IBM Pervasive Computing Visit June 9, 1997

1
IBM Pervasive Computing VisitJune 9, 1997

• David E. Culler
• Computer Science Division
• U.C. Berkeley

2
Looking Forward ...

• What kinds of problems should Computer Science
  research be focusing on?

3
Future Internet-Scale Systems

• 10 Billion of Information Appliances
• 100 Million of Stationary Computers
• Million Scalable Servers

4
Millennium Testbed (started 9/97)
Business
SIMS
BMRC
Chemistry
C.S.
E.E.
Biology
Gigabit Ethernet
Astro
NERSC
M.E.
Physics
N.E.
Math
IEOR
Transport
Economy
C. E.
MSME
5
Millennium Basic Components

• Hierarchical Cluster of Clusters
• 6 M Intel Equipment (1,000 PCs)
• participant desktops gt dept SMPs gt group
  clusters gt campus cluster
• TBs of IBM storage
• Gigabit Ethernet (Bay Nets. 70 dis.) between CAN
  within
• NT and Solaris/x86
• Staff through Campus and NSF
• Network and High-end Visualization through NSF
• Common Technical SW env. For computational
  Science and Engineering
• NOW cluster system SW gt distributed
  computational economy

6
Expanding the Vision

• Small devices will be pervasive.
• They can be powerful when they draw upon the
  infrastructure.
• How do they change they way we think and work and
  learn?
• cuts through many research areas

7
Partnership gt Testbed for I.S. Research
Wireless Infrastructure
Future Devices
PDAs
Cell Phones
8
Plan for Today

• 100 - 115 Greetings
• 115 - 145 Mission for IBM Pervasive Comp.
• 145 - 150 UCB Intro
• 150 - 215 Big Infrastructure Small Clients
• Demos Wingman and MediaPad
• 215 - 235 Informal Collaborative Note Taking
• Demo notepals
• 235 - 250 Experimental Plan
• 250 - 315 Strategic Issue Discussion
• 330 - ??? Tactical Planning
• IBM tech. Road map, Wireless options, gt TODO
  list