The weakness here is the very strength AMD is hoping for. This chip is targetted at what might be, not what is. AMD is looking forward and banking on the way software is heading, that is great but at the present this does not translate well. What makes this worse is that while AMD is likely aimed at the right spot, it is a spot still a few years down the road for MOST applications.
For me the FX4100 is the real story of this release. The 4100 is priced right at the same point as the 955. The higher clock speed means it is actually, likely, faster than the 955 and so becomes an awesome budget choice. I am hoping we can get a 4100 to put this theory to the test. However I am concerned that with the lackluster numbers when using most of todays apps the 4100 might end up being a wash.
That is the way I feel the 8150 ended up, being a wash. If you have an existing Phenom II system with a 965 or higher quad core or a 1090T or so in the six core then there is really not a lot that suggests this is a good upgrade, it is more of s step sideways. The FX can beat both chips but only if the app can fill it's thread pipeline. The same goes for the 2500K the target of the 8150. If you do a lot of heavily threaded apps the 8150 is a solid choice but if you are a typcial user of your PC in any way then the 2500k is just a better value.
AMD is investing it's future in this architecture and that future needs developer support to really be leveraged. If things go the way AMD is aiming this could end up being a smart move, it could also blow up in their face. This is not a bad chip and in the haze of the future it might even be a great chip, in the reality of today it is an okay chip but not anything to get excited about.