November 20, 2006
Voltaire has announced new scalable
fabric management software that is designed to deliver peak performance
for high performance clusters and grids. GridVision Enhanced is a
server-based version of Voltaire's InfiniBand fabric management
software that, according to company, delivers high performance and
speeds deployment of large InfiniBand fabrics consisting of hundreds or
thousands of servers. Voltaire claims that GridVision Enhanced allows
users to bring up a multi-thousand node cluster in less than one minute.
and scalable fabric management software is a critical element for grids
and supercomputers -- such as those in the Top500 -- that are scaling
to hundreds and thousands of servers. With petaflop-scale computing in
sight, it is important to be able to scale both hardware and software
performance," said Patrick Guay, senior vice president of marketing,
Voltaire. "GridVision Enhanced offers all of the mission critical tools
required for an InfiniBand cluster in a commercial-grade, host-based
solution that enables the utmost scalability and performance for large
GridVision Enhanced incorporates a highly scalable subnet manager, which is a required component for InfiniBand fabrics that manages operational tasks such as routing, addressing and fabric initialization. In addition to subnet management, the software enables better fabric utilization and performance through advanced routing, central provisioning and configuration, and hierarchical root cause and analysis tools. For larger clusters and grids, as the environment scales, reliability, scalability and performance become increasingly important. Using server platforms to host the software, GridVision Enhanced is designed to provide faster configuration and response time for very large fabrics, simplified integration with standard Linux software tools and new routing algorithms.
GridVision Enhanced encompasses all of the performance and scalability features required by the top clustered supercomputers. The new host-based software encompasses the plug and play features of Voltaire's switch-based fabric management software which is used in some of the world's largest InfiniBand fabrics including supercomputers at Japan's Tokyo Institute of Technology and the U.S. Department of Defense. By running on a host instead of on the switch, GridVision Enhanced aims to further accelerate the speed of all fabric management related tasks including fabric initialization or "bring up", troubleshooting and accessing critical operational information. GridVision Enhanced can scale from a few to thousands of nodes in both stand alone and highly available redundant configurations and can bring up a multi-thousand node cluster in less than one minute, according to the company.
As a standalone software package, GridVision Enhanced manages heterogeneous fabrics that may include InfiniBand switches from multiple vendors. It also can access the advanced device management capabilities of Voltaire's InfiniBand-based hardware solutions including all managed switches and Ethernet and Fibre Channel router modules for fast I/O and virtualization.
GridVision Enhanced is available as a standalone software package and is also included as part of Voltaire's recently-introduced GridVision Enterprise service oriented infrastructure (SOI) management software. GridVision Enterprise automatically maps the relationships between application requirements and the physical or virtual data center environment consisting of servers, storage and networks to create a logical and efficient data center. Employing an abstracted object model, an intuitive graphical user interface and an open web services API, GridVision Enterprise integrates with leading server and storage virtualization solutions, as well as provisioning and scheduling tools, creating a single, logical data center environment consisting of both physical and virtual resources. The software enables automation of repetitive IT tasks associated with network, server and storage provisioning to deliver efficiency. GridVision Enhanced will be available in early 2007.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?