-
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.
Digital Foundry: PlayStation 5 GPU Comparable To RTX 2060 Super Performance
- Thread starter GamerEDM
- Start date
- Status
NXGamer
Member
We do not know yet, RDNA2 itself is an unknown, this is just DF speculating I assume.
From the base specs of the PS5 more likely 2070/2070s - 2080 Level as a guesstimate.
RT is 100% unknow but would be shocked to see it perform better RTX on Turing/Ampere as they are the same on this aside more RT cores on Ampere.
Last edited:
STARSBarry
Gold Member
Oh don't worry the new SSD will make it far more powerful than any PC /s
Gen.Grievous
Member
You do know that the 2060 super is pretty similar in performance as a 2070.
UltimaKilo
Gold Member
Honestly, as a baseline for a console, this is pretty damn good. Obviously being a closed system, the performance will be better than just raw numbers.
However, this might be a bit underpowered for VR, expecially 6 years of it.
However, this might be a bit underpowered for VR, expecially 6 years of it.
"Regarding the 5700 Thing I put in the Video, that is a just a circa equivalent that I wrote in there - where you could also put in the 5700 xt if you wanted. GPUs that over of a large grouping of titles perform similar to eachother. In that case I chose the RTX 2060 Super for the Video because I confered with the Team about what GPU to use and we thought the 2070 S Was too close to the 2080 and we do not think the PS 5 GPU will be that high up there necessarily (but it is just a guess). It is more of an informed guess based on what we saw of XSX GPU and our expectations for the scaling between the the PS 5 GPU and XSX GPU using our GPU ladder.
https://www.eurogamer.net/articles/...27-gpu-power-ladder-all-graphics-cards-tested
Somewhere floating around a 5700-5700xt, 2060S, 2070. We cannot be more precise without having an unlocked PS5 game to test with (hopefully hitman 3 when that comes out as we have great contact with IO and they would tell us the exact GPU settings for that game I think). That is why in the Video I said the 2060S is likely a great comparative point for the PS5 GPU - not the same or exact same, but something where we will be using it in comparisons against the PS5 when it comes out. "
forum.beyond3d.com
There is no story here.
https://www.eurogamer.net/articles/...27-gpu-power-ladder-all-graphics-cards-tested
Somewhere floating around a 5700-5700xt, 2060S, 2070. We cannot be more precise without having an unlocked PS5 game to test with (hopefully hitman 3 when that comes out as we have great contact with IO and they would tell us the exact GPU settings for that game I think). That is why in the Video I said the 2060S is likely a great comparative point for the PS5 GPU - not the same or exact same, but something where we will be using it in comparisons against the PS5 when it comes out. "
Digital Foundry Article Technical Discussion [2020]
DF said (a typo IMO) that a 5700 was about the same as a 2070, but that's actually a 5700 XT. I think it came from there. That's in terms of normal, none RT graphics of course. In terms of RT ... well ... we'll see, but I don't think it'll be above the 2060. Yes, and that makes the PS5 gpu...
forum.beyond3d.com
There is no story here.
supernova8
Banned
In a lot of games 2060S is really indistinguishable from the original 2070.
With that said, the original 2060 isn't exactly shit either.
Stilton Disco
Member
Looking it up, the difference between the 2060S and 2070 is about 5-10 fps, and about a quater of the time that's in the 2060S's favour.
There's bugger all difference between the two basically, except the 2060S is newer and more popular, so more people will be able to get a handle on the ballpark of power DF are seeing.
There's bugger all difference between the two basically, except the 2060S is newer and more popular, so more people will be able to get a handle on the ballpark of power DF are seeing.
DonJorginho
Banned
And it begins.....
Elog
Member
Clickbait title. Alex doing the Alex thing. He knows it is a disingenuous statement since we do not know the efficiency of RDNA2. Nor does he know what degree of efficiency gains that are seen since it is a console (always more efficient than a PC) and from all the customisations (I/O, cache management and geometry engine). And finally he does not know what the RT solution looks like.
DF should know better. It is almost time for them to silently drop Alex - he drags down the brand they have built up. I have no beef with the rest of the DF crew - Alex however says these plain stupid things over and over.
