I doubt it. Jason has purposefully said he doesn't turn all these things said in confidence into a story. Given his track record, he can actually afford to be discretionary.
100% true. Cerny himself said that the PS5 was 4 years in the making in the orig Wired article.
I haven't seen a single person state that early dev kits were GCN. Possibly you can provide this source.
www.reddit.com
www.google.com
He said GPU's like a 2080, and that performance equivalent is good enough for me.
I don't think anyone that says 10TF RDNA = 12TF GCN means it literally. I think when they say that they are talking about performance equivalent... or at least I hope they are. Cause that is the only way that such a statement would make sense.
Exactly. And that's is what makes what MS did with the XSX ambiguous. They made a statement that literally does both things.
I think that is because that 13Tf number was given at a time when no one knew of RDNA. I don't personally believe the PS5/XSX wold be 13TF RDNA, but 13TF GCN (which would be equivalent to around 10TF+ RDNA) sounds about right to me.
We must respectfully disagree. Also, tomwarren had already hinted 12 and 4 multiple times before the reveal. This would indicate Tom was using GCN-equivalent flops for the SX and RDNA for the SS?
There has been anything notable since Dictator discussed Raytracing but even then we have no official anything from either company. We are currently in the desert walking around the same mountain, the promised land is close.
You think they're going to have any large scale briefing during this health crisis?
I don't know enough about the first sentence but the second doesn't make any sense to me. That's not why you would put quotations around something like that. There could be different reasons to do it, but the one you give isn't one of them, imo.
Please don't. Let us see this slowly burning and how much will stay sane.
21th of March ?! One day AFTER GDC seems to be a little too late to the party. There is a reason why conference at events are made before opening.Transistor
So, number of voters seems to have stabilized (169 votes total), and a near final analysis can be performed:
So, according to the ERA hive mind we have two most probable dates: 25th of February, or 21th of March. So far there is an single vote for the latter day (jikevgw), is he/she our ultimate champion and god of prediction? :) Stay tuned. :)
I think it's not a proven fact yet that purely gaming h/w ("console") needs specialized ML h/w. DLSS may be great now but there are still other resolution reconstruction options which can arguably give comparable results in both quality and performance uplift while not requiring as much specialized h/w - which may not be that useful for gaming otherwise.
Both? Yeah. EA mentioned it in the 2nd wired article about PS5, and there's no reason why MS wont so as well.
I think that he was hedging in March and when XSX validated that it was above 10.7 TF of Google power he said something about it. For the record, I think both are above 10.7TF if we are comparing it to Google's Stadia, so Jason would be right about that. There was some debate if that would be the case back then, though. It wasn't a for sure thing and that was what Jason was talking about, imo. His sources were telling him they were aiming above Stadia performance. That is why the quotations were used. My take on what Jason said in March of 2019 is that "these things are more powerful than Stadia" and not "these things are 10.7TF RDNA or above."
