Originally Posted by skupples
Actually, right now mantle is only slated for FROSTBITE engine products.(EA onry)
Several other studios have signed up to support Mantle on their engines, there is a thread about it here in the news section.
Originally Posted by fateswarm
I wouldn't be so sure. I looked at the science behind it and I'm convinced there is a concrete limit on conventional technology, e.g. at a point you simply can't make the transistor contraption any more because the building blocks are too small, without even going to quantum mechanical limits creeping in before even reaching that, which they already start showing up, e.g. notice how it's harder to overclock a processor nowadays (it's not just because "intel goes for laptops", that's a fallacy) and they are generally slower per new iteration (even if they seem to have tried to slow it down even more with smaller dies to save some (market) time).
Then if you go to completely different paradigms - e.g. Graphene, even Intel's own optics, or Graphene-Hybrids - you notice that it is very unclear if they will have a competitor to conventional silicon in years or even decades and even if they are competitors in time, concerning the subject at hand, the stock of these companies, it is not very likely (and I would say it's unlikely) that they would be able to adopt them in their technology before someone else takes their position in the market.
The problem isn't the building blocks being too small, the problem is the transistors themselves getting too close to each other. The current smallest node is 22nm, which means the transistors are 22nm apart from one another. Between these transistors is a gate that is controlled by an electrical signal, opening and closing the gate allows current to flow between the transistors.,
The problem we run into is as these transistors get closer together, smaller manufacturing node, we lose the ability to control the electron jump from one transistor to another. Eventually transistors are going to get so close, that they will just jump on their own, we won't be able to control it. This is expected to happen between 5nm and 7nm; which really isn't that far out.
So what options do we have? New materials that will let us get down to that level and still control the transistors, a new type of transistor completely, or actually some say stacking transistors. In stacking the transistors we would take silicon to the smallest controllable level, say 8nm, and then begin stacking transistors to put more on a die.