By using this site, you agree to our Privacy Policy and our Terms of Use. Close
curl-6 said:
Darc Requiem said:
The Xbox CPU was basically a Celeron. They called it a Pentium but the cache said differently. The Gamecube essentially had a Power PC G3.

 

As someone not too familiar with CPUs, how does this make them fare aginst each other power wise, wth the Gamecube's CPU clocked at 485MHz and the Xbox one at 733MHz?

 

 

 

I orded that wrong; they didn't say that the GC needed the GPU to do physics, but rather that it had to useit's CPU to do geometry calculations, leaving the Xbox CPU with more power to spare.

Here's the quote: 

"The GCN is more powerful than the Xbox at Floating Point calculations all by itself. It is significantly weaker doing Integer Calculations though.


HOWEVER, this is misleading, unless viewed in the proper context.

The GCN uses it's CPU for raw triangle setup, and base 3D rendering. The GPU only handles the final stages of performing Transforms, and any fixed hardware lights.

The Xbox uses it's GPU, NOT the CPU to perform all 3D rendering, which means that the Xbox CPU isn't being weighted down with doing 3D calculations. 

Since it is the 3D rendering that uses the Floating Point processing power the most, the advantage the GCN has in that one area is totally lost. The GCN relies on it's CPU to perform the Floating Point calculations for 3D rendering, while the Xbox does NOT use it's CPU for the same job. This leaves the Xbox with more Floating Point processing power to use on non-3D rendering tasks, as well as it's superior Integer Calculations performance."

http://forum.teamxbox.com/showthread.php?t=98510


What would you expect from a person posting on teamxbox in favor of the Xbox? He's selling a mistruth. The CPU does indeed do the calculations for it but it really doens't take that much of a hit for it at all. Rogue Squadron, once again, shows this all to well. That statement is a common fallacy known as inconsistent comparison. http://en.wikipedia.org/wiki/Inconsistent_comparison

He is stretching a fact beyond its implications. The CPU in the GC is nearly twice as capable as the one in the Xbox if not more, so even if it used a quarter(which is unlikely) of its processing power for calculating geometry it would still have a lot of power left over. The CPU in the GC also has  a lot of special capabilities beyond sheer brute force that the Xbox1 processor does not posess. The comment is simply stating that because x does this, y happens without stating how x results in y.

People take raw numbers and give you all sorts of conlusions, but the real world results always tell an entirely different story. Microsoft claimed the Xbox1 could push 120mil polygons but the most ever gotten in a real games was 12 million at 30fps.

 

The first thing I usually see when someone is comparing the Xbox1 to the GC or even the Wii is a copy and past of the clock speeds with no understanding of the cycle rate differences between the PowerPC and the Pentium or any technicle aspects. They see a higher number on the Xbox1 and immediatley conclude that it is stronger as an absolute fact. Ever since the AMD and PowerPC processors came into existence, clock speed has been little more than a measurement of how much electricity it takes for your CPU to do something, not how well it can do it. That is another issue though.