DF should know better. It is almost time for them to silently drop Alex - he drags down the brand they have built up. I have no beef with the rest of the DF crew - Alex however says these plain stupid things over and over.
Clarissa
Banned
DonJorginho
Banned
You have a scar fetish I see.But Alex is hot!
New Yorks Finest
Banned
Not a PC gamer never was nor do i know whats hot out there in the market, vid cards and so on..
SO, that RTX Whatever model PS5 has blowing us all that smoke, good or bad ? High end video card?
SO, that RTX Whatever model PS5 has blowing us all that smoke, good or bad ? High end video card?
Stilton Disco
Member
Honestly, looking up the numbers, and ignoring the PS5 stuff, the 2060S is a really good GPU, pretty much a generational leap in output over the original 2060, giving basically a 2070's worth of performance for a cheaper price tag.
Makes me even more excited for the 3060, which should be up there with the 2080, especially with better ray tracing and DLSS efficiency gains, all at the standard mid range 'good enough' cost to performance sweet spot for PC gaming.
Makes me even more excited for the 3060, which should be up there with the 2080, especially with better ray tracing and DLSS efficiency gains, all at the standard mid range 'good enough' cost to performance sweet spot for PC gaming.
LeftThumbstick
Member
Why does everyone get offended when DF does any PS5 speculation? I don’t see people doing the same on DF Xbox content.
they aren’t saying this is definite either, just what they are able to tell based on the games they’ve been able to test against.
Some people in here need to get off their high horse and stop freaking out whenever the PS5 gets mentioned.
They analyse things, and this is their current analysis based on what they have seen.
Get a grip people, they aren’t saying it’s bad because they think it’s the equivalent to a 2060s... that’s a GOOD thing.
they aren’t saying this is definite either, just what they are able to tell based on the games they’ve been able to test against.
Some people in here need to get off their high horse and stop freaking out whenever the PS5 gets mentioned.
They analyse things, and this is their current analysis based on what they have seen.
Get a grip people, they aren’t saying it’s bad because they think it’s the equivalent to a 2060s... that’s a GOOD thing.
xcalibersa
Member
With Alex, everything has an agenda.
Night.Ninja
Banned
DF are shills if they did their research they would know it's closer to a 3090Ti overclocked
Elog
Member
To the people above: I agree that 2060 Super is a good card. However, why would you compare the 10.2 TFLOPs PS5 with a last-gen card from an architectural point of view - in a console that has in previous generations gained roughly 40% in efficiency from that stand-alone - that comes in at 7.2 TFLOPs? And then we have not even touched all the other optimizations and the larger VRAM pool.
It is so wrong on every single dimension. There is no excuse to make that comparison - it is 100% wrong.
I do not care if it is a PS5 or the XSX - it is just wrong.
It is so wrong on every single dimension. There is no excuse to make that comparison - it is 100% wrong.
I do not care if it is a PS5 or the XSX - it is just wrong.
Last edited:
Kenpachii
Member
5700xt sits at the 2060 super performance. PS5 GPU is of a new architecture and sits at higher clocks while at it. There is no way in hell it performs like a 2060 super its faster.
For raytrace, its a crap shoot either way. RT will probably be done on a fraction of the quality of any RTX gpu anyway. So comparisons going to be hard on this one.
For raytrace, its a crap shoot either way. RT will probably be done on a fraction of the quality of any RTX gpu anyway. So comparisons going to be hard on this one.
Last edited:
regawdless
Banned
Why are you so harsh on him? He's taken the 2060s as a reference point to show the effects of optimization regarding raytraced reflections. Assuming it's in the same ballpark as the PS5 GPU. He's not saying that it will be, just that it's his guess. And that's not actually the point of this video.
Don't understand what's so bad about this that DF should drop him.
amstradcpc
Member
we still have to see them comparing XSX to a 2070 no super or something. That's why.
Reindeer
Member
Actually Alex explained on Era that he spoke to DF team and it was s team decision to go with 2060S instead of 2070S, so let's please stop this Alex bashing. Alex also explained that this is approximation and NOT ONE TO ONE comparison (he used caps). The whole point of comparison was to get close approximation of how PS5 GPU might perform in ray tracing, nothing more.
