Limitations
Parimics' technology and IP is based on two very basic axioms. One is that for a certain class of computational problems, dedicated special-purpose processors are more energy- and instruction-efficient, and as a result, they are easier to program than general-purpose processors for these classes of problems. Our second axiom is that simplicity is key, and not complexity, for these categories of computational problems.
We do not claim or assume that all problems can or even should be solved by dedicated special-purpose processors. Current general-purpose processors offer a tremendous performance at a fairly low cost - for general-purpose applications. They do not solve all problems equally well, or even at a reasonably low level of programming effort.
There are a variety of reasons why image analysis applications are not well-served by existing general-purpose processors, DSPs or GPUs and software. This is especially surprising as the algorithms have been known for a long time.
We have made sure our customers understand the benefits of Parimics' architecture. If you want to know what distinguishes Parimics' solution from general-purpose processors, DSPs and GPGPU computing, then we'd be happy to send you a whole series of white papers that explain what we do differently, and why. Just send us an email to info@parimics.com requesting the white papers, and we'll send you a set that does not require an NDA.
