• 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.

Performance Test(GigaPixels/s): PS5 outperforms RTX 2080 by a wide margin under heavy load

Yoboman

Yoboman

Member
DynamiteCop! said:
I can't even begin to unfold the complex levels of delusion where people find value in this kind of synthetic nonsense. Furthermore beyond people trying to make the PS5 seem to be more than it is, this man is not only a previous developer at Naughty Dog, he's plugging his own test software...

Get out of here... Jesus..
Click to expand...
This you?


2erZ5J5.gif
 
Bo_Hazem

Bo_Hazem

Banned
Rikkori said:
That's just RDNA2 wiping the floor with Ampere in Valhalla, but it is not representative of other games' performance.

1440p-ACV.png


XL27Ste.gif
Click to expand...

Damn, and that's while the big brother is still not in the party, 6900 XT. But raytracing performance matters so much now though.
 
LostDonkey

LostDonkey

Member
Eric Lengyel is credited on the following games:



iazb0uL.gif
 
Bo_Hazem

Bo_Hazem

Banned
Agnostic2020 said:
Yea thats where amd shits the bed 😅
Click to expand...

Maybe add his Bio?

PhD/CS. Gamedev & math author. Creator of @SlugLibrary, @C4Engine, @The31stGame, http://opengex.org. Former Naughty Dog, Apple, Sierra.

Eric Lengyel
From Wikipedia:

Eric Lengyel is a computer scientist specializing in game engine development, computer graphics, and geometric algebra. He holds a Ph.D. in Computer Science from the University of California, Davis and a master's degree in Mathematics from Virginia Tech.

Lengyel is an expert in font rendering technology for 3D applications and is the inventor of the Slug font rendering algorithm, which allows glyphs to be rendered directly from outline data on the GPU with full resolution independence.[1] Lengyel is also the inventor of the Transvoxel algorithm, which is used to seamlessly join multiresolution voxel data at boundaries between different levels of detail that have been triangulated with the Marching cubes algorithm.[2]

Among his many written contributions to the field of game development, Lengyel is the author of the four-volume book series Foundations of Game Engine Development. The first volume,[3] covering the mathematics of game engines, was published in 2016 and is now known for its unique treatment of Grassmann algebra. The second volume,[4] covering a wide range of rendering topics, was published in 2019. Lengyel is also the author of the textbook Mathematics for 3D Game Programming and Computer Graphics[5] and the editor for the three-volume Game Engine Gems book series.[6][7][8]

Lengyel founded Terathon Software in 2000 and is currently President and Chief Technology Officer at the company, where he leads development of the C4 Engine. He has previously worked in the advanced technology group at Naughty Dog, and before that was the lead programmer for the fifth installment of Sierra's popular RPG adventure series Quest for Glory. In addition to the C4 Engine, Lengyel is the creator of the Open Data Description Language (OpenDDL) and the Open Game Engine Exchange (OpenGEX) file format.[9]
Lengyel is originally from Reynoldsburg, Ohio, but now lives in Lincoln, California. He is a cousin of current Ohioan and "Evolution of Dance" creator Judson Laipply.

Games
Eric Lengyel is credited on the following games:
 
Last edited:
C

CuteFaceJay

Member
what does he need a 3080 for just to test when people have been doing non stop tests with them for months?

The data is already out there. He's just beating around the bush for no good reason.
 
Stilton Disco

Stilton Disco

Member
Black_Stride said:
Indeed not taking anything away from the PS5 its just this specific test isnt exactly something we need to be praising the sun for if you catch my drift.

What leads to "capable" performance isnt down simply its pixel/s rate.
Theres a ton more going on so this test alone says very little to even get an idea of what the GPU could do for actually rendering a game.

Im sure Unigine and Unreal engineers have more "solid" examples of what the PS5 GPU is actually capable of.
So its alittle hilarious to see a bunch of people praising the PS5 because its font rendering is better than an RTX 2080.

Effectively if you looked at this test alone the PS5 can draw a fontbased UI faster than an RTX 2080......is that really something we need to be up in arms about?
Click to expand...
The fact the PS5 is able to do any kind of graphical rendering at 4K better than a bloody 2080 is damned impressive.

It does however give an idea, when comparing console apples to PC oranges as to where the PS5's strengths lie, with the most likely real world benefit being with texture streaming and fillrate, two things we're already seeing evidence of with the unexpected results in comparisons between PS5 and XSX.

Dismissing these kind of benchmark tests out of hand is dumb and shortsighted. This relative speed and ability is being used in PS5 game development, even if it's not for this purpose, and seeing what a really good PC component is capable of in comparison adds real world context to that, even if we don't know what it's actually being used for.
 
