To make informed decisions about what computers to purchase and/or use for their applications, computational scientist need access to benchmark performance data. The process of gathering, archiving, and distributing computer benchmark data has typically been carried out by various users and vendors with little coordination. The NSE prototype provides access to a publicly-available centralized database of performance data for all ranges of machines from supercomputers to personal computers [2]. The Linpack benchmark, Perfect benchmark, and Genesis benchmarks are included. A number of benchmark reports and a benchmarking bibliography are also available.