The definitive Fall Refresh

After NVIDIA released the TNT2 Ultra, we saw the first incarnation of the now common “6-month product cycle.” The strategy was the exact one used to dethrone 3dfx, and is based on a very simple principle of using parallel design teams. If you have three design teams, one working on the current generation product, one working on the 6-month follow-up and one working on the next-generation solution, assuming all teams work efficiently, you should be able to maintain a stream of GPU releases in 6 month intervals.

To make the job a bit easier, you only work on inventing new architectures every 12 months, giving you a little break in between the hectic lifestyle of a GPU design engineer. But in order to maintain competitiveness you have to have a product every 6 months, so in the time between architectures you simply “refresh” your current generation architecture. A refresh is generally a higher clocked GPU, potentially with some faster memory, made possible due to more experience with manufacturing that particular GPU (yields improve over time) and the availability of faster memory. Sometimes we get advancements in process technology that allows for a boost in clock speed as well.

When NVIDIA introduced the 6-month product cycle the idea was that new architectures would debut in the Fall, and refresh products would hit in the Spring. The delay of NV20 (GeForce3) changed things around a bit and the GeForce2 Ultra became the first “Fall refresh” product. Since then, little attention has been paid to when various GPUs hit, as long as we get something new every 6 months we’re happy. Earlier this year we heard that both ATI and NVIDIA would be releasing their true next-generation hardware next Spring, leaving this Fall as the refresh cycle.

ATI’s high-end refresh was the Radeon 9800 XT, and as you can guess their midrange refresh is the new Radeon 9600 XT. Much like the Radeon 9800 XT, the 9600 XT only adds two features: a higher clock speed and support for OverDrive.

The Radeon 9600 XT GPU now runs at 500MHz, a 25% increase in clock speed over the 9600 Pro’s 400MHz clock. The memory speed of the Radeon 9600 XT remains at 300MHz DDR (effectively 600MHz), so there is no increase in memory bandwidth over its predecessor.

The hefty increase in clock speed is due to improvements in process technology as well as the introduction of a low-k dielectric. As we briefly explained in our 9800 XT review, the benefits of a low-k dielectric are mainly related to shielding from crosstalk in high transistor density chips, which gives us the clock speed boost we see with the 9600 XT. Because we’re just talking about an increase in core clock speed, the games to receive the biggest performance boost from the XT would be those that are GPU-limited, which unfortunately are few and far in between these days. Games that are largely shader bound such as Half Life 2 will definitely enjoy the 9600 XT’s increase in clock speed, but for now we’ll see most of the performance benefits go to waste.

We explained OverDrive technology in our Radeon 9800 XT review and tested it in our Catalyst 3.8 driver update. The Radeon 9600 XT includes an on-die thermal diode that measures the temperature of the core; when the temperature is cool enough the driver will instruct the core to overclock itself by a set margin. The Radeon 9600 XT will run at one of three speeds depending on its temperature: 500MHz, 513MHz or 527MHz. The combination of this driver and hardware support makes up ATI’s OverDrive feature.

OverDrive is currently not enabled for the Radeon 9600 XT in the Catalyst 3.8 drivers, we will have to wait for the Catalyst 3.9s before we can test the 9600 XT with OverDrive. If you’re curious about the performance implications of enabling OverDrive, have a look at our Catalyst 3.8 review – it’s nothing to get too excited about.

Index The Test
Comments Locked


View All Comments

  • Anonymous User - Friday, October 17, 2003 - link

    As an owner of a GeForce2 Ultra (a card that continues to hold its own!), just about *any* card today is going to be a noticeable performance gain. Based on the fact that I can get the 9600 XT for $200 with Half Life 2, and I could get the 9600 Pro for $150 (w/o HL2), I'm gonna run out and grab me an XT.

    btw: I ran the FF XI benchmark last night and scored 1650 on high resolution and 2444 on low resolution (on an Athlon 1.4GHz, 512MB PC2100 DDR RAM, and a GeForce 2 Ultra)... heh, a far cry from the 5000+ generated from the 9600 XT.
  • Anonymous User - Thursday, October 16, 2003 - link

    It's stupid not to test ALL video cards using the best system possible, otherwise you can't isolate performance differences, in fillrate, etc.