Last edited:
Black_Stride

Black_Stride

do not tempt fate do not contrain Wonder Woman's thighs do not do not
Stilton Disco said:
The fact the PS5 is able to do any kind of graphical rendering at 4K better than a bloody 2080 is damned impressive.

It does however give an idea, when comparing console apples to PC oranges as to where the PS5's strengths lie, with the most likely real world benefit being with texture streaming and fillrate, two things we're already seeing evidence of with the unexpected results in comparisons between PS5 and XSX.

Dismissing these kind of benchmark tests out of hand is dumb and shortsighted. This relative speed and ability is being used in PS5 game development, even if it's not for this purpose, and seeing what a really good PC component is capable of in comparison adds real world context to that, even if we don't know what it's actually being used for.
Click to expand...

I could totally dismiss this benchmark.
It doesnt show anything we could reasonably extrapolate game rendering performance from.
So it isnt exactly useful when we are talking about a videogame console.

The PS4pro has a higher (absurd) Gigapixel/s rate than the Xbox One X but did that translate into anything us gamer could sit and say "see, that gpx/s rate is finally showing how much of a beast the PS4Pro is versus the XB1X"? it didnt.
Because this test doesnt really translate into anything gamers would be able to even remotely gauge performance from.
 
S

spyshagg

Should not be allowed to breed
LostDonkey said:
Eric Lengyel is credited on the following games:



iazb0uL.gif
Click to expand...

Instantly looking for any kind of detail to spin the news around. You guys should work for politicians. Scum of the earth.

What I see are 3 of the most amazing games ever made on his portfolio. IMO.
 
Last edited:
Coulomb_Barrier

Coulomb_Barrier

Member
wipeout364 said:
Is this true? Seems questionable to me do you have a link and the PC setup?
Click to expand...
Bernd Lauert said:
Imagine actually believing this lmao
Click to expand...

Oh shut up, it's true, do the maths. Or tell me, what graphics card runs Valhalla at the same fps and res as the PS5/XsX? Go on I fancy a good laugh this morning.

It's 3070 level, in fact PS5's performance was capped at 60 fps in that mode, it would run higher. Digital Foundry and NXGamer also concluded the same but I'll only dig around for the quotes if you change your whole attitude.
 
wordslaughter

wordslaughter

Banned
Coulomb_Barrier said:
Yep, the PS5 is running Ass Creed Valhalla equivalent to a 3070 (2080 Ti) at minimum. So much for Ampere graphics cards blowing the consoles out of the water, like some claimed.
Click to expand...

This is just not true. A better comparison would be to the 1080ti, at best.

The PS5 has a performance mode that does a good job at sticking to 60fps, but the resolution has to drop quite substantially to stay there and it's not on Ultra settings. It's hard to directly compare to the PC version in this mode.

But the quality mode makes it easy to directly compare. We know the resolution is locked to 4k and the framerate to 30fps ... but are the settings all Ultra? And do we even know for sure if the PS5 is NAVITELY rendering at 4K or are they just using temporal reconstruction to get there?

Anyway, a 1080ti can do all Ultra 4K (native 4K) and lock it to 30fps. I know this because I own one and Valhalla. A 2080ti can do quite a bit better.
 
Last edited:
Stilton Disco

Stilton Disco

Member
Black_Stride said:
I could totally dismiss this benchmark.
It doesnt show anything we could reasonably extrapolate game rendering performance from.
So it isnt exactly useful when we are talking about a videogame console.

The PS4pro has a higher (absurd) Gigapixel/s rate than the Xbox One X but did that translate into anything us gamer could sit and say "see, that gpx/s rate is finally showing how much of a beast the PS4Pro is versus the XB1X"? it didnt.
Because this test doesnt really translate into anything gamers would be able to even remotely gauge performance from.
Click to expand...
Well of course we didnt. No games were made for the XOX or PS4Pro.

Those were all base XO and PS4 games running on them fundamentally, just given extra power for slightly better frame rates and higher resolutions, with maybe a few more effects and details. We never saw what either machine was truly capable of were they the target specs for development.

Conversly, this is the base PS5 specs developers are already using we're looking at here, where everything the machine is capable of will be leveraged by at the very least first party Developers.

And as I and other people keep saying, this is indicative of the PS5's abilities when it at least comes to pixel fillrate and texture streaming, two things that we know it's really bloody good at that corroborate these benchmarks being useful.

But hey, if you want to ignore context, logical deduction and existing game data analysis, and just dismiss this out of hand, you do you
 
ZywyPL

ZywyPL

Banned
Interesting, although doesn't mean much in real-life scenarios where 2080 has no issues with pushing 60+ FPS while PS5 has to settle for 30, or cut down the resolution in half to push more.
 
longdi

longdi

Banned
Bo_Hazem said:
Games
Eric Lengyel is credited on the following games:
Click to expand...

seems like a pattern? :messenger_open_mouth:

where can i download the slug5.0 benchmark? like to try it out.
 
ethomaz

ethomaz

Banned
Black_Stride said:
The 3080 is about 2.5x faster than the RTX2080.
So you could approximate that it would score around 6.5 in the absurd test and absolutely level everything in the ordinary test.
The XSX should score similar to the PS5 maybe slightly higher.

