In summary, this data example clearly demonstrates the potential for enhanced reservoir characterization by high-resolution surface seismic velocity analyses. The stacking velocity analysis took about 1 cpu second per CMP gather on 4k processors, and about 7 cpu seconds per semblance panel for the migration velocity analysis, not including I/O times. The stacking velocity analysis is heavily I/O bound, but could be made faster by a factor of about 2.5x with the use of a 10 Mb/s data vault or fast disk array. The migration velocity analysis is mostly cpu bound, and could only be marginally improved by about 20% with fast I/O.