DDR Or Not?

There’s an article over at Anandtech which talks about the wonders of the KT133A vs. DDR systems. Let’s get a few things straight.

  1. Memory by itself is not a huge performance booster

    Even under optimal conditions (which we don’t have yet), we’re looking at a maximum performance boost of roughly 10% from DDR over SDR options (probably somewhat more than that for games).

    Nonetheless, 10% is nothing to sneeze at. It’s the equivalent of almost two speed grades nowadays. Going from 66Mhz to 100Mhz, or 100Mhz to 133Mhz FSB only got you roughly 3%; so as memory goes, that’s a pretty big boost.

    Does that mean you should run out and buy DDR all by itself as soon as you can? No, we’ve said before that you shouldn’t, but rather should buy it as part of a Palomino upgrade.

  2. Gamers take note

    Games take greater advantage of memory bandwidth than most other applications. They’ll do so even more when increased bandwidth becomes generally available to them. You won’t get that sticking to SDR.

  3. The 133A makes little sense to most people, at this point

    If you have a current socket A system and socket A CPU; upgrading to a 133A board makes no sense at all; if DDR isn’t worth it, this is even less worthy.

    If you are considering a new system now, you have to ask yourself if you can wait for a Palomino/DDR system in a few months. If you can, then wait! Putting money aside is not a crime!

    Buying a current Athlon chip and a DDR board doesn’t make a whole lot of sense, either, because you’re going to pay a lot more than you should for RAM. To add insult to injury, the first-generation DDR boards look
    like they’re going to be doozies to deal with. Unless you like pain, why do it? Let people like us suffer instead. 🙂

    If you can’t wait, even then, the 133A is not an obvious choice. The reason for that is motherboard maturity. If you have to go through all sorts of conniptions to get another immature platform to work; that’s not worth a few percentage point improvement.

    From what we’ve seen so far of “evolutionary” steps of the Via Athlon platforms, we would not bet the 133A will build on the experience of earlier chipsets. If it does, fine, but we say “Show me” (and we’ll certainly test and run a couple so you’ll know).

    If you’re going to buy a Palomino, stop thinking about this until one actually comes out. Then evaluate the situation. No doubt some of you like your current RAM very much and don’t feel like buying more. OK, there will be hybrid boards that will take either SDR or DDR, that might be for you. That may be a more prudent option than buying a motherboard you’ll have to toss once you can’t resist DDR.

    The next few months will be unsettled ones; I have no idea what I would recommend three months from now. I would hope DDR systems will mature enough to earn a recommendation, but events could turn out quite differently.

    Let’s see what actually works well first. Let’s not be impatient, and respond to every siren call to buy something new until we see if it’s worth it or not.

    If the 133A proves to be a rock-solid product, it may well be the best choice for many of you. Let’s see that first. If the 133A proves to be rock-solid, and DDR boards flounder around, it may turn out to be the best choice for many more of you. Let’s see first.

Email Ed

Be the first to comment

Leave a Reply