  • rms - Thursday, October 16, 2003 - link

    Inquirer says a newer considerably faster 9600XT is coming out:

  • Anonymous User - Thursday, October 16, 2003 - link

    The packaged benchmark in Halo? I sure wish I could find it....
    Any help would be appreciated.
  • AlteX - Thursday, October 16, 2003 - link

    #52, I didn't say AT should test the cards on all the possible systems or on the exact system I intend to buy. I just said that it seems more appropriate to me if they test the mainstream cards on a typical mainstream system and the high-end cards on a typical high-end system.

    It seems weird to me that a dude would shell out say almsot $800 for an Athlon 64 FX just to get crippled by a $200 video card. I think that most "real-world" systems have components that all belong to a similar price/performance range (low-end/mainstream/high-end), and testing a mainstream video card with a high-end CPU won't really show "real-world" performance, even though real games are used.

    What interests me personally (and maybe many other mainstream gamers ;)) is what is the cheapest mainstream gfx card with which I can still get playable framerates at nice IQ settings in the tested games rather than how all these GPUs compare to one another, which is more of an "academic research" than "practical information"... but that's just my very humble opinion... :)
  • Anonymous User - Thursday, October 16, 2003 - link

    Hey, just curious, I need a PCI only video card. What's reccomended for this type of application?
    IE: A good cheap well rounded video card with PCI?
  • Anonymous User - Thursday, October 16, 2003 - link

    #51 It is not my monitor. A Radeon 9000 Pro does not show this problem and neither does a GeoForce2 MX or Matrox G200. Another monitor showed the same problem with those two Radeon 9600 Pro cards.

    The moire effect is caused by the monitor, but I was not referring to that.
  • Revolutionary - Thursday, October 16, 2003 - link

    Update for 9500 Pro lovers:

    I just checked out HotHarware's review of the 9600XT.

    On Unreal Tournament 2003 Citadel FlyBy, at 1024x 768, "maximized graphical (sic) settings", and no AA/AF, the 9600XT scored 111.38 average FPS.

    On my 9500 Pro, at the same resolution, using HardOCP's High Quality settings (which I am assuming are similar to HotHardware's "maximized" custom INI settings) I get 119.44.

    Its not a lot, and its not as scientific a comparison as I'd like to make, but in UT2K3, anyway, the 9500 Pro seems to still top the 9600 line.
  • Revolutionary - Thursday, October 16, 2003 - link

    The Zalman heatpipes are the way to go. I've got one on my 9500 Pro and its passive (no fan). The fan bracket that comes with it can be used to attach an 80 or 90mm fan easily.

    I don't think its practical a handful of dudes to try to scale all their tests to a variety of different platforms. They aren't interested in telling you exactly how well it is going to perform in your system based on your systems' specs, and I'm not interested in reading it. They are going to tell you how well it performs objectively, by isolating the GPU and keeping the rest of the system as advanced as possible-- the constant. Any variation is the result of the variable: the GPU. If you want to know precisely the performance you can expect on your system, buy the system and run the tests. AT's methodology is intended to show you what you might expect, not what you should experience.

    They stated that they couldn't test the OverDrive OC feature because it isn't supported for the 9600XT in the current drivers.

    As for comments, I too wanted to see the 9500 Pro in there. It typically outperforms the 9600 Pro, and for my money, that's where the real challenge for ATI's mainstream products lies: not in besting the 9700 Pro, but in actually besting the previous generation mainstream product. The 9600P didn't do it. Can the 9600XT? Guess I'll have ot buy one to find out.
  • Anonymous User - Thursday, October 16, 2003 - link

    #47 ... any card will do that, and the higher the res, the worse the problem -- its your monitor.

    Along with the problem you mentioned, it is also very easy to see the moire effect in your line of crap. that's not vid card eithre.

    your welcome, have a nice day

Log in

Don't have an account? Sign up now