Xbox 360 and PS3 GPU's revisited?

NamelessMC

Senior member
Feb 7, 2007
466
0
0
Okay, 2 years ago, mid 2005, everyone was up in arms asking for PS3 and 360's to be dissected and analyzed. The general consensus was that they were roughly equal in power, but the PS3 supported higher resolutions and had beefier processing power.

The problem was, in 2005, the Xbox 360's GPU really couldn't be classified in real processing power compared to a real video card.

We all know, the PS3 used a G70 chip similar to the 7800GTX.

The 360 on the other hand, used the Xenos chip which was supposedly the technology ATI was adopting for its "at the time, future" planned R600.

Now, after seeing the performance difference of the R600 compared to the 7800GTX for example, one could argue that the unified shader technology in newer video cards is more efficient, but the question is how much more efficient?

I want to gather information, to write up a "revisited" comparison of the two, so we can find out, now after having a larger sample size, which video game console does in fact, have better hardware.

Responses I don't want to see-
"Who cares? They both look good". Uh, I care, hence making the topic. If you don't, why read and why bother posting? You could tell from the title what the topic was going to be about.
"X game on Y system looked better" The problem with using actual games as comparisons is that there's always the question of whether or not a game was ported from one system's development to another. Perfect scenario is Need for Speed: Carbon. We don't know if the PS3 version is a port of the 360 version, or a re-developed console version based on original code. The developer won't say either way so it's really a wild goose chase.

Wiki information on "Xenos", 360 GPU developed by ATI
http://en.wikipedia.org/wiki/Xenos
Wiki information on "RSX", PS3 GPU developed by Nvidia
http://en.wikipedia.org/wiki/R...Reality_Synthesizer%27

Other useful information - The PS3 was originally going to be designed to use Cell processing for GPU, but close to the end of production, Cell was found to be terribly inefficient in real-world API developing. They needed something with better developing ease of use that had real-world testing. They went to ATI first (That's unconfirmed), but ATI claims they told Sony, "We would need more time to develop a GPU." Sony didn't have time, so they went to Nvidia. Nvidia said, "Well if you don't have time, we can put our best GPU in it with 256mb of Video-ram." That part IS confirmed.

Also- Despite the fact that the RSX is listed as having 22.5 GiB of bandwidth, one has to note that the Cell FlexIO interface has a bandwidth speed of 20 GiB reading and 15 GiB writing.
 

AzN

Banned
Nov 26, 2001
4,112
2
0
I wouldn't say PC graphics card unified shader are more efficient. They are just more powerful beasts compared to xenos.

Xenos is really a midrange GPU with 8 rops running @ 500mhz comparable to 8600gt/gts.