• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Next-Gen PS5 & XSX |OT| Console tEch threaD

Status
Not open for further replies.
lmgd9xw.jpg

117 now, is there a record 😂

One ticket please!
 

Bogroll

Likes moldy games
It was not my intention that it seems that I do not consider some users and I am sorry for it. Undoubtedly I am very grateful to all of them. I am nobody but a simple human worker. And although I may be right in some data, I can also be wrong as a person when choosing inappropriate words. So my sincere apologies in case I have offended someone who shouldn't have.

I feel highly valued and respected by the vast majority, both those who respond to me and those who do not. And I am grateful to all of you.

And let me publicly thank Bo_Hazem Bo_Hazem for giving me a month of Gold. I didn't have to, but it was an unexpected detail.



I do not know if I have finished understanding your question very well, but if not, excuse me and ignore the message.

If we talk about PC yes we could affirm (obviously) that if you buy a 1070 (base) you will not get better performance than if you buy a 1080 (base). Then it is true that you can buy versions with OC and better cooling that can increase overall performance. Still, the normal thing is that those cards are still less than 1080 (base). Although there could be rare cases (as a possibility).

But if we talk about consoles it is not easy to ensure that. You go to the factory and ask the manufacturer what technologies it has. And you choose a base. Then you modify that base to your liking in all aspects. Therefore, although both consoles start from the same base, it is possible that the final result (insurance) will be totally different. In the same way it is also possible that even starting from different bases (from the same cycle, such as a 1070 or a 1080), better results can be achieved in the 1070. But not because it is a 1070, but because once the customization is finished that it is no longer a 1070. It is something else. Because we are not talking about customizing the GHz or the amount of memory. We talk about customizing the entire architectural ensemble. Removing things, putting things, changing components, adding components, etc. Logically if one part of a 1070 and the other part of a 1080 and you do not customize them then the 1070 will always get worse results. But is not the case. Other than that this is just an example. And also, in the same way that one can customize a 1070, the other can also customize a 1080. In this case, the important thing is the "amount" of customization and the "intelligence / experience" when it comes to doing it. If you add to all this that you possibly have your own technology that you can add to those bases, the result is even better if you know how to combine it well without generating bottlenecks. So, in this case, it could be that it was indifferent from where you start. Always within limits, of course.

I repeat, all this has been an example, it is not the numerical reality. But I think it is understood.

For this reason it is absurd to try to compare these systems with supposed equivalencies on PC. It does not make any sense. Because you can start from a "low" base and get "high" results. If the job is done well, of course.
I'm on about PC cards as its a fairer way to compare the TF and in no way using it to compare to the consoles just how useful it is as a yardstick. I get that people use PC cards as a rough comparison like DF did with the 2080 super (i think) in the Gears 5 XSX video.
I was thinking along the same lines but with the 5700 being overclocked and outperforming the higher cu of the 5700XT but in turn does pushing the clocks higher on the 5700 push up the TF.
 

IntentionalPun

Ask me about my wife's perfect butthole
TFs don't paint the whole picture; but the one thing that confuses me:

On PC some of the highest end cards accomplish their power by having more parallel processing at a lower frequency. Like they outright perform better that way without any dev having to code for them.

So how does the statement from the CryTek dev even make any sense? CUs aren't like CPU cores.. it was not a great analogy from the perspective of how a dev would or wouldn't take advantage of them.
 
Last edited:

Bo_Hazem

Banned
Is it possible to find out who did that? My profile went gold weeks ago but until you said this I assumed it was down to post frequency or something else.

My apologies if anyone gave me that and I ignored them. I have all email notifications etc off and didn’t realised it was a paid thing.

He probably gave it to you anonymously, on the other side I'm a bragging mofo who's feeling high with this hype dosage.

149590.jpg
 
Last edited:
Man I'm excited for the reveal. With my state opening back up our cases haven't really been going up much from virus. Life is feeling like it's finally slowly returning to normal. I hope Sony drops a bomb and reveals the price+date so I can pre-order at retails now!

giphy.gif
 

geordiemp

Member
I'm on about PC cards as its a fairer way to compare the TF and in no way using it to compare to the consoles just how useful it is as a yardstick. I get that people use PC cards as a rough comparison like DF did with the 2080 super (i think) in the Gears 5 XSX video.
I was thinking along the same lines but with the 5700 being overclocked and outperforming the higher cu of the 5700XT but in turn does pushing the clocks higher on the 5700 push up the TF.

Are you timdog, its been covered a million times already, but for 1 million and 1,,,,

RDNA 1 and GCN are limited in frequency and older, and have limits on speed and performance. AMD hada slide for this already, + 50 % perf / watt.

This has NOTHING to do with RDNA 2, which Cerny says gives performance to 2.23 Ghz after which logic is impaired.

There are no PC RDNA 2 cards, so stop talking crap;
 
Last edited:
Status
Not open for further replies.
Top Bottom