Documented results of 1 to 256 node runs on Chiba City
1/01
+ runs with fast ethernet/UDP, Myrinet/MPI, with & w/o
algorithms
Reconfiguration of models to conform to Phase 1 (S)
3/01
+ should include test measurements with
component timing and processor utilization -
this places
a demand on phase 1 (time stamps in and out
for each
component, network inferred from these results)
Preliminary UML diagrams describing HLT/ROS/Data Coll.
5/01
+ Interaction diagrams and others required to characterize
system timing and component interactions
Review of paper model and update to current parameters
6/01
+ include updating of trigger tables and expected performance
Generic model update 6/01
Evaluation of model frameworks
7/01
+ review of suitability to task
o includes completeness, accuracy, speed,
flexibility, people
(which will get the job done
- if both then which is best)
Selection of primary and secondary framework
7/01
+ primary framework should have complete set of components
& results
UML interaction diagrams for all system components available 7/01
Comparison to phase 2A results 8/01
Tuning of parameters to 2A results as needed complete 9/01
Full set of component models 10/01
Documentation of component models (includes 2A comparison) 11/01
Comparison to phase 2B results (S) 1/02
Documented phase 2B comparison (S) 2/02
Complete system model 5/02
Review of paper model and update to current parameters
6/02
+ include updating of trigger tables and expected performance
Review of phase 2B->2C changes and model updates complete (S) 7/02
Documentation of system model 8/02
Comparison to phase 2C results 8/02
Complete system characterization 8/02
Documentation of system model results (includes phase 2C)
9/02
KEY: (S) means the milestone is not ABSOLUTELY necessary and
can slip or disappear
without causing critical damage
to the program