Microsoft Word is going to be so smooth on my PS5!
Cant wait!


This is what the test looks like for those curious, the viability of this information for actual gaming applications isnt particularly high versus other benchmarks.
EoVgJx7UUAAKDBT
Click to expand...
Probably not.
Series X fillrate is 22% lower than PS5 so in theory it should score less.

RTX 3080 probably do 1.5x better than RTX 2080... not 2.5x.
 
Last edited:
FearlessBelgian

FearlessBelgian

Member
Rikkori said:
jgsfe06.jpg
Click to expand...

Wow, they made a meme of that so often it happens that Ubisoft games are better optimized on AMD cards because of their partnership ... Strangely Watch Dogs legion runs better on a 3080 than a 6800 XT ... As if by chance, the only recent game from Ubisoft not sponsored by AMD ... And don't show me a graphic in 1080p, in 1440p / 4K the 3080 is ahead of the 6800 XT on Legion ... :)
 
RobRSG

RobRSG

Member
What I see is 3080 in native 4K breezing, aside from Asscreed, through all the lastest games, and PS5 relying heavily in DSR to get 60 done.

But the real competition is XSX, and the secret sauce keeps beating it though.
 
LostDonkey

LostDonkey

Member
spyshagg said:
Instantly looking for any kind of detail to spin the news around. You guys should work for politicians. Scum of the earth.

What I see are 3 of the most amazing games ever made on his portfolio. IMO.
Click to expand...
Right...see the whole point of me reading GAF is so I can take whatever someone says on twitter as gospel without questioning it. Without looking at the source and seeing if they have an ulterior motive or an interest in pushing a certain agenda due to company ties etc.....

That will get you really far.

Not to mention this is his own program he is pushing, that he invented and wrote.

Yeah, i'm a little sceptical.
 
Bernd Lauert

Bernd Lauert

Banned
Agnostic2020 said:
Lol u don't understand how funny u sound. U basically wrote a meme without trying 🤣


I m afraid of false hope wishing for better future instead of enjoying that right now. But keep waiting bud.
Click to expand...

I don't have to wait for anything, thanks to VRR games already play best on the World's Most Powerful Console. The clock is ticking for you and you seem worried about it.
 
Kholinar

Kholinar

Banned
Bernd Lauert said:
I don't have to wait for anything, thanks to VRR games already play best on the World's Most Powerful Console. The clock is ticking for you and you seem worried about it.
Click to expand...

Wow, the world's most powerful console needs a thousand dollar TV just to deliver on it's promises. Naw, I'll take the one with more multiplat performance.

PS5. It eats monsters that eat monsters for breakfast.
 
S

Superayate

Member
ethomaz said:
Probably not.
Series X fillrate is 22% lower than PS5 so in theory it should score less.

RTX 3080 probably do 1.5x better than RTX 2080... not 2.5x.
Click to expand...

It's possible but it does not depend only of the Fill rate in fact in this test, just need to see the results between PS4 pro Vs One X.
 
Last edited:
Bernd Lauert

Bernd Lauert

Banned
Coulomb_Barrier said:
Oh shut up, it's true, do the maths. Or tell me, what graphics card runs Valhalla at the same fps and res as the PS5/XsX? Go on I fancy a good laugh this morning.

It's 3070 level, in fact PS5's performance was capped at 60 fps in that mode, it would run higher. Digital Foundry and NXGamer also concluded the same but I'll only dig around for the quotes if you change your whole attitude.
Click to expand...

Remind me again, what are the detail settings on the PS5? Hint: it's not ultra like on every PC benchmark.
 
You must log in or register to reply here.

Similar threads

Xbox Dev Demonstrates NVIDIA GeForce RTX 2080 Ti & Xbox Series X Mesh Shaders Performance With DirectX 12 Ultimate API
3 4 5 6 7
310
36K
Three Jackdaws replied
Unreal Engine 5 revealed! Real-Time Prototype Gameplay Demo Running On PS5
30 31 32 33 34
2K
134K
Mr Wonderstuff replied
PS5 – Every Official News, Leak & Rumor on Sony’s Next-Gen Console
5
2K
SoloKingRobert replied
Xbox Series X CPU Performance May Be on Par With Ryzen 5 1600/1600 AF, Analysis Suggests
2
56
8K
JareBear: Remastered replied
C4 Engine 4.0 now has new subscription model
11
4K
blu replied
Top Bottom