I am surprised by only one thing, its AMD's unbelievable will to win! Intel, for fear being labelled anti-competitive and for higher margins, has always allowed AMD to rule the budget gaming segment. If anyone wants a quad-core for less than 6000, even now the Phenom II X4 955 is the way to go. With a decent graphics card like the Radeon 6850/70 there are very few games that do not run smoothly on the Phenom II but run smoothly on Core i5 2400. And the situation is likely to remain the same at least for a couple of years. AMD could have easily maintained this status quo by rolling out a tweaked Phenom II which is around 20% slower than a i5 2400 and priced competitively at around 150$ and stay live for the next couple of years. But, instead they come out something radically different from Phenom and attempt to upstage Intel again. In my estimation they have failed horribly, but the fact is they tried something very different and took a huge risk and it backfired.
Bulldozer, as it stands, is not a good buy primarily because of the very bad single threaded performance, and even worse power consumption. AMD is lying when they say the TDP is 125w for FX-8150. It seems to be close to 150, but of course no actual way to measure this. If pile driver can improve single-threaded performance by 20% and the same time bring down power consumption by 20% (I am asking a lot), AMD will get to a situation where they can make some money. Pile driver will probably decide the future of the CPU industry as it will launch around the time Windows 8 launches. With the release of Windows 8, low power and ultra cheap ARM desktops might start eating away AMD's market share in the low-end spectrum, and Intel with Ivy Bridge might be able to produce a Quad core at 65w TDP will dominate the high end. AMD must find a way to squeeze in between these two markets with pile driver.
