Go Virtual Solutions
Go Virtual’s deep knowledge about CAE applications, compute systems, and IT infrastructure, allow us to configure a system solution meeting your bussiness needs and requirements. We are able to help you during your entire project ranging from defining the need, propose a solution, justifying the investments, deliver the system, integrate the system, educate and train IT professionals and end users, as well as support the system during its life time.
We work with system solutions including workstations, servers, clusters, and surrounding infrastructure such as networks and storage and administration and management tools. The sizes of the clusters we have delivered ranges from 2 to 2.128 nodes.
Go Virtual Turnkey Cluster Solution
Irrespective of the size of your compute system Go Virtual can always deliver your compute system in a ready-to-run condition called Go Virtual Turnkey Solution. The base is a HP certified platform to which we add software and services from our partners and ourselves.
Before the delivery to the your site we:
- assemble the hardware including rack internal cabling,
- install operating system, queuing system, and other software needed
- apply site specific settings such as IP numbers, host names, DNS settings, etc.
- perform basic queuing system configuration
- perform cluster testing to verify the functionality and the performance of the cluster
After delivery to your site we complete cabling between racks, connect the system to power, connect LAN, boots up the system, and again verify the functionality. A high level system documentation as well as introduction for IT staff and end users is also included if requested/needed .
Workstations, Servers, Compute Clusters – Hewlett-Packard
Hewlett-Packard (HP) is a technology solutions provider to consumers, businesses, and institutions globally. The HP offerings span IT infrastructure, personal computing and access devices, global services, and imaging and printing for consumers, enterprises and small and medium businesses.
Why HP?
- HP is the market leader in High Performance Computing, HPC.
- HP is the market leader in Workstations.
- HP offers the broadest spectrum of HPC-solutions, spanning workgroup computing, enterprise computing to supercomputing. From blades and departmental servers, to systems designed for the engineering enterprise and supercomputing centres, the HP portfolio provides leading performance, solutions support and investment protection.
- HP offers a unique delivery model with factory integration and pre-staging of complete clusters from the HP factory.
- HP has the broadest offerings for graphics intense workloads.
Interconnect
In order to use your licenses efficiently for parallel applications, a good interconnect with low latency and high bandwidth is needed. Go Virtual works with Mellanox and Qlogic to include the best solution for you in your system.
Go Virtual is a partner to Mellanox who provides server and storage switching and software solutions that enable high-performance grid computing within the data centre. Mellanox’s fully integrated family of switching hardware and network virtualization software delivers the highest performance, intelligent backbone for grid computing architectures. Leveraging the InfiniBand standard, Mellanox solutions offer improved performance, utilization and scalability across compute clusters, storage and IP networks.
Why Mellanox?
- Mellanox is the market leader in InfiniBand.
- Mellanox supports InfiniBand solutions for both Windows and Linux environents.
- Mellanox provides the cost efficient multi fabric solution for connecting for example HPC storage systems direct to a Cluster InfinBand network.
- State of the art software for fabric managment.
Operating Systems
The choice of operating system is usually driven from the application needs and the experience of the IT professionals who will operate and manage the compute environment. Go Virtual works with both Linux operating systems (RHEL, SLES) and Windows HPC Server 2008 and can assist in selecting what operating system is best for you.
Cluster Management
The importance of efficiently managing your cluster increases with cluster size. For Windows HPC Server 2008 based environments we work with the integrated tools in the product and for Linux based environments we normally work with HP Cluster Management Utility (HP CMU).
HP CMU is an efficient and robust tool for the management of Linux based nodes in HPC clusters and compute farms. An intuitive and easy to use graphic interface enables centralized console management, monitoring of cluster-wide and node-specific metrics, and software installation. HP CMU enables rapid provisioning and management of multiple images, making it ideal for a site with frequent changes to the software configuration across the cluster or on a subset of the cluster.
Resource Management Systems / Queuing Systems
We have experience from working with many of the resource management systems on the market such as MOAB, Scali, LSF, PBSPro, Windows CCS, Windows HPC Server 2008, Torque, and GridEngine, and are happy to assist you to make certain your resource management of choice works in your compute environment.
Go Virtual recommends MOAB as your scheduling tool.
Why Moab?
- Integrate/Unify management across diverse resources and environments in a cluster
- Linux/Windows hybrid clusters
- Control/Share resource usage between users, groups and projects
- Simplify use, access and control for both users and administrators
- Scalable architecture that is grid ready, compatible and extensible
- Automate tasks and processes to streamline job turnaround and reduce administrative burden
MOAB is developed and distributed by Go Virtual’s partner Adaptive Computing. Incorporated in 2001, Adaptive Computing Inc. is a leading provider of workload and resource management software and services for cluster, grid and utility-based computing environments. Adaptive Computing draws on more than a decade of industry experience to provide market defining products and services to thousands of customers. Clients include many of the largest and most complex cluster and grid sites in the world. Top commercial, government & academic sites have licensed Moab® to maximize performance and value, ease management burden and create competitive technical advantages for themselves.
Storage
The amount of data generated usually grows with the size of your cluster and it is important to have a storage solution with sufficient performance (MB/s) and capacity (TB) to handle the data generated. As part of the compute system design we propose a storage solution matching your requirements and needs. For the smaller clusters it may be sufficient with a standard NAS server while for larger clusters a segmented or parallel storage solution may be needed. Go Virtual works with parallel storage from Panasas.
HPC Parallel Storage – Panasas
Panasas sells completely integrated software/hardware storage systems and related services. The Panasas Parallel Storage Cluster is designed for engineering and scientific customers using clusters that demand a highly scalable, yet easily manageable storage network
Why Panasas?
- Record-setting performance: up to 7X random I/O performance and 30X data throughput relative to alternative NFS servers.
- Seamless Scalability from Terabyte to Petabyte scale: Storage capacity can be quickly and easily added and configured in less than 30 minutes. As capacity is added system performance scales linearly.
- Low Total Cost of Ownership: Easy to use appliance-like management of a virtually boundless system.
- Data Availability: Breakthrough ActiveScale Operating Environment for Predictive Self Management and Object-based RAID maximizes system and data availability.