The heavy-ion collision generator HIJING was used,
in various modes:  "central" refers to an impact
parameters distribution cut off at 2.0 fermi;
"min-bias" refers to a distribution out to 14.0 fm.


A variety of light-light and light-heavy ion collision
classes were simulated, with a particular emphasis on
data sets of use in preparation for a deuterium-gold
collision program at RHIC.  Here is a summary of the
events generated:
                                                         
------------------------------------------------------------------------------
 Au+Au Central:                 ~40,000

  D+Au Minbias:               1,201,372   (*1/52=  23,103 central Au+Au equiv)

  D+Au Central:               2,057,527   (*1/25=  82,301    "      "     "  )
 Au+D  Central (reversed):    1,407,827   (*1/33=  42,661    "      "     "  )


 Si+Si Minbias:                 340,557   (*1/26=  13,098    "      "     "  )

 Si+Si Central:                 187,331   (*1/9=   20,814    "      "     "  )


 Cu+Cu Minbias:                 198,755   (*1/13=  15,289    "      "     "  )


         TOTAL:               5,433,369   (237,266 central Au+Au equiv)
------------------------------------------------------------------------------

The simulations were orchestrated on the Los Lobos
machine via a custom-designed script system (three
tiers of "online" control and several pre- and post-
processing "offline" steps) that made use of the pbs
batch system and various job control components of MPI.

The produced MC data were transferred to Brookhaven Lab
(BNL) using the 'bbftp' program and stored in BNL's
HPSS for further processing.  Modest (by Grid standards)
transfer rates of up to 5 MB/sec were seen over UNM's
current Internet-II connection to the ESnet.



Possible future work at the AHPCC includes:

 o Local implementation of several post-processing
    software stages.  These require connection to
    PHENIX's central (at BNL) Objectivity data base.
    An alternative scheme is to connect to a mirror
    of that data base, maintained by colleagues at
    Vanderbilt University.

 o Implementation of the PISA (GEANT 3) step via
    the Condor wide-area shared-desktop management
    system [ http://www.cs.wisc.edu/condor ], which
    is becoming an integral component of the Globus
    grid toolkit.  I have successfully deployed other
    particle physics applications, including some
    using GEANT 3, in this way.  (These experiments
    were facilitated via the AHPCC's research
    relationship with the University of Wisconsin,
    where Condor was developed.)  I also would like
    to make use of Condor Kangaroo and GridFTP for
    data gathering at AHPCC and transmission to BNL.

 o Exploration of the use of various data-grid software
    technologies to better manage such complicated issues
    as file cataloging and the multi-step data processing
    flows that are characteristic of PHENIX's large-scale
    heavy-ion simulation projects.



The Albuquerque High-Performance Computing Center is
located on the campus of the University of New Mexico.
I wish to express my gratitude to the AHPCC for the
time grants and other assistance which they have
provided to make this work possible.