We present a scalable framework for real-time data-intensive systems on com
modity multiprocessor workstations. The framework is an extension of the pr
ocess network model, which captures parallelism, guarantees determinate exe
cution, and executes in bounded memory. We implement the framework using li
ghtweight POSIX threads and prototype a 4-GFLOP sonar beamformer on a 12-pr
ocessor 336-MHz Sun Enterprise server. The beamformer scales nearly linearl
y from 1 to 12 processors.