This paper describes a new data acquisition system at JET which uses T
exas TMS320C40 parallel digital signal processors and the HELIOS paral
lel operating system to reduce the large amounts of experimental data
produced by fast diagnostics. This unified system features a two level
trigger system which performs real-time activity detection together w
ith asynchronous event classification and selection. This provides aut
omated data reduction during an experiment The system's application to
future fusion machines which have almost continuous operation is disc
ussed.