity. Further, data explosion has morphed application workloads in traditional technical computing www.cabotpartners.com environments from compute intensive to both compute and data intensive. Also, there continues to be , an unrelenting appetite and need to increasingly solve problems that evolve, grow large and are complex, further challenging the scale of technical computing processing environments. These newer domains cover a broad range of emerging applications including fraud detection, anti-terrorist analysis, social and biological network analysis, semantic analysis, financial and economic modeling, drug discovery and epidemiology, weather and climate modeling, oil exploration, and power grid management1. Although most technical computing environments are quite sophisticated, several IT organizations are still challenged to fully take advantage of available processing capacity to adequately address newer business needs. For these organizations, effective resource management and job submission that meets stringent service level agreement (SLA) requirements across multiple departments because sharing IT infrastructure is an extremely complex process. They require higher levels of shared infrastructure utilization and better application processing throughput, while keeping costs lower.
    [Show full text]
  • Installing Platform LSF on UNIX and Linux Chapter 1
    Platform LSF Version 9 Release 1.2 Installing on UNIX and Linux SC27-5314-02 Platform LSF Version 9 Release 1.2 Installing on UNIX and Linux SC27-5314-02 Note Before using this information and the product it supports, read the information in “Notices” on page 39. First edition This edition applies to version 9, release 1 of IBM Platform LSF (product number 5725G82) and to all subsequent releases and modifications until otherwise indicated in new editions. Significant changes or additions to the text and illustrations are indicated by a vertical line (|) to the left of the change. If you find an error in any Platform Computing documentation, or you have a suggestion for improving it, please let us know. Send your suggestions, comments and questions to the following email address: [email protected] Be sure include the publication title and order number, and, if applicable, the specific location of the information about which you have comments (for example, a page number or a browser URL). When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you. © Copyright IBM Corporation 1992, 2013. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Chapter 1. Example installation directory Chapter 8. install.config .......19 structure ..............1 About install.config............19 Parameters ..............19 Chapter 2. Plan your installation ....3 EGO in the LSF cluster ...........4 Chapter 9. slave.config ........31 EGO_DAEMON_CONTROL.........32 Chapter 3.
    [Show full text]
  • Platform LSF™ Foundations
    Platform LSF™ Foundations Platform LSF Version 7.0 Update 5 Release date: July 2009 Last modified: July 17, 2009 Copyright © 1994-2009 Platform Computing Inc. Although the information in this document has been carefully reviewed, Platform Computing Corporation (“Platform”) does not warrant it to be free of errors or omissions. Platform reserves the right to make corrections, updates, revisions or changes to the information in this document. UNLESS OTHERWISE EXPRESSLY STATED BY PLATFORM, THE PROGRAM DESCRIBED IN THIS DOCUMENT IS PROVIDED “AS IS” AND WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT WILL PLATFORM COMPUTING BE LIABLE TO ANYONE FOR SPECIAL, COLLATERAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING WITHOUT LIMITATION ANY LOST PROFITS, DATA, OR SAVINGS, ARISING OUT OF THE USE OF OR INABILITY TO USE THIS PROGRAM. We'd like to hear You can help us make this document better by telling us what you think of the content, organization, and usefulness of the information. from you If you find an error, or just want to make a suggestion for improving this document, please address your comments to [email protected]. Your comments should pertain only to Platform documentation. For product support, contact [email protected]. Document This document is protected by copyright and you may not redistribute or translate it into another language, in part or in whole. redistribution and translation Internal You may only redistribute this document internally within your organization (for example, on an intranet) provided that you continue redistribution to check the Platform Web site for updates and update your version of the documentation.
    [Show full text]
  • Running Jobs with IBM Platform LSF
    Running Jobs with IBM Platform LSF IBM Platform LSF Version 8.3 May 2012 Copyright Note: Before using this information and the product it supports, read the information in Notices on page 49. This edition applies to version 8, release 3, modification 0 of IBM Platform LSF (product number 5725G82) and to all subsequent releases and modifications until otherwise indicated in new editions. © Copyright Platform Computing Inc., an IBM Company 1992, 2012. US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents 1 About IBM Platform LSF ............................................................................................................. 5 Clusters, jobs, and queues ............................................................................................. 6 Hosts ............................................................................................................................... 9 LSF daemons ................................................................................................................ 11 Batch jobs and tasks ..................................................................................................... 14 Host types and host models .......................................................................................... 16 Users and administrators .............................................................................................. 17 Resources ....................................................................................................................
    [Show full text]
  • HP HPC Linux Value Pack
    QuickSpecs HP HPC Linux Value Pack Overview HP HPC Linux Value Pack V4.1 (HPC-LVP) is a specially-priced software bundle that is specifically designed for the development, use, and management of high performance applications running on HP servers and workstations. HPC-LVP includes Platform LSF, software, which delivers a robust set of workload management capabilities for distributed computing environments. HPC-LVP also includes a web-based interface, workload monitoring and reporting, a simplified application integration process, and the HP MPI parallelization library. Why use HP HPC Linux Value Pack? Platform LSF Workload Management Powerful, comprehensive, policy driven workload management Topology and resource aware, including GPU scheduling Rich job and system information for ease of use and management Comprehensive workload accounting and reporting MPI integration for quickly launching MPI tasks, precisely tracking MPI job resource usage, and reliable MPI job management. High performance, throughput and utilization Highly available and secure Worldwide HP support HP- MPI Reduces development effort through the use of the same MPI library for multiple architectures on both Linux and Windows Optimizes performance without compromising portability or flexibility. Support for NVIDIA GPUDirect™ with CUDA drivers. Supports a wide range of Operating Systems, interconnects and workload managers, including support for interconnect protocols such as RoCE, RDMA, and OFED. Provides the quality and robustness that come from over a decade of production
    [Show full text]