Last edited by a moderator:
Damigos
Member
You cant compare apples to oranges. Console vs PC benchmarks are stupid.
Console can have the same result with less horsepower because its a closed unique system and is optimized to do just that. PC needs to have more power as there are endless combinations.
The only example i will tell is The order 1886 running on a 1,84 TF base PS4.
Take a 1,84 TF grqphics card on PC and laugh at the results.
Console can have the same result with less horsepower because its a closed unique system and is optimized to do just that. PC needs to have more power as there are endless combinations.
The only example i will tell is The order 1886 running on a 1,84 TF base PS4.
Take a 1,84 TF grqphics card on PC and laugh at the results.
LisaMariaMartin
Banned
SSD off = 2060 super
SSD on = 3090 TI super ultra edition
SSD on = 3090 TI super ultra edition
Bitmap Frogs
Mr. Community
What do you expect? A ps5 costs 499, and a 2060 super is around 400 bucks. Disc drive, cpu, 1tb PCI SSD, power supply, ram and motherboard for 99 bucks.
I swear to god some of you act as if Sony or Microsoft have no budgets and are free to lose 400 per console sold.
I swear to god some of you act as if Sony or Microsoft have no budgets and are free to lose 400 per console sold.
Last edited:
martino
Member
what is DFbusters specialty again ?
Build argument taking things out of context and/or adding warring one onto it....
Elog
Member
The first is that the comparison does not make any sense as a base-case for a comparison from any single parameter. Secondly, he has been doing these really disingenuous statements a few times now. I am a data-driven person and when you try to create an analog you need to base that on something. I can honestly not pick a single number from the 2060 Super as a starting point for the comparison - can you? It has 30% less power, 50% less VRAM and 15-30% lower frequency. And an architectural disadvantage. And without the console efficiency gain taken into consideration (same for PS and Xbox). And without the customizations. Etc.
On what planet does 2060 Super make sense as an analog?
Bonsaibäumchen
Member
Its fine. As far as we have seen Sony with a 2060 performance GPU on PS5 is creating miracles . Lets see how MS does on XSX with their GPU.
Also:
Also:
Last edited:
Reindeer
Member
Actually technical head of Unreal Engine engineering who knows more about PS5 than anyone on this forum said it was a very good video and gives an idea of how ray tracing will work on consoles and what sacrifices will need to be made. The only people that are console warring are tech illiterates on forums such as this one.
Last edited by a moderator:
Elog
Member
He commented on what to RT and what not to, and how to make the trade-off between RT and traditional shader work. That piece if the video was good - including the discussion with the UE engineer.
That is not the discussion however - it is the 2060 Super comparison that is plainly non-sensical.
itsnotme
Member
Apparently the Xbox is 2080-2080 super, so you're telling me a 2TFLOP difference will get me from there all the way to 2060 super level of performance? The 5700xt is already Inbetween 2070-2070 super level (albeit no RT) of performance so why is the PS5 GPU which has a higher TFLOP count and is built on a newer architecture have less performance than the already existing 5700xt
oldergamer
Member
This doesn't surprise me, nor does the fan-ish overreaction and automated defense systems kicking into action.
Last edited:
Reindeer
Member
If he thought RTX 2060S was an unfair point of approximation or that PS5 ray tracing was far superior then he wouldn't be praising that video and speaking well of it.
Last edited by a moderator:
The 5700 XT, which is the closest point of comparison available, performs almost perfectly in between a 2060 Super and 2070 Super. Unfortunately it couldn't be used for this comparison because it doesn't support hardware accelerated raytracing. So really it's a choice between the 2060 Super or 2070 Super. Neither is equivalent to the PS5 GPU, but both can be used to illustrate the performance hit of using full resolution raytraced reflections on PC, and the gains from reducing the resolution of raytraced effects.
VFXVeteran
Banned
Good post. And likewise with my own analysis that I do here on these boards. The defense force is too irrational and shows the mindset "we gotta win this war" to be repulsive. It is what it is for better or worse.
Alex was talking more about RT performance comparisons than raster performance. People need to take that into account.
