I find it interesting that the contributors work for semiconductor companies. I would've thought that something like this would be more likely to emerge from a team with a mathematics/CS/academic background more so than hardware. Perhaps these algorithms can (and will) be going strait into some new hardware cryptography IP cores?
I only tangentially follow the semiconductor companies...
However, I know several advanced algorithms in the realm of cryptography can be computationally expensive, and as demand for security grows, several of these semiconductor companies have implemented things in the hardware to optimize certain forms of encryption. (I can't remember if it was Intel or AMD that released a chip with some sort of built-in support for security).
This is just a theory, but if they're going down this road, it would make sense for them to develop a highly secure algorithm that they know can be easily implemented and optimized on the hardware level. If they're developing specialized algorithms AND hardware, they can probably take it into account during the design process and further optimize their boards.
Or, a bunch of semiconductor guys are also crypto-geeks. Which is pretty likely anyway :D.
All of the SHA-3 candidates were designed to be implementable in fast hardware. Here's a comparison of some of the hardware designs evaluated during the competition:
