IBM High Performance Services

© 2015 IBM Corporation

IBM High Performance Services July 2015

Chris Porter – WW HPC Cloud Product Manager – [email protected] Jeff Karmiol – SDI Cloud Portfolio Marketing Manager – [email protected]

© 2015 IBM Corporation 2

High Performance and Technical Computing

• Compute and data intensive workloads used for simulation, modeling and analysis •  Parallel applications developed with MPI • High Throughput Computing environments for embarrassingly parallel workloads •  Parallel Service Oriented HPC environments used for Monte Carlo simulations and analytics • Covers a wide range of

–  Homegrown, commercial, and open source applications –  Government and Commercial vertical markets

Banking Aerospace & Defense

Chemical & Petro

Automotive Education & Research

Financial Markets & Insurance

Electronics Life Sciences


High Performance and Technical Computing

Challenges • Unceasing demand for Compute

–  Increased complexity and scalability requirements • Data growth

–  Non-linear growth of data, structured and unstructured

• Rise of shared services / convergence of Compute + Data

–  Rise of Big Data clusters driving need to support any/all distributed applications (heterogeneous compute + data) in a common infrastructure

• Clusters à Grids à HPC Clouds –  From single purpose clusters of static resources

to workload-driven dynamic hybrid HPC resource clouds

Painpoints • Unable to meet business objectives (delay to

market, etc.) •  Existing resources insufficient to meet peek

compute demand –  Long run times on existing cluster or grid –  No access to local technical computing resources

(workstation users) •  Technical resources are expensive and time

consuming to acquire •  The skills/staff to architect and manage a

technical computing infrastructure can be difficult to acquire


Characteristics to Consider

High Performance Infrastructure

Low Latency

Networks High

Performance Storage

Optimized Resource

Use

Capacity to scale

Data Movement

Hybrid Cluster

Security

Software Licenses

Business Advantage Provisioning

Speed

Enterprise Support & Reliability


IBM High Performance Services Ready to Use Clusters in the Cloud

Client and ISV Applications

SoftLayer Infrastructure

Cloud Operations – 24/7 Support

IBM High Performance Services for Data

IBM High Performance Services for Analytics

IBM High Performance Services for HPC

IBM High Performance Services for Hadoop

IBM High Performance Services for EDA


IBM High Performance Services Complete solutions for HPC clouds

Expertise Decades of experience: Deploying and running HPC environments, building clouds, interoperability, consulting, implementation, enterprise grade support

Flexibility Middleware on demand (Platform LSF, Platform Symphony, Spectrum Scale), Hybrid interoperability, bare metal and virtualized, scale compute and storage as needed, hardware customization and specialized components

Performance high-performance infrastructure, bare-metal servers, InfiniBand support, the latest high-performance processor technology, storage designed for HPC environments that can handle thousands of simultaneous connections

Reliability Market proven workload management, multiple data centers, redundant power and networks, clear price structure, proven support

Security Dedicated non-shared servers, isolated networks, encryption, a pedigree of providing secure compute and storage services

Value Time to use, interoperability between on-premise and cloud, software to ensure high compute, storage and application license utilization, resources that meet the specific needs of your applications


Thank you For more information: http://ibm.biz/high-performance-services

IBM High Performance Services

Technology