Retrieving homologous sequences from existing databases, for instance, is a fundamental routine for the bioinformatics research community. Programs of the NCBI BLAST family have been widely used for retrieving homologous sequences from existing databases. This article briefly introduces and evaluates a parallelized version of the BLAST algorithm using Message Passing Interface (MPI) on a multi-node compute cluster, paraBLAST and its web server, CyberparaBLAST. The CyberparaBLAST web server can be accessed through web browsers on intranet web servers. CyberparaBLAST enables the researchers of institutes to query sequence databases in the real time when needed. Parallelized computing together with cyber access make the bioinformatics research more efficient with the aid of CyberparaBLAST.