You Gotta Have It? . . .

Intel’s quad-cores have been released, and inevitably, there are plenty of propagandist “reviews” that tell you, “You have to have this now.”

Uhhh, noooo.

Our objection to quad-core has been that there’s no point laying out an arm and a leg for a processor until it either does you some real good, or it stops costing an arm and a leg. Problem is, neither is going to apply to most people any time soon.

Now those who insist you need one now aren’t saying, “You need to buy this now because CPU companies need to make a lot of money off somebody these days,” but it’s pretty hard not to come to any other conclusion after seeing their benchmarks. Excluding a few professional apps which specifically use multithreading (and even most of them aren’t terribly well optimized for four yet), there’s little improvement from four core (and in all honesty, it’s not like everything is tweaked up for two cores yet, either).

But of course they can’t say that, so instead they begin by saying what their own numbers force them to say, but then, to make the marketing people happy, they pay the piper and tell you that games will be taking advantage of four cores next year.

You have to read these articles closely to discover that:

  • only a handful of games plan on using four-cores
  • they won’t show up for about a year
  • the games won’t take full advantage of all those cores

    A Better Strategy

    We’re not against four-cores forever. Its time may well come, it’s just not now.

    A good rule of thumb for most technoproducts is, “If they think the average person needs it, then they should be ready to sell me one for an average price.” For CPUs, let’s be generous and say $200-$300.

    Yes, this will come to pass sometime in 2008, once Intel has 45nm down. In early 2008, it would be a Yorkfield, and I’m sure AMD will come up with something to match that. It’s also safe to assume that whatever it is, that $200-300 processor will be better than today’s $900-$1,000 budget-buster. In the case of Yorkfield, it will include SSE4 instructions, which I suspect will be of great use to game designers at some point.

    By then, too, those games promised for the end of 2007 may actually be around, and even if they get delayed a bit, it’s a lot better spending $200 for something a little down the road than spend $1,000 only to increase your neighborhood’s unemployed core rate.

    The Peeve

    It’s hardly a new pet peeve, but I get irritated when manufacturers treat the word “gamer” as a twenty-first century magic spell; that if you can just use the word in a marketing sentence, people just lose all their mental and financial inhibitions and zombishly buy.

    Well, what really irritates me is that it apparently works well enough for those manufacturers to keep doing it.

    What’s next? Gamer clothes? Gamer condoms? Gamer bread? (Don’t laugh, there’s already an “Xtreme Bread” out there.) Really, can’t people realize that if they just spent the time to learn how to PTFG (hint, it’s a descendant of RTFM) they’d find that a PC Lexus is unnecessary 99.99% of the time?

    I know, that’s heresy, blasphemy, treason; but it’s just the truth: don’t be a tool all the time.

    Ed


  • Be the first to comment

    Leave a Reply