Can Denebs hold their own against Penryns?
Fudzilla reports that Phenom IIs should be “as fast or a bit faster than 45nm Yorkfield based Core 2 Quad.”
I’ll take that to mean “as fast or at least close, except where Phenoms are really bad.”
Such an estimate does seem to be in accord with this AMD slide, provided you read it very carefully and ignore the apples-to-oranges comparisons.
AMD has also been muttering that their Deneb platform will be a lot cheaper than the Nehalem package, but let’s face it, that’s not saying much.
Denebs aren’t going to be for everyone. They’re not going to be for people willing to pay for top performance. They’re not going to make anyone replace a current Penryn or even Kentsfield system. Their potential market are those rather behind the curve now who want four-cores and Penryn-like performance for an un-Penryn-like price (or at least out-of-pocket expense, people may well recycle DDR2 memory they already own).
However, these Denebs do look like they should be able to beat the venerable Q6600, and at least hold their own against the low-end Penryns like the Q9300, even overclocked. It might even keep it close against a Q9550. If nothing else, that ought to retire the 6600 and push the 9300 down in price.
But what do you think? Is there any realistic chance you would buy a Deneb system, provided it got at least close to a Penryn system? Would you need to save money to buy such a system, and if so, how much?
The comments section is below, please consider adding your two cents. Of course, if you prefer, you can always email me. What I really want is to get a feel as to whether Denebs have a real chance with this audience or whether it’s too little too late.