Brief Summary of trip to FNAL for the US_CMS Software meeting

FNAL Computing Division

FNAL have committed to providing manpower and computing infrastructure support to CMS. J.Bunn, R.Mount and H.Newman spent a day discussing with people in the FNAL Computing Division what this commitment might mean quantitatively.

The FNAL computer centre sports around 3000 CERN Units of processor capacity, 1000 tape drives, more than 2.5 TBytes of disk, and a few tape robots (3490 and STK). In the near term, the possibility of allocating CMS simulation time on the Sloan Sky Survery machine was suggested.

FNALs highest priority presently is support of CDF and D0. However, Joel Butler (in charge of the Division) had appointed Stephan Lammel and Michael Diesberg as coordinators between the Division and CMS.

Software meeting

This is my personal list of the highlights of the meeting:

Worries and frustration about WAN bottlenecks between the US and Europe, and within the US itself.

Some concern about divergence between FNAL and CERN on various software topics, for examples CAP/RD45 and FUI/HEPIX.

Some apparent overlap of effort between FNAL and CERN in the area of investigations of mass storage hardware and software.

Very interesting and thorough work presented on background and trigger simulations by Breedon and Dasu respectively.

A highly topical presentation on the BaBar C++ Reconstruction framework by X.Shi ... what particularly appealed to me was the flexibility with which the user could configure which modules the reconstruction program used. Although the framework currently uses the CLHEP class library, it was intended to migrate to the STL.