Parimics' strongly believes that most problems can be solved well by traditional general-purpose processors. Image analysis - especially real-time image analysis - is not among them. So far, the road to machine vision is littered with failures, but most of them were predictable. Here are a few links to the reasons of the failures:

Even the design and manufacturing powerhouse Intel had to concede to the problem of power consumption:

As a result, many CPU vendors have turned to parallelism instead of Moore's Law for scaling performance without incurring the thermal problems. This helps for a variety of general-purpose processing problems, but specific applications still cannot be properly addressed.

While multi-core processors prove to be the right solution for a wide variety of computational challenges, they are not the solution to all ailments. Among the problems are load sharing, load balancing, and even very fundamental issues such as programmability.

The generic Multi Processor or multi-core Programming Model and Real Time Operating Systems (RTOSes) for these are among the challenges that a general-purpose multi-processor or multi-core system encounters. See these links for more information:

2006 went down as the year the electronics industry woke up to the full breadth and significance of the trend to multicore processors. See the article here:

Neither hardware nor software of Parimics' processors are subject to these challenges and restrictions.