They kind of are, though. And here's the thing about speculation: there's a
time and a
place. Back in fall of 2019, when we had no official confirmation of anything and the consoles were a year away, pretty much any speculation was valid because no one actually knew anything. The closest we had to official info were leaks on Github, found through data-mining. That was the only hard data we had but there was enough still left open to allow for a sea of other speculation even surrounding what the leaks had, let alone other aspects of the consoles not mentioned in the leaks.
Then, gradually, we started getting confirmations, like MS mentioning Series X was 12 TF at TGA. And almost just as quickly, new rumors conveniently springing forth that it wasn't "really" 12 TF, that they were stating 12 TF of relative GCN performance so they were closer to 9 or 10 TF. Why? Well, because even earlier rumors from some insiders said the systems were close in power, and since the Github stuff had already happened then the assumed conclusion was that MS weren't speaking about the actual RDNA number. There was even doubt cast over if it was RDNA 2, even back in December 2019. And what did so many people (mainly a lot of Sony people) say back then? "Wait for MS to officially confirm."
Which turns out was a perfectly valid thing, but you already see where this is going. Eventually MS did confirm it was 12 TF of RDNA 2, and then we started seeing new rumors pop up that the PS5 was actually also 12 TF, even 13 TF, even 13.3 TF, even maybe 15 TF! Wonder what sparked those rumors to suddenly appear
? It didn't help that you had various insiders like Matt, OsirisBlack etc. pretty much VERY quickly be dismissive of the Github leaks (and subsequent data) in absolute terms; this swayed the vast majority to also dismiss Github and if you even referenced the Github stuff even as a supplemental to what the insiders were trying to push at the time, you risked catching a ban.
Almost all of the rumors, speculation etc. around PS5 would focus dominantly on TF all the way up to late March, once the Road to PS5 event happened. Keep in mind, the vast majority of insiders still stuck to their guns regarding PS5 TF even by the time of the show, this also included people like Jason Schreir. The only insider I recall giving a TF estimate for PS5 that was pretty close to what actually came to be was Heisenberg, but he never stuck with that statement, mainly due out of fear from backlash. So Road to PS5 happens, and almost immediately the narrative shifts from TF (where it had been for months), to the SSD and 3D audio, and new rounds of rumors, speculation etc. hinged on those aspects of the consoles.
You can probably guess why a lot of those rumors came about: because most of the people who were churning them out, had a good feeling these were areas the PS5 had an advantage in, and areas the Series X was weak in, at least in terms of paper specs. The fact MS did a blog dump for Series X mentioning the SSD bandwidth, then Sony doing their own SSD I/O breakdown at Road to PS5, affirming at least one of the earlier rumors as being true (out of a sea that were ultimately false), played into this. It even would get to a point where rumors surrounding Series X said it didn't have dedicated audio hardware, or that the memory system was "split" and MS would never really hit 560 GB/s. Some took that and ran with it, even claiming effective Series X bandwidth would be around PS4's level!
By the way, quick aside: at this point more than enough ridiculous rumors surrounding Series X were about, yet I don't recall channels like RGT dedicating near hour-long videos to disprove these particular rumors and speculations...just saying
...
Anyway back on topic; as you can see, the nature of rumors and speculation for the longest time regarding these systems has always, by and large (with some exceptions) always favored Sony and PS5. That's how the vast majority of insiders have played it, and that's how the vast majority of people discussing the systems have come to prefer to view it, even if the reality isn't as the rumors and speculation try saying. And that's ultimately the problem: the nature of these rumors and speculation have ever-increasingly changed into an echo chamber of reinforcing one's preferred fantasy, in spite of what a simpler reality suggests to be the case.
Generally, as the launch of new consoles draws near, the reality should become dominant, for obvious and logical reasons. But it seems that is only by and large happening with people discussing the Xbox platforms, from what I've been noticing. We should also look at the timing of when some of these rounds of PS5 rumors and speculation pop up: they're almost always usually timed whenever some type of perceived good news pertaining to the Series systems (on a technological level) manifest. I still remember a lot of people saying "wait until the end of October" for us to get the full rundown on PS5 specs, since Sony wouldn't be bound by NDAs any further.
