It's been nearly two years since Lenslet, an Israeli firm, announced the public availability of it's Enlight256 (pdf) optical digital signal processor (DSP). This revolutionary DSP utilized a complicated array of Galium Arsenide lasers, lenses, mirrors and spatial modulators to perform computations with photons rather than electrons. This breakthrough design allowed the Enlight256 to support dynamic instructions unlike previous designs that are hardwired to perform a single function only. Optical DSPs offer performance boons because entire operations can be performed in a single clock cycle. Control Engineering describe the DSPs ability to perform operations by altering light patterns in multiple locations in a single pass which eliminate instruction queue bottlenecks in silicon based microprocessors. Such a design lends the Enlight256 architecture to remarkable speed in processing vectors, matrices and transforms while the more mundane comparison and addition operations remains on par with contemporary electronics. Those speed benefits have allowed computational speeds of up to 8 TeraOperations/sec to be performed ... in October of 2003.

This technology has not yet emerged into consumer electronics, despite it's public availability and accessible architecture. Has the development of such radical technologies been replaced by industry's bottom line for profit? Are the benefits of nanotechnology and quantum processing so much more exciting than optical systems? Or is the inability of the R&D to build the Enlight256 as an x86 replacement the reason it's remained relatively unknown for the past two years?