Preview: IBM Deep Computing Visualization to add ability to run on Microsoft Windows, and a new Scalable Parallel Visual Networking interfaceIBM United States Software Announcement 206-285
November 7, 2006
|Table of contents||Document options|
|At a glance|
Deep Computing Visualization plans to expand their solutions with future new software offerings, which will:
- Include the ability to run on the Microsoft Windows platform
- Enable large sets of complex, multidimensional application data to be rendered into a meaningful form that can be shared and utilized across worldwide networks
- Help enhance the 3D graphics interface of software applications
- Provide a low-cost, open standards-based alternative to proprietary visualization systems
- Support with OpenGL up through version 2.0, including support for shader programming capability
- Provide clean visual integration and seamless elimination of artifacts
- Provide improved responsiveness and visual performance
Scalable Parallel Visual Networking (SPVN) will be a new component added to the Deep Computing
Visualization software family in 2007. It is an application programmer's library for implementing
real-time, high-performance graphics on distributed memory systems.
Back to top
IBM Deep Computing Visualization will soon offer a Microsoft Windows-based solution which will be based on industry-standard workstations, commodity networking, and cluster rendering software that enables both local and remote visual collaboration. Similar to the current Deep Computing Visualization for Linux product, this Windows-based solution will provide a scalable, collaborative middleware infrastructure to help enrich the graphics function of OpenGL software applications.
A Deep Computing Visualization implementation running on Windows can be attractively affordable and highly adaptable. With Deep Computing Visualization running on Linux, high-end graphics can be rendered in either Scalable Visual Networking (SVN) or Remote Visual Networking (RVN) mode. On Windows, high-end graphics can be rendered in RVN mode.
These visualization modes can help provide improved performance and function at a fraction of the cost of many proprietary offerings. SVN will support multiple high-resolution monitors or projectors for immersive visualization. RVN uses rich security to distribute graphical images to remote (collaborative) clients called endstations.
The next new release of Deep Computing Visualization plans to include the following new capabilities and features to build upon those which already exist in Deep Computing Visualization for Linux V1.2:
- Support with OpenGL up through version 2.0, including support for shader programming
- Accommodation for latency-sensitive networks
- Improved documentation, installation, debugging, and error recovery/reporting
- Support for the new IBM IntelliStation® Z30 workstations, as well as the new NVIDIA FX 5500
- Clean visual integration and seamless elimination of artifacts
- Improved responsiveness and visual performance due to implementation of a handshake protocol to meter frame flow rate
- Enhanced user interface
- Better responsiveness for high-latency (long-distance) application
Deep Computing Visualization will be offering a new component, SPVN, to its software family in 2007.
Initially for the Linux platform only, the SPVN interface is designed to enable an application to
partition data, redistribute data, load balance, synchronize, composite, and control I/O
functionality. The advantages of the SPVN distributed system approach will be that it is scalable,
modular, flexible, capable of upgrade and price/performance competitive.
Back to top
- Microsoft Windows XP and NVIDIA 3D graphics cards, or
- Red Hat Enterprise Linux and OpenGL on IBM workstations or selected eServer® servers running Linux
Previews provide insight into IBM's plans and direction. All statements regarding IBM's plans,
directions, and intent are subject to change or withdrawal without notice, and represent goals and
objectives only. Availability, prices, ordering information, and terms and conditions will be
provided when the product is announced.
Back to top
Effective visual solutions are becoming business-critical requirements as the volume and complexity of data increases. In what has become a virtualized world, data is often collected at different locations, where it is analyzed by a team collaborating at various locations. Deep Computing Visualization supports applications that take large sets of complex, multidimensional, nonlinear data, and helps render that data into something meaningful that can be shared and utilized worldwide.
Deep Computing Visualization helps enhance the 3D graphics functionality of software applications. These applications represent and display 3D images to provide a better understanding of the real world and to derive insight from complex data.
A Deep Computing Visualization solution is based on clustered graphics workstations coupled with differentiating IBM visual systems software that enables both local and remote collaboration. It includes:
- Remote Visual Networking: A technology for enabling remote delivery of unmodified OpenGL applications to one or more client endstations over commodity networks
- Scalable Visual Networking (Linux only): A technology for the network transport of streams of OpenGL graphics from a single host, executing an unmodified 3D graphics application for the purpose of displaying in pixel-rich, multi-pipe environments
Deep Computing Visualization can provide a low-cost open standards-based alternative to the proprietary, monolithic visualization systems that have been a standard in the market.
Applications can coexist for both batch computation and interactive visualization, using industry-standard job scheduling and cluster management software.
Industries such as chemicals and petroleum, life sciences and health care, government, higher education, PLM, and digital media can utilize Deep Computing Visualization to help gain intelligent differentiation from their data. As a result they can potentially optimize revenue, minimize time-to-market, minimize product costs, and maximize effectiveness.
Deep Computing Visualization helps increase screen resolution and will allow remote use of the application, all while maintaining performance and scalability. Deep Computing Visualization will also enable applications to be displayed on large multi-projector display walls and high-resolution monitors, with minimal cost and effort. This feature can enable more accurate decisions to be made based on the higher resolution of data contents.
Deep Computing Visualization makes graphics applications much easier to manage, by keeping the application and its data centralized, and by using rich security to transfer graphic images to remote collaborators anywhere on the network.
Deep Computing Visualization can help enable clients in industries such as petroleum exploration production, automotive, aerospace, life sciences, and digital content creation to enhance decision making and derive new insights from complex sources of data.
Deep Computing Visualization uses commodity, off-the-shelf components, is scalable for growth, and is modular for enterprise flexibility all key components to help provide IT investment protection and cost savings.
Deep Computing Visualization will support RVN when running on Microsoft Windows XP SP2® on applicable graphic workstations. The new release of Deep Computing Visualization will continue to support both RVN and SVN on Red Hat Enterprise Linux and OpenGL on IBM workstations or selected eServer servers running Linux.
Scalable Parallel Visual Networking (SPVN) will be a new component of the Deep Computing Visualization software family. It is an application programmer's library for implementing real-time, high-performance graphics on distributed memory systems. It is targeted to developers of software applications to engineer, or re-engineer, domain-specific programs for use on stand-alone systems or graphics rendering clusters that employ commodity graphics accelerators in workstation or server class systems. The library can help address performance or problem size issues that might not otherwise be resolvable. The SPVN interface is designed to enable an application to partition data, redistribute data, load balance, synchronize, composite, and control I/O functionality. The advantages of the SPVN distributed system approach are that it is designed to be scalable, modular, flexible, capable of upgrade, and price/performance competitive.
SPVN is fully object oriented and embodies two core, yet architecturally separate, component objects: a distributed scene graph and a distributed rendering engine. A novel design feature of the scene graph (the hierarchical data structure for encoding graphics objects to be modeled) is that it can be fully and readily extended by programmers to support node objects, transformations, and attributes specific to an application domain. The extensible render engine class of SPVN will initially support both sort-first and sort-last rendering and both polygonal and volumetric data, but users can derive their own custom renderers that optionally can be independent of OpenGL.
Here are some of the other outstanding features of Deep Computing Visualization with SPVN:
- Ability to be used in conjunction with high resolution or multi-tile displays
- Support for surface and volumetric data
- Support for static and dynamic data
- Support for out-of-core techniques for data problems that exceed aggregate system memory
- Use of a distributed, shared memory model for primary data storage
- Flexibility to work with many GUI toolkits
- Support for remote display
Graphics optimizations, such as:
- Visibility culling
- Adaptive load balancing, caching and pre-fetching
Planned availability dates
- December 2006: IBM Deep Computing Visualization RVN option on the Windows operating system
- First half 2007: IBM Deep Computing Visualization SPVN option on the Linux operating system
- First half 2007: IBM Deep Computing Visualization SVN option on the Windows operating system
Back to top
The primary benefit of Deep Computing Visualization's RVN capability is that geographically dispersed experts have access to the computational and visualization capabilities of a scalable graphics computer, meaning:
- RVN could help improve communication between team members and shorten cycle times in decision-making processes.
- When multiple scientists or engineers are accessing a centralized database, RVN also could help improve data integrity and security, because multiple copies of an organization's core data is not being replicated in multiple geographies.
- RVN could help enterprises to remove expensive workstations from every engineer's desk, and consolidate them into a managed research that can be shared among a larger number of workers.
For companies whose business relies heavily on organizational knowledge combined with massive and growing amounts of data, the IBM Deep Computing Visualization solution running on Windows could help bridge the gap between having better information and making better decisions.
IBM has the breadth of immersive, collaborative visualization technologies that will link your experts to your data, and the depth of expertise to help streamline the infrastructure of your organization, helping you to make intelligent decisions faster and better.
Business Partner information
If you are a Direct Reseller - System Reseller acquiring products from IBM, you may link directly to Business Partner information for this announcement. A PartnerWorld ID and password are required (use IBM ID).
Back to top
Most product media is shipped only via Customized Offerings (that is, CBPDO, ServerPac, SystemPac®). Non-customized items (CDs, diskettes, source media, media kits) will continue to be shipped via the stand-alone product.
Back to top