The problem is that Fusion is pretty much a busted flush in a desktop scenario since Bulldozer looks like it isn't going to be causing SB too many problems. Fusion has done well in the notebook sector by dominating Atom based netbooks and being cheaper than the cheapest i3 notebooks. The problem is that they have no coherent tablet/phone strategy.
Nvidia for all of their issues in the GPU sector have realised that the traditional GPU is dead and have already started to make the transition to HPC with their GPUs. In the mobile sector they have got the Tegra platform which is extremely competitive in Android tablets and phones right now, and Tegra 3 has begun to roll out with rumours that we could see the first quad core tablet in Q1 2012 from Asus. The brand power that Nvidia have given Tegra is beyond any other mobile SoC in those terms and just like people look for the Intel Inside sticker on notebooks/desktops they will begin to ask whether a tablet has Tegra or not.
Back to AMD - without a tablet/phone strategy they are dead in the water as Intel dominate the x86 sector and Nvidia have better brand power in the discrete GPU sector despite their relative weakness of late. The only bright spot has been mobility. They must expand into tablets/phones or face the bankruptcy of the company. This is the next frontier of computing and the notebook market will go in reverse as regular consumers who go on facebook, YouTube and Gmail while watching TV move to tablets. That move will cause their one bright spot in computing (Brazos) go into terminal decline with nothing to replace that revenue.
All in all, AMD have been concentrating so hard on catching Intel in x86 and Nvidia in discrete GPUs that they have missed the next big sector of computing. Nvidia have let their GPU dominance lapse, sure, but Tegra gives them a real business model going forwards, one which AMD doesn't have right now. Management have a choice between burying their heads in the sand or getting a real mobile strategy going. I personally think it is too late, and in my experience (investment analyst specialising in tech companies) a company that enters a mature sector so very late will find it difficult, especially if they have a handicap like x86 in an industry dominated by ARM.
The course for AMD is to acquire an ARM licence in short order and put out a Tegra competitor using their knowledge gained from Fusion, and they need to get a product using it out onto the market before the end of 2012, some kind of tablet, even if it means paying Acer or Asus to do so. If they make these moves their company will survive in the longer term as the tablet market grows and it will give them more time to address their x86 processor woes against Intel.
Sorry about the longish post, but this is the view about AMD from my industry of banking and finance and many people are very worried about their future and are looking to pull the plug if they can't come up with a business model that taps into the fast growing smartphone and tablet market.
What does this have to do with ATI?
