Current Generation Hardware Speculation with a Technical Spin [post GDC 2020] [XBSX, PS5]

Status
Not open for further replies.
Meh. Talk to me in a year. These early jitters are nothing IMO.

It's click-bait gold for DF, but I think all these unoptimized ports on incomplete dev kits are telling us almost nothing right now.
 
https://www.game-debate.com/news/29...c-performance-report-graphics-card-benchmarks

2080 GPU at 1440p, high setting. The "1% low fps" of RTX 2080 is 41.4 fps. Average fps is 61.5.

ACV is 1728p~1440p on PS5 and xsx. Assuming consoles uses medium~high setting.


It does't seem xsx "underperforms" compared with 2080.

It looks like PS5 is exceptional and "outperforms" RTX 2080, which is very surprising.

Maybe there is a common denominator here between PC and XSX that makes them not perform as well as we expect them to? The tools maybe? :D
 
I don't get the impression that matt guy is an engineer (certainly not a graphics dev) so I'm not sure what significance him being a real insider has. Unlikely he's going to get to the bottom of horrible api level bugs choking framerate or the exact headaches caused by early tools via water cooler talk with his coworkers.

He is not an engineer but he works with them. They can explain to him easily. Jason Schreier said the same thing he is journalist. If someone told me the PS5 is a very good hardwarwe and the performance between PS5 and XSX will be close now and during all generation you don't need to be Einstein to understand.

EDIT: But I think it will be favorable to XSX long term with on paper spec
 
Last edited:
https://www.game-debate.com/news/29...c-performance-report-graphics-card-benchmarks

RTX 2080 at 1440p. 1% low fps is 52.5fps with medium setting and 41.4fps with high setting.
Doesn't it look like xsx version which is 46fps under heavy load?

Basically xsx has 2080 performance.

The exceptional case is PS5 which has crazily good performance.

I'm not sure how comparable this is. The PC results there are from the in game benchmark which isn't necessarily comparable to the scenes we see the XSX dropping frames in. We also need to understand the settings mix although I expect the consoles are probably running at higher than a flat medium.

On the face of it the PS5 does indeed seem to be performing crazily well in this game vs Nvidia offerings but compared with AMD PC offerings it's probably more in line with what we'd expect. See the 6800 averaging 81 fps (56 1% low) at Ultra settings 1440p in the link below. You can pick up roughly another 10-15% going to high settings or 35-40% going to medium.

XSX has worse performance then 5700XT.

In this game the 5700XT is faster than the 2800S at 1440p! Valhalla is insanely AMD friendly (like Dirt 5).

https://www.techspot.com/review/2146-amd-radeon-6800/
 
Maybe there is a common denominator here between PC and XSX that makes them not perform as well as we expect them to? The tools maybe? :D
There's clearly issues with the title.

https://www.overclock3d.net/reviews...c_performance_review_and_optimization_guide/9

For instance the 1% low frame is significantly better on the 5700 (53.9 fps) than the 1% on the 5700XT (48 fps)

Both 5700 and 5700XT outperform XSX 1440p @ high preset.

It does seem like every card is suffering from some terrible drops, 2080TI included.
 
It also looks like RTX2080 is underperforming here.

Maybe there is a common denominator here between PC and XSX that makes them not perform as well as we expect them to? The tools maybe? :D

Nvidia massively underperforms in this game vs AMD. the 5700XT and 6800 both look to be roughly where I'd expect vs the PS5. Obviously settings and resolution could make a massive difference though. We'll know much more once the DF face off is published.

https://www.techspot.com/review/2146-amd-radeon-6800/
 
I'm not sure how comparable this is. The PC results there are from the in game benchmark which isn't necessarily comparable to the scenes we see the XSX dropping frames in. We also need to understand the settings mix although I expect the consoles are probably running at higher than a flat medium.

On the face of it the PS5 does indeed seem to be performing crazily well in this game vs Nvidia offerings but compared with AMD PC offerings it's probably more in line with what we'd expect. See the 6800 averaging 81 fps (56 1% low) at Ultra settings 1440p in the link below. You can pick up roughly another 10-15% going to high settings or 35-40% going to medium.



In this game the 5700XT is faster than the 2800S at 1440p! Valhalla is insanely AMD friendly (like Dirt 5).

https://www.techspot.com/review/2146-amd-radeon-6800/

In Ultra settings, John and Alex told in the video the game is running like Xbox One X settings but at 60 fps. I suppose PS5 and XSX run at a higher resolution because of DRS but with inferior settings.

Better wait Digitalfoundry analysis for being sure, they are the only one to test PC and console at the same setup.
 
My first impression when PS5 was revealed was that it looked like Cerny really focused on removing all major bottlenecks and provide developers with a platform easy to use, probably PS5's advantage comes down to tools + architecture. If it's way easier to extract performance out of the PS5 maybe devs find it easier creating first the PS5 version and the moving to other platforms which would explain inefficiencies in the use of XBSX hardware.
You know what's funny. I was hearing this.

Then I heard from others(supposed neutrals) that the PS5 was hard or terrible to develop for and the XBSX was so easy to develop for(like DuskGolem and the BS he sold) and I was thinking, "Well that's not what I heard. I actually heard the opposite."

Colteastwood was one of the many I saw also spewing so much bullshit about the PS5, and I was hearing many a Xbox Fanboy(not same as Xbox Fan) spewing the same shit all over the internet.
 
