That's what I'm saying. Cerny came out almost a year ago and said the PS5 had HW raytracing and people are like "Nah, he lied."
What?
Microsoft entered into an agreement with Sony a while back to help them with their cloud stuff(I don't recall when exactly).
Based on that, even if it was a mutual thing for one purpose, we really believe neither one had or has any idea what the other was shooting for in specifications?
I'm a bit hard pressed to believe that, honestly.
The biggest thing that gets me is "b-b-but Github." Nothing about it was conclusive on anything.
Also,
@OsirisBlack has sent numerous validations to Mod of War, including a screenshot that was *verified* as legitimate and somehow, he doesn't know what he's talking about when he speaks on what each system has.
He even said the XSX has a slight advantage at this juncture, though kits aren't final for either system. But slight was stressed and *slight* is not 9.2TF vs 12TF to me.
Actually there was more to Github than just that one sheet that got around with Oberon and Arden listed on it. Apparently it had information on a ton of other AMD products, too, some in quite a lot of detail. The GPU benchmarks Komachi and Rogame (among a couple others) have been uncovering reaffirm a lot of what was in the larger Github info while providing more details and context, as well as some clarifications. And several actual products have emerged officially with specifications matching the benchmarks that have been uncovered.
I think a problem a lot of people are making is looking at simply one metric (teraflops) when taking the mention from some insiders that the systems are "close". There's more to a console than its GPU and raw output. Architectural features, optimizations, memory system, CPU core/thread count, CPU architectural features, I/O subsystems, system memory bandwidth, OS features, OS optimization, audio capabilities, video output support, custom storage setup and optimization, storage bandwidth and speed, etc.
All of those things factor into a system's overall performance, so when people are saying the systems are close, I take all of those things into consideration as well as TF count. Now, none of those things can necessarily "replace" or "make up" for the other, but as another example, it won't matter much how many TFs your system has if it's bandwidth-starved to keep the GPU fed, and the storage drive for the game assets it very slow, meaning it'll take longer to populate RAM with required access (for things that need the speed of main RAM to be operated on). See where I'm getting at here?
Even in a "worst-case" scenario, if you're looking at a 9.2 - 10.24TF super-clocked Oberon-powered PS5 to a 12TF XSX, the systems are still going to be pretty close because TF doesn't mean everything. They will
both still have similar storage systems,
both similar memory,
both relatively close amount of memory, still
both using Zen 2 and virtually identical clocks,
both having a great ratio of TF to system memory bandwidth,
both featuring comparable ray-tracing,
both having VRR and VRS,
both on 7nm/7nm+ benefiting greatly from RDNA/RDNA2 and Zen 2 efficiencies, and
BOTH being notable leaps over even their respective mid-gen refreshes, let
alone the base systems.
That's what it means when both systems are close: if they are near-identical or practically identical in most of the key areas, and benefiting from the same advanced features and able to provide similar high-end functionality to the end-user, then they're practically going to be twins. That reality isn't suddenly shattered if in case one has a notably higher raw TF count than the other, because raw TFs don't mean much on their own. If they did, we'd still be using GPUs with the same architectures in PS3 and 360 but powering near 100TF. What good would that be on its own when such a GPU would lack all modern-day architectural and graphical features? Even a low-end card on the market today would handily smoke such a GPU.
The people incessantly obsessing over TFs (almost to the point of defying common sense) are acting the same way people did back when "bits" were the measuring metric. People would just
constantly use bits as the only measurement of how "next-gen" or powerful a system was, when, just like with teraflops today, they never told the full story. And just like with people who look only at how high the TF number goes and stops there, the ones who only knew how to talk about consoles in the past in terms of bits, by and large, just weren't very knowledgeable about technology or console technology in particular. Bits were yesteryear's marketing goto, yesteryear's power metric into having some type of grasp on console power by the layman. Today it's teraflops, and next gen it'll probably be something else.
So I think we've now got a full generation of gamers who don't understand how to look at technology (let alone console technology) in the scheme of the overall picture, who are eager about next-gen, may have one preference over another, and are clinging to TFs as a power measuring metric because it's the easiest thing to understand and, yes, GPU power is vital to a system's performance. So you combine all of that, then combine with the fact we have a set of sources, then combine
that with the fact that a side with one preference is only listening to one type of source because they tell them what they want to hear, and
another side with a
different preference only looking at another type of source because that's the one reaffirming what they want to believe. Combined with the fact that one side in this next-gen console race has (imho) been more directly communicative of features that diehard fans of the other side wants that other side to actually talk about, and we have the chaos that's ensued.
This is why you have some Xbox diehards clinging madly to the Github leak, because it reaffirms what they want to believe narrative-wise. And that's why you've got PlayStation diehards clinging madly to the insiders, because most of them are telling them what they want to believe in. I think people clinging hard to exclusively one or the other are going to be disappointed with how things actually play out,
but (and this is just my opinion) I feel the ones clinging super-hard to some of the insiders will end up more disappointed. I just base that off the trends of messaging from both MS and Sony, plus the fact they both likely know a more of each other's specs than we (us, insiders, dataminers, tech specialists etc.) do. And also the trends/pattern in certain data reoccurring, and staying up-to-date.
I have put my main belief into the benchmarks, because of their consistency and the fact several actual products have already come forth matching them. It just so happens to be reaffirming prior leaks, but if anything that connection shows a trend. That doesn't mean the benchmarks are the only thing I'm looking at, though. Almost everything I hear from reputable insiders, or articles from big sites that are published, what some tech-focused Youtube channels that have a record of insider sources (or just are really good at understanding tech) estimate, I take into account as well. I try seeing what they're all saying that seems to have something in common, and see how those fit with the benchmarks and leaks. If they don't exactly fit, I try thinking of what context could be missing that when added will make everything basically agree with each other.
So that's my takeaway on how to best approach all of this next-gen talk. Go with what's proven to be the most pertinent, don't exclude speculation or rumors from other source (as long as they are modestly reasonable and come from reputable people), and do some of your own work to find a way all of that might fit into agreeing with each other. Because short of actual confirmation from Microsoft or Sony, it's the best way for you as an individual to come towards a very probable outcome for both systems in terms of specifications and features. Oh, and don't let any bias blind you, that one's very important. Definitely okay to have a preference but don't let that preference blind you to actual data, especially if it's been persistent and recent (and has a track record of actual products coming forth matching the data).