What next?
Exploring what other competitive advantages King-Smith thinks Imagination has, he mentioned ‘tile based deferred rendering'. Apologising for our ignorance, we asked him to explain, ideally in words of one syllable, what that is.
"Tile based deferred rendering divides into two parts," said King-Smith. "The ‘tile based' says we divide the screen into a set of tiles, we do a lot of processing of each tile on-chip, which means we need to use much less memory bandwidth in order to process much more efficiently.
"The ‘deferred rendering' says that when we start processing the graphics, the first thing we do is get rid of anything possible that isn't going to end up resulting in a pixel. So we never waste any processing power."
On the back of that description, tile based deferred rendering seems like a good idea. So we asked King-Smith why everyone doesn't use it. "It's hard," he said.
To conclude we asked King-Smith what we can expect to see from Imagination in the future. The performance you're seeing from us today is at the low end of what we're capable of," he said.
"The intersection point between what we do and what the likes of NVIDIA and AMD do right now is in that netbook space. But we see ourselves as a sort of wave-front moving out across across these markets and you will definitely see us in new market areas, which those guys probably think should belong to them. I'd say we're one of the most important companies you've never heard of."
The only way to put Imagination's claims to the test is to benchmark a product we know uses its IP, like the Nokia Booklet, and see how it performs in the areas of graphics and video processing. What seems clear, however, is that we're likely to hear a lot more from this company as the SoC market goes exponential.
A lot of this is likely to be manifested in future Intel products and, with Imagination Technologies as its secret weapon, Intel seems to have significantly improved its chances of success as an SoC player.
