|HOME | INFORMATION | SPONSORS | PROGRAMME | EXHIBITORS SUBMISSIONS | CONTACTS | REGISTRATION | COMMITTEES | PROCEEDINGS | ARCHIVE | PRESS RELEASES|
Birds of a Feather Session: Visualisation
Date: Tuesday 19th September 2006, 18:05 - 19:00 Venue: Conference Room 2
Organiser: Ian Grimstead, University of Cardiff
The National Grid Service (NGS) currently provides computational resources for researchers, enabling large compute jobs to be handled which are far beyond the local resources available to most researchers.
However, it is noted that large compute jobs can produce more information than local resources can easily display, either through lack of graphic processor power or network bandwidth constraints. We are wishing to investigate the possibility of providing a National Visualization Service, enabling researchers to observe results from their desk in the same manner as submitting compute jobs. Such a service needs to cope with a wide range of client devices, source datasets and presentation formats.
In tandem with this, computational steering can be utilized to modify running simulations to prevent wasted compute cycles when a simulation fails to converge, or misses a target attribute. With computational steering, a simulation can be retargeted, verified or interrupted whilst it is in progress; this is in contrast to batch job submission, where a job may run unattended until completion, when the user may discover a large amount of compute has been wasted.
The aims of this BoF include the discovery of how people currently use visualization and/or computational steering in their work, taking into account best practice and particular requirements (e.g. frame rate, resolution/image fidelity, presentation techniques, etc.) and whether additional researchers would make use of viz./comp.steer if they could access it from their desks without requiring additional (potentially expensive) resources. Other aspects to be determined include ease of access and any barriers to take-up/accessibility of viz./comp.steer.
The output of this BoF will be a list of user requirements to be taken forward and used in an implementation of a National Visualization Service (NVS), supporting computational steering. We aim to pull in as much information from the potential user base as possible, with the ideal result of eventually producing an NVS suitable for a wide spectrum of users.
In summary - the output of this workshop is the sum of all its inputs, so if you have a view on visualization, you use visualization, are possibly intersted in a National Visualization Service, or would like to hear suggestions on what on earth visualization could do for you, please come along!