Bo_Hazem
Banned
I'm happy that they claimed so. But didn't they say that BC mode on XSX is "immense"?
Sekiro 1800p on XSX:
Sekiro: Shadows Die Twice also runs at 1800p on both of these platforms but the image quality looks better on the Xbox One X with better shadows offered along with it.
gearnuke.com
.
Sekiro at a higher resolution, native 4K, most likely ultra settings, on RTX 2060 Super:
BC mode was "ok", not "immense". It's in the level of RX Vega 64. But did they say so? Nope.
Sekiro 1800p on XSX:
Sekiro: Shadows Die Twice also runs at 1800p on both of these platforms but the image quality looks better on the Xbox One X with better shadows offered along with it.
Sekiro: Shadows Die Twice Performance Is A Mess, Runs At 1800p On PS4 Pro And Xbox One X
Sekiro: Shadows Die Twice is available tomorrow and we already have our first look at the performance profile of the game, and it doesn't sound that great
gearnuke.com
Sekiro at a higher resolution, native 4K, most likely ultra settings, on RTX 2060 Super:
BC mode was "ok", not "immense". It's in the level of RX Vega 64. But did they say so? Nope.
LeftThumbstick
Member
Why don’t you do a fair comparison for us then?
The Shepard
Member
They would get a massive discount with the sheer amount of chips they'd buy at once but you are right, They have budgets to keep.
We are getting some very good hardware for the price we are paying.
Reindeer
Member
BC on Series X works without the RDNA2 enhancements as Microsoft explained, so the Series X GPU is actually performing like a GCN 12.15 GPU would. I think PS5 BC will work the same unless Sony codes their BC to take advantage of new architecture, but that's probably too much hassle.I'm happy that they claimed so. But didn't they say that BC mode on XSX is "immense"?
Sekiro 1800p on XSX:
Sekiro: Shadows Die Twice also runs at 1800p on both of these platforms but the image quality looks better on the Xbox One X with better shadows offered along with it.
.Sekiro: Shadows Die Twice Performance Is A Mess, Runs At 1800p On PS4 Pro And Xbox One X
Sekiro: Shadows Die Twice is available tomorrow and we already have our first look at the performance profile of the game, and it doesn't sound that great
Sekiro at a higher resolution, native 4K, most likely ultra settings, on RTX 2060 Super:
BC mode was "ok", not "immense". It's in the level of RX Vega 64. But did they say so? Nope.
Zathalus
Member
It is a "informed guess" as Alex put it. That still does not make it any less of a wrong one however. Assuming a 1950Mhz boost clock on the 2060 Super (which is around average and most models get you that) lands you at 8.5 TFLOPs. You cannot directly compare Turning and RDNA, but looking at average gaming performance, the 5700 XT is 9.8 TFLOPs of RDNA 1 and beats the 2060 Super by roughly 8% or so. The PS5 GPU is 10.28 TFLOPs so a straight 5% over the 5700 XT. Thus, assuming RDNA 2 has zero IPC gain over RDNA 1 (which AMD has already said it does, but that is just confirmed for the PC parts, so lets assume the same IPC), the PS5 GPU should be equal to a 2070 Super in pure rasterization. RT performance is a unknown still, so I can't make any guesses based on that.
Another point of comparison is the Gears 5 and Minecraft RT demos that were run a number of months back on XSX hardware. On a console design that was not completed with a 2 week old port, the XSX was matching a 2080 in Gears 5. Minecraft RT is a bit trickier to estimate, but it was stated to be 1080p running at high 30fps - 60fps, which once again matches 2080 / 2080 Super performance numbers.
Based on the above I expect the PS5 GPU to at least match a 2070 Super, while the XSX will match a 2080 Super.
Another point of comparison is the Gears 5 and Minecraft RT demos that were run a number of months back on XSX hardware. On a console design that was not completed with a 2 week old port, the XSX was matching a 2080 in Gears 5. Minecraft RT is a bit trickier to estimate, but it was stated to be 1080p running at high 30fps - 60fps, which once again matches 2080 / 2080 Super performance numbers.
Based on the above I expect the PS5 GPU to at least match a 2070 Super, while the XSX will match a 2080 Super.
- Status