More than a few of you have wondered out loud in emails about Dothans possibly being the next Great Blue Hope while PressHot stays, well, hot.
A few manufacturers seem to be having the same thought, just for one example, Hitachi is going to have a line of desktops powered by notebook chips.
Will this be a new overclocking frontier?
Don’t get your hopes up too high, for the following reasons:
Dothans Desktops Will Be Expensive
The slowest Dothan available (the 1.7) has an official price of $294. The quickest (the 2.0) sports a $637 price tag.
By the last quarter of the year, at best, the 1.7s ought to cost a bit over $200.
There will be Celeron versions of the Dothans eventually, and they’ll carry a sub-$200 price tag, perhaps approaching $100. However, like their desktop equivalent, Celeron-Ms only have 1/4 the cache of their big brothers, and that’s not exactly what you want in a performance machine.
Unfortunately, the real sticker stock comes at the mobo level. A current socket 479 mobo is both expensive and rare. The prices I saw looking quickly around for some came in at about $400.
No doubt mass production would drive down the price, but would we ever see mass production for an item like this?
Dothan Desktops Will Have Performance Handicaps
Right now, the Dothan platforms are 400MHz, single-channel memory, or the same as Willamette. Need I say more? Well, just a little more. Intel plans to raise the FSB to 533MHz at the end of the year.
So far, any benchmarking done with Dothans have been done with benchmarks in which CPU power is everything. Move to a benchmark where memory speed or video counts, like a game, and the results will likely be rather less impressive.
Not surprisingly for a notebook platform, mobos for these processors are big on integrated graphics, which is not exactly bleeding-edge even under the best of circumstances. It’s possible to get a reasonable decent video card, but expect to pay more and get less.
Dothan Desktops Probably Won’t Overclock All That Much
The whole point of Dothans is more instructions per cycle, fewer GHz. If this sounds like “back to AMD’s way of thinking,” well, you’re right. 🙂
Intel doesn’t plan on scaling Dothan very much. By the end of the year, the fastest Dothan will be only 2.13GHz. The best overclocking attempt with a Dothan I’m aware of so far is 2.4GHz.
Maybe they can do more than that, but probably not much more. What’s important to understand is that Dothan is a rather different design than a standard PIV. On the one hand, you get more bang per GHz, but the downside to that is that you can’t crank up the volume like you can with a standard PIV.
Conclusion: Not Yet
The Dothan CPU itself is fundamentally fine as a speedy (if not speediest) processor, and in a year or so, Intel probably will be able to fashion an excellent desktop home for it.
However, if it’s going to take Intel a year to craft a performance infrastructure around it, that ought to tell you this isn’t really a DIY job today.
For now and the near future, a Dothan desktop seems much more feasible as a Quiet PC that has more oomph than the most silent and less noise than the most oomphy rather than as a highly competitive performance machine.
Perhaps we’ll see this infrastructure appear incrementally as work progresses, maybe we’ll see a desktop mobo with features and specs more like the current PIV standard than exists now towards the end of the year.
Until we do though, for the average and even not-so-average overclocker out for performance, Dothan really isn’t a good choice. Yet.