By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Machiavellian on 02 October 2013
Edit Thread
Scoobes said:
Forget console bias, the first article was simple incompetence. How are ROPs not part of the "graphics performance"? They're a pretty essential part. If he only wanted to test the performance of Compute Units he should have made that clear, but he said "graphics performance based on available GPU specs". That's not restricted to compute units, he needs to take into account ROPs and texture units which he completely fails to even mention. The following choice of cards was ridiculous at best and the result was a useless metric, even "in theory". It's a sloppy investigation or a preliminary experiment at best.

As for the second article, you have developers who've worked on the PS4 coming out and claiming the article was false:

If you don't want to believe a PS4 developer then I don't know what to tell you.


Actually ROPs does not play into the scope of the article.  The article was not on total GPU performance as the quote I posted explained.  It was scoped to guage GPU scale based on the number of CUs.  This is what I am talking about.  You went into reading the article based on what you thought it was about, instead of what the scope of the article tried to accomplish.  One of the main things the net has argued over is that the PS4 has a significant amount of GPU TFlop power.  The article tried to guage how much scale does having more GPU grunt actually provide.

 Unless the PS4 and X1 is going to be using displays past 1080p, ROPs really do not play that much of a difference.

For the second article, Richard stated he has developers working on the PS4 giving him information.  Whats the difference.  The difference would be who you believe more than anything else.  The difference I looked at is from the source.which is Sony.  Sony could have easily cleared all that up but went the no comment line.  If no comment speaks volumns for MS, it should do the same for Sony.

As to your link from Brian Provinciano, who said he was writing a full editorial on the subject.  Did you read that editorial.  He stated NOTHING to disprove the Ram allocations.

Here is the link from Brian editorial

http://retrocityrampage.com/blog/2013/07/522/