(he means gamecore)
Yea I think it's possible that could be a representation of what's happening. But then you'd have to find equal dips across all video cards where XSX dips. An interesting check worth pursuing
 
Yea I think it's possible that could be a representation of what's happening. But then you'd have to find equal dips across all video cards where XSX dips. An interesting check worth pursuing

Or maybe isolated to only RDNA2 -- like possibly related to how it handles a few things differently? Maybe something with translation to NGC Primitive Shaders?

EDIT: Oh, I see that doesn't have any RDNA2 GPUs, so it's just general DX12 cards.
 
Or maybe isolated to only RDNA2 -- like possibly related to how it handles a few things differently? Maybe something with translation to NGC Primitive Shaders?
Yea I'm only catching up on that now. I'll need a better overview of that conversation, it's not exactly clear to me what is being implied wrt that test.
 
Meh. Talk to me in a year. These early jitters are nothing IMO.

It's click-bait gold for DF, but I think all these unoptimized ports on incomplete dev kits are telling us almost nothing right now.

I agree that over time these early days issues will be ironed out, but it's hard to disagree with the idea that having several 3P games running better on what MS's own messaging for months now has suggested was the "weaker" system, isn't really that good of a look, regardless of what SDK and API maturity issues could be going on.

Because the thing is, most people are simply going to pay attention to what these systems are offering them...right now. And that's an area I have to admit PS5 is winning in. Even in terms of PS4 BC it's doing a lot better than others were thinking it would (though maybe not really considering the Ariel and Oberon leaks kept showing regression testing for PS4 BC, so they were obviously testing BC for a while by that point). The SSD advantage is nowhere near what it was hyped up to be, but outside of that they've got a system either staying on-par or outperforming Series X in multiple 3P games (even games MS has marketing rights on like Valhalla), and 1P launch games that are more or less the best-looking/performing next-gen games currently out, that can either only be played on PS5 or play best there (Demon's Souls, Miles Morales). Not to mention the neat bonus Astro's Playroom turned out to be.

So yeah, maybe DF are kind of playing into some of the more fanboy-ish debates raging on, but I don't think they're anywhere near leaning too far into it, and the results are the results at the end of the day. This kind of performance stuff means nothing for the larger mass market, but IMO hardcore/core gamers are still the most crucial in the early years of any new console gen, they're the ones who are going to drive where interest from casuals and mainstream eventually go, and if hardcore/core gamers are choosing PS5 over Series X because of some of these early results (and the possibility this could continue for several more 3P titles going into at least early next year), that could influence the mass market to follow suit.

That's a danger for MS in my honest opinion, because we know they want the Series S to appeal to those casual/mainstream types, that's why it was developed. I guess the question is, how strong can Series S maintain its sales outside of big holidays like BF or Christmas? Is enough of the casual/mainstream market at a point where they don't follow the trend the hardcore/core gamers set beforehand, as it's traditionally been? I would like to think so, and I know the Wii might be a good candidate to point to, but the Wii actually got a ton of hardcore/core gamers on-board from Day 1 as well, and that probably helped a lot more with encouraging the mainstream to jump in than some people might think. Plus it had its unique controller; can MS position Gamepass & Series S in a way that hits the way Nintendo was able to position the Wii to mainstream/casual gamers back in its day?

That's what they need to ensure not just the consoles but Gamepass itself hits the way it should. But if the performance differences we're seeing in most 3P multiplats between Series X and PS5 persists (heck, even into a year from now), and that in turn drives more hardcore/core gamers to go with the latter...again, are there enough casual/mainstream gamers out there who don't necessarily follow where the hardcore/core go, and therefore be suitable folks to still push Series S and Gamepass to? This is a bit of a contingency scenario I'm playing out, fwiw, if the performance trend continues for longer than expected.

If so, will that create a shift to start deemphasizing Series X in marketing/production numbers? Will that shift 1P base platform focus from Series X to Series S if so? Will that create a run-off effect of hardcore/core gamers in the Xbox ecosystem drifting away if they feel their needs aren't being catered to, in these first few years? Am I asking too many questions?

At least for the last one, probably :)
 
MS is good as long as relative parity or advantage comes in ~6 months and it reflects in general media and/or Series S becomes 2nd console for alot of people. If either of two happens, they are good.

Series S is frankly very nice because (well, anecdotal evidence) some people like myself or my friends will get S for GP and MS exclusives, and PS5 as main console. They will still get people in their eco system so that is lure they saw in this strategy.

For Sony, PS5 had to be home run (and it looks like it is).
 
The SSD advantage is nowhere near what it was hyped up to be

As I posted previously, the games optimised for PS5's storage - like, Astrobot and Spider-Man Remastered and Spider-Man files Morales - are loading in a few seconds. It's apparent most other games continue to store and load their assets as they have been for years. Both Spider-Man games are dense city open worlds and you can go from the console's main menu to swinging about the city in seconds. Seconds. :???:
 
Quite a few of these Series X|S titles seems to have been built with XDK (Durango) and are "XboxOneGen9Aware" titles. Gears 5 and Bright Memory have been built with GDK (Scarlet) though.

Assuming these "XboxOneGen9Aware" games are Xbox One binaries + libraries but can have higher frame rates and target resolutions etc. if env calls are showing it's running on Xbox Series X|S?
 
This is what I mentioned the other day in the DigitalFoundry thread @ https://forum.beyond3d.com/posts/2176214/

The XboxOneGen9Aware is the XDK with minor additions to allow the game to see if they're running on Series X or Series S. It does not allow access to newer APIs or to RDNA2 functionality. The title is still operating in "Backwards Compatibility" mode. It's used mostly to allow higher resolutions or refresh rates. This is also used by Sea Of Thieves.
 
Status
Not open for further replies.
Back
Top