AMD announced its “Black Edition” Phenom yesterday. It is a 2.3GHz Phenom with the multiplier removed.
It’s easy to see why AMD would want to put out this product. It lets them sell more of these troubled B2 stepping chips, and lets them sell these chips for more than they would otherwise.
What is not so easy to see is why most people would want to buy one, at least not at the initial prices.
They aren’t going to overclock as much as Penryns or even C2Ds, so any money you save will be paid for with lower performance in any new system.
Yes, you’re supposed to be able to drop one of these inro a current AM2 board, which certainly increase the cost benefit, but if you have an AM2 board, you most probably have a relatively new X2 processor in it.
Even if an overclocked Phenom runs as quickly as the overclocked X2, you’re going to end up paying somewhere around $250 for about a 15% performance improvement. That’s pretty pricey from a price/performance standpoint.
Given that these Phenoms probably won’t run quite as quickly as current overclocked X2s, the performance increase will probably be more like 10%, which is definitely not much bang for the buck.
To be fair, the exact same argument can be made for a Penryn upgrade from a Q6600, if anything, the argument might be a little stronger.
In all honesty, if I had to choose between Phenom and Penryn, I’d pick Nehalem. Given its integrated memory controller and a new design which takes full advantage of Intel’s 45nm process technology, it ought to provide more of an improvement step-up than the next offerings. Having a more mature/less expensive Vista64/DDR3 environment won’t hurt, either.
Of course, “twiddle your thumbs for a year” may not be the most popular advice to give, and if these Phenoms were cheap the way K6-2s and Thunderbirds or Thoroughbreds were cheap, the advice would be, “Knock yourself out.”
But $250 or even $200 is a lot to pay for what will end up being a disposable CPU.