Edge

Honestly, I wouldn't be surprised if they used a Radeon on purpose so that the XBox comes out on top.

Microsoft and EA do have some partnership ties dealing with the XBox. I wouldn'd doubt there is some bias involved in the tests setups.

Gallant Pig

You mean like not having Madden on XBL and instead of bringing the PC version of MOH to the Xbox porting the overrated (IMO) PS2 MOH to the Xbox as single player only?

Sorry but EA and MS aren't in bed together and your conspiracy theory doesn't add up.

Tamrok

What would be the reason that EA would be biased toward the Xbox when the majority of it's sales are on the PS2?

ScandalUMD

Based on their description, they're using graphics chips entirely for graphics, and no CPUs here. That means it's the Xchip vs the PS2 GS vs the Flipper vs the Radeon. The CPU's don't come into play. Incidentally, the CPU is the Xbox Achilles heel. I don't think EA is biased against anyone; after all, their program is limited by the PS2 GS, and will fail to utilize the Xbox. I think they just cut corners.

ChrisKnudsen

Where's the Dreamcast? I wouldn't mind seeing the times compared with this just to see how next generation systems compare to the first.

Kellehair

Did EA ever make games for the DC? Even if they did I'm sure they don't anymore.

ten41

Let's end the conspiracy theories.

belboz

Just to clarify some points. The NV2A core in the XBox technologically falls somewhere between the NV20 core in the Geforce 3 line and the NV25 core used in the Geforce 4 line. The R200 core used in the test is a little more advanced the NV20, but as I mentioned above, it lacks the second vertex shader which is what probably accounts for much of the NV2A's edge.

There are a couple other reasons why the XBox is better optimized for gameplay than a standard PC. As Tamrok mentioned above, the XBox runs a stripped down version of Win2K where everything runs in the kernel memory space. In regular Win2K and WinXP, the kernel runs in a privileged memory space for memory protection, so that applications can't corrupt the kernel. However, for applications that have to work intimitately with low-level hardware (ie 3D accelerators), this memory protection exacts a significant performance penalty.

The other main reason the XBox has an advantage in games is that it uses a UMA design. The port from PC to XBox isn't as simple as most people think for this reason. Unified Memory allows for some efficiencies in the way that memory is managed in a game which allows for improved performance with 3D games (duh, or MS wouldn't have designed it that way).

Also, it shouldn't be surprising that the much faster CPU in the PC isn't being reflected in this benchmark. This benchmark is purely a test of the graphics subsystem. There's no game physics, or AI going on. In terms of 3D graphics, the CPU hasn't been the bottleneck since hardware T&L was introduced in GPU's.

The real benefits of having a faster CPU won't be apparent except in actual gameplay and for the most part, it won't be manifested in better graphics. It'll be in more subtle things like more realistic physics, smarter AI, etc.

Groucho

Doesn't make much sense to me. It's obvious to even a casual observer that the GCN is much more powerful than the PS2, but these numbers show just the opposite.

Tamrok

I think you're confusing raw polygon performance with actual graphics quality. The PS2's strength has always been in pumping out a lot of polygons. Where the PS2 runs into trouble is with it's tiny amount of texture memory which results in low resolution, bland textures in it's games. On the other hand, the Gamecube, while not quite as adept at producing a lot of polygons, is much better with textures and can accomplish numerous graphics techniques in hardware (such as bump mapping, multi-texturing, etc.) that must be done in software (if at all) on the PS2. The result is that while PS2 can output more raw polygons, the Gamecube excels in producing fewer, but higher quality polygons, with higher resolution textures. Thus, the graphics end up looking much better on the Gamecube. This is a perfect example of why polygon performance is only one part of the overall graphics performance picture.

jeffdsmith

Shoot, the gamecube is the weakest of them all. (yeah right)

All that fun I've been having wasn't as good as I thought it was...

joltaddict

If you read that anandatech link I posted on the first page youll see how much graphic muscle that little box has. Better opimization beats raw specs 99% of the time.

http://www.anandtech.com/showdoc.html?i=1566&p=1

Tamrok

What exactly is your point with this post? I don't recall anyone (other than you) saying that the Gamecube is the weakest. Clearly, it isn't.

joltaddict

*shrug* Maybe thats how he interpreted that.

Kellehair

Lighten the **** up. It was a joke.

WOWZY

There was an interview with Nvidia late last year on their NForce technology that's in the Xbox.

I've tried to find the article, but to no avail.

Just going by memory on what Nvidia said, they said that their new NForce chip had beaten their GeForce Ti500 3to4 fold and had beaten there upcoming (out already) GeForce4, 2 fold in all of their benchmark tests and that developers should have no excuse to putting out top notch games that blow away HALO graphics wise. I guess they were hoping developers would change from their lazy ways.

Nvidia also said that when Nforce cards come out for the PC that the rest of their competition will have alot of catching up to do.

Sorry that I can't find the article. but those were the main things that stuck in my mind when I read it.

I'm not an expert but it seems that the Nforce is doing everything that Nvidia said it would based on those tests.

belboz

The nForce* is nVidia's chipset for AMD Athlon motherboards. It's extremely similar to the chipset that's used in the XBox. The south bridges are identical, but while the north bridge in the nForce chipset only has an integrated Geforce2 MX core, the north bridge in the XBox chipset has the NV2A core.

The NV2A is indeed more powerful than a Geforce 3 core at the same clock speed, however, there's no way it's 3 to 4 times faster. Nor is there any possibility that it's 2 times faster then the NV25 core used in Geforce 4 Ti boards.

*There's also an nForce2 chipset that's been announced, but won't be shipping for a few weeks yet. It's big improvement over the nForce is that it supports dual channel DDR400 memory and integrates a NV17 (Geforce 4 MX) core in the north bridge.

Tamrok

A poor one, perhaps. Seems more like sarcasm to me.

jeffdsmith

Indeed it was. Perhaps it was ill concieved for some, but it wasn't that bad.

Joltaddict, I've read over Anandtech's summery long ago, I appreciate you trying to clear the matter for me but I was just messing around.

I guess my point was to poke fun at almost any of these "Comparison tests" because they will always have the GC at a sever disadvantage because of the archatechture Nintendo and IBM chose to go with.

I did not intend for so much confusion. My apologies.

joltaddict

It seems you can make a joke about anything except the Cube around here.

jeffdsmith

hmm. okay. I'm just going to let this one die out.

Shawn

Because you know it's true?

Kellehair

jolt: I'm sure you and Tamrock were just defending the Cube.

Shawn

Umm...I haven't read this whole thread but I saw this originally posted by Tamrock And originally posted by jolt: Also posted by Tamrock: I don't understand how if you say something bad about the Cube, you are considered bashing it, making someone show that all your posts aren't anti-nintendo, but if you say something bad about the Xbox, it's okay.

Kellehair

I don't see why they can't take a joke.