BRAHMS Activities (RCF meeting October 1 1997).

Presentation by F.Videbaek

Software meeting Copenhagen August 4-5

Oslo, Bergen, NBI, Krakow, Strasbourg +fv

Framework, data structures, data flow discussion

 

Collaboration meeting August 23-24, Brookhaven

Agreement to use ROOT for presentation purposes. Will also fits into other ongoing analysis activities.

Additional investigation with regard to STAF.

event format (XDF very in in-efficient for small data structures)

present tables and AMI are very restrictive for OO approach.

 

Will future development be sufficient clarified in short term so our actual writing of can continue.

Activities using RCF

mainly in detailed simulation of detector response for finalizing design using GBRAHMS (Cherenkov, Beam-Beam counters, H1/H2,) More collaborators are utilizing RCF rather than local machines.

DQS (AFS password problems) and MDS experience generally positive

 

 

Possible Benchmark contribution from Brahms

Use the standalone Sonata++ which already compiles on different platforms using g++ and/or native C++. Input can be ASCII hit-files.

Run in 3 different configuration (digitization and tracking code). These do reflect a good part of the BRAHMS analysis goals, but not all.

FS small angles- cpu demanding Drift Chamber C and TPC an fair amount of floating point calculations.

MRS high multiplicity cpu demanding, pattern recognition

FS large angles - low multiplicity mainly integer (logic)