Transcript CineGrid - Marshall Ikonography

1. Visualization and Collaboration and
Network Evolution
Increasing importance – a continuing and big impact
– For example: CarletonU’s Eucalyptus project for architectural
collaborative design that uses UCLP to reconfigure the network
based on resources required. 4x1Gbps. Tools include
• shared high resolution models,
• uncompressed SD or HD conferencing,
• HD document camera
– McGill’s Shared Spaces -- Multi-screen uncompressed HD real
sized VC: 5Gbps
– New apps need reconfigurable networks or at least lots of
available capacity to make this work
– Dedicated lightpaths
2. Role for CineGrid to Improve
Content Quality & Interactivity
> Biggest challenge is last mile/kilometer
– Many campuses and other sites don’t have capable
infrastructure
– CineGrid provide a key role if it helps stimulate this through
funding or even through peer pressure
> CineGrid should use its connection to wellestablished “science” research groups to identify
and implement key cross-over applications
– Must be clearly superior to existing services for acceptance
3. Artists and Technologists
Working Together
> We’ve seen such collaborations happen many times
(Moveable Feast (next gen Cinema?), SAT in
Montreal, Eucalyptus)
– Fun! Interesting! Exciting! Can show well – or be obscure.
– Results in new approaches but unless there’s a concerted effort
to get the results in use they often result only in demos and only
peripheral long term effects – critical mass issue?
– But some things outside the (traditional) research community
can have some significant effects – Azureus (HD YouTube) –
perhaps only when community gets large enough
– New models/languages for storytelling may emerge – a new
immersive cinema
– New highly distributed workflow models (modulo latency)
4. New R&E Net
Developments
> Discipline specific IP networks on optical substrate
such as CANARIE with UCLP essential
> Huge increase in the numbers of places that can be
reached with gigabit or multi-gigabit connections will
really push new visualization and cinema, dance,
performance, teaching and mentoring apps
> Not so much new technical developments but a
wider deployment of existing high-end infrastructure
such that it reaches a critical size
– How much is enough?
5. Network Researchers
Learn
> Special/defining characteristics of high-bandwidth
time critical streams – handle well (but hard too see
that this would be too significant)
> Justifies the existence of our networks
6. 5 Year plan @ CANARIE
> Coverage/Access: With renewed funding… will
extend direct lightpath connections and capabilities
directly to many new and existing facilities that
could be CineGrid users
> Security/Support: In our model, support and security
is a local or consortium responsibility
7. GLIF lessons for CineGrid
> Get real world applications happening ASAP
> Include the right players ASAP
> Get visibility first, then work on regularizing and
developing the key processes & services
– Use initial experiences to inform that development