ABSTRACT The MPEG-4 Audio standard provides a toolset for synthetic Audio generation and Audio processing called Structured Audio (SA). SA permits to describe algorithms through its Structured Audio Orchestra Language (SAOL) programming language. Unlike some other languages of the same type, SAOL has a sample-by-sample execution structure, and this makes particularly important the overhead computation in the case of an interpreted decoding. This paper describes the design of a virtual DSP architecture able to exploit the data level parallelism contained in many audio synthesis and processing algorithms and to consistently reduce the implementation overhead.