proactive mind?
You make it sounds like AMD want to kill off a market with a single chip. Unfortunately, that is not proactive mind. Who with the right mind will try to combine a market by using one chip? If you think AMD's fusion is going to kill anything, it is going to be itself.
Fusion is a new type of CPU, at best it will stay in the market, but can't kill a market. Let say there are 3 markets only. A rich, B mid, and C low. If a single chip can be used in A, B, and C, then this chip must be cheap and powerful at the same time. What does it mean? Market B and C will buy the chip and won't buy another one for years to come. Profit? To customer yes, big time. For the manufacturer? Short term yes, but will quickly dies out.
The definition of proactive is to create new markets. Nvidia opened the professional graphics as well as GPGPU, the high end computing with GPU at the supercomputing segment with tesla at the top end, tegra at the mobile segment, and ION to the mini PC segment. Those are all new markets for a GPU company. Harvesting those will make money until opponent steps in, which will saturate overtime.
Back to Fusion. It must create its own market, or it won't go far. It will eventually create a segment on the mini PC/notebook market if it is done right, which is what AMD needs to survive.
People with reactive mind will believe AMD will eventually kill off everything else that ever created because of the success of cypress. However, cypress' glory has long pass due and if they don't come up with something great again. Otherwise, they will lose whatever they have gained with cypress. Nvidia believed they are save with the 2xx series and believed that AMD is far away until they were hit by the cypress. In six months time Nvidia suffers huge hit in the discrete video card market until they finally come back with a decent card GTX460 to stop bleeding uncontrollably. But that is far far away from a knock out. If AMD don't have some surprises under its sleeves, the market will bounce right back to before cypress.
Seriously, usually people don't kill people. We simply try to survive. The same thing happens to companies. AMD won't want Nvidia to die or else Intel will kill it shortly after. Nvidia won't kill AMD either as Intel will kill it too. What they do want is to kill Intel, which they really can't. As soon as AMD successfully sued Intel, others immediately follow, including Nvidia. They can be best friend forever if the discrete market is big enough to feed all the employees. Unfortunately, it isn't enough. In fact, AMD is good at CPU, but since its opponent is too strong, they brought ATI so they can cover more ground. Unfortunately, Nvidia is very strong at GPU and Intel is very strong at CPU, leaving AMD very small room to breath. Fusion was the first project right after AMD and ATI become one, but for 15 years I still don't see AMD make well use of its advantage of having 2 blades. They are not dump, it is just too funky hard to do it.
Intel is the only one who have the muscle to kill others, but laws prevents them to do so. By killing i don't mean making better goods, but to use connections and relations to stop its target from functioning. Otherwise, their opponent will not die just because of a good product. "Duh, they simply cost more to buy!"