Well, October came, but it turned out not to be the case, did it? Instead we have some of the same insiders once again regurgitating old rumors in new coats of paint, with details that are just as overall murky as they were when these rumors and speculation first popped up, despite a great deal of time having passed. And once again the angle of these "new" rumors and speculation take on the form of implicating some type of strength for Sony and PS5, this time in the form of "custom solutions", that are supposedly much better than what AMD themselves were able to come up with (along with their partners) for RDNA 2, despite the fact that if these technologies were so much better....why would AMD not use them for RDNA 2? If these features are instead RDNA 3, why would AMD not find ways to pull them ahead into RDNA 2? If RDNA 3 is over a year away, doesn't that suggest they are still in early phases, so how would Sony have gotten implementation of RDNA 3 features if even AMD aren't 100% sure how those features would be implemented? If Sony indeed finalized their spec before Microsoft (the fact their devkits have been consistently more mature than MS's heavily suggests this), would that not have given AMD more than enough time to gleam what custom features Sony had they could then implement in RDNA 2?
Because here's the truth of it, all console warrior BS aside. Look at this quote from Mark Cerny from Road to PS5:
That gives us an exact time frame to work with. What PC cards would this be? RDNA 2. When is the PS5 releasing? Next week. When are these RDNA 2 cards going to start releasing? Sometime this month. Some may be December but...that would still fit the timeline Cerny gives here.
What seems to be a common trend with the RDNA 2 cards? 2 GHz+ Game and Boost clocks. What clock is the PS5 GPU at? 2.23 GHz. What was one feature AMD focused on last Wednesday? Dynamic sharing of power between CPU and GPU to push very high clocks at Boost modes. What did Cerny describe the PS5 GPU as in terms of clock? A "continuous Boost mode". What does AMD's description of power load sharing between their CPUs and GPUs sound like? Variable frequency. What has Sony called their setup? Variable frequency. What allowed Sony to break past 2 GHz GPU clocks? Switching from fixed frequency to variable frequency. What is one thing we've consistently seen between the Ariel and Oberon GPU revisions? Increasing GPU clocks. What is one thing Sony have said they designed the PS5 with in mind from the beginning? High GPU clocks.
Like, c'mon, the evidence is right in your face. It's right in RGT's face, right in MLiD's face, or whoever other's face falls into the trap of trying to push some crazy speculation regarding some super-crazy RDNA 3 features on PS5. You have the fruit of Sony and AMD's collaboration being successful staring you right in the eye, and manifest with RDNA 2 GPUs on the way, but because it doesn't fit the typical power narrative or typical "secret sauce" idea, you are not seeing the forest from the trees.
This has nothing to do with fanboyism; at least on my end, it's all about looking at the reality for what it is. Any specific features Sony wanted to highlight regarding PS5, were already a focus during Road to PS5. What guys like RGT, MLiD etc. are doing is trying to keep a certain narrative going strong by spinning up old rumors and marrying them with yet more speculation, teases, etc. that have been more recent. Yet virtually none of the stuff they are bringing up, with a week or so to go before launch, have in ANY way been confirmed officially by Sony, and this is with the additional understanding that they aren't really beholden to strict AMD NDAs anymore to even make a blanket statement of official feature support, the way MS have for example.
Chances are some of these guys aren't connecting the dots because they have tunnel vision, but that's me being generous. After all, the variable frequency stuff isn't "new" anymore, it's been known for months, so how would you drive traffic to your channel by going further into that, knowing in order to go further in, you'd have to start getting into VERY particular details that could be beyond your scope of understanding?
That's the way I'm looking this now. The clothes are slowly falling off.
This sounds like a spin of what MLiD said in their latest video, but he (as usual) went into instigating that this was MS not developing a console with "next generation game design" in mind. It's actually kind of a ridiculous notion to push on his end because there's no way to conclude that from a system being designed to run 4 1080p game instances simultaneously (plus I don't think you'd need to physically separate the L3$ on the CPU in order to enable this; you could virtualize banks of cache on the L3$ mapping each instance to a slice within it if desired, for example), and it's not like console design is an either/or game, either: you can realistically do both.
AFAIK the PS5 is only confirmed to run one game instance on its design setup; you could easily take MLiD's point and flip it against PS5; if it can't run multiple game instances on the same system simultaneously, does that mean it's not a next-gen system design? It was a really bad assertion on MLiD's part because he didn't think about it to its logical conclusion, because he's
kind of a fanboy :S
