-
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.
DF: Unreal Engine 5 Matrix City Sample PC Analysis: The Cost of Next-Gen Rendering
- Thread starter adamsapple
- Start date
alabtrosMyster
Banned
There must be some clear limits, and sometimes you are told to say nothing or provided with something credible to say in public.
And there are the very secret projects that you are not supposed to even hint at.
alabtrosMyster
Banned
Well designed really is not too commital, so they can say it and not lie to you if it disappoints in some games.
I would not put any money on things getting better for the S. Despite the buzzword salad, I only expect it to get more services and advertising on its dashboard.
VFXVeteran
Banned
Your father was working on a classified project. My project's aren't classified. Lockheed Martin has 4 branches : Space, Aero, Missile and Fire Control and Rotary Systems. By the mere mention that I work under the Missile and Fire Control division should be pretty self-exclamatory on what I do as a software engineer.
The gaming world takes this secrecy shit too far and it's a bit annoying to me. Most of my colleagues don't even want to mention what company they work for. I miss the days when John C. and Tim Sweeney would give blogs about what they were doing in the production of a game. Those days are long gone but the work is still the same. Oh well...
MasterCornholio
Member
Well they usually have an entire division dedicated to telling people what they can say to the public. If you post something on Facebook that’s not approved by the public relations office you will get in huge trouble.
alabtrosMyster
Banned
You are right tho, I don't think anyone seriously in these circles is normally that open about it.
MasterCornholio
Member
It’s definitely something I wouldn’t brag about. Accidentally releasing something you shouldn’t can get you fired, ruin your reputation and prosecuted. It’s why I really don’t use social media and also why I don’t talk about my line of work in detail.
Md Ray
Member
What is this magical I/O?
MasterCornholio
Member
SatansReverence
Hipster Princess
Except clocks affecting performance is irrelevant because no one is claiming it doesn't.
alabtrosMyster
Banned
It depends, putting that kind of information in linkedin may bring you the benefits of honey pot type spies.
I would spill the beans!
DenchDeckard
Moderated wildly
I'm not too sure myself, I hear murmers of it on the internet and about but unfortunately due to a shit virus and cross gen games it hasn't erupted from its chrysalis yet. Maybe its a huge butterfly with Mark Cerny's head on it...Please someone photo shop this. lol
/joke
Last edited:
winjer
Gold Member
DidNXGamer provide an explanation for why his numbers differ from others? If he did I missed it.
avin
He blamed other people of being mean and fallacious.
And spouted some stuff about the engine being heavy, while ignoring that the game is heavy on all machines.
Yeah, PC's can vary a bit. But not like 60% performance difference or more, with very similar hardware.
And using a vast amount more of ram more. I just tested my system with no programs open in the background, except MSI Afterburner. Ram usage was 7.5GB, almost half of what his PC was using.
WTF was he running in the background to have such ram usage?
We even had a user, with a 2700X, underclocking his ram bellow the spec of what NXGamer was using, and still got a much higher frame rate.
Now seeing these results with this demo, it kind of makes sense why his results always seemed off.
Last edited:
Negotiator101
Member
Stop with the brown nosing for gods sake
VFXVeteran
Banned
The claim is that you are saying that the engine isn't highly multithreaded thereby basing that a reduction of cores is what's the "unoptimized" part. However, you haven't given any proof that adding more cores to do work with the Nanite/Lumen architecture would be more beneficial than increasing clock frequency and keeping the same number of cores working. That would be the meat and potatoes of this argument.
Also, you saying that if all the cores aren't at 100% all the time, that the code is unoptimized. But that claim is impossible to declare as being accurate because under the hood of any application, things are dynamic and sometimes all the cores don't need to be used. There are hotspots like iterating through a grid of data where multiple cores are absolutely paramount to getting good performance and then there are areas where using multiple cores won't do much good and clock frequency would be more sensitive to performance.
Last edited:
Negotiator101
Member
We're talking about a WIP demo, it's not up for debate it's unoptimised. Work needs to be done on UE5 across the board, I don't understand why everyone is arguing over this?
VFXVeteran
Banned
I have no doubt that it's unoptimized but we don't know what is unoptimized. To make such a detailed claim that it's unoptimized simply because it's not using all the cores when you have no idea where the code branch is that forces single threading is being an armchair programmer IMO.
We should not be so careless to make such claims without proper code profiling to find out where the bottleneck is. And that's actually possible for anyone to do if they are good enough software engineers that understand a realtime graphics pipeline as the source code is there for all to see.
winjer
Gold Member
Seems like it has some of the heaviest shaders ever in a game, or tech demo.
Almost as if it was made for offline rendering.
Unreal Engine 5 revealed
Yeah it doesn't feel right... so you dont even look at whats happening, and end up with old tech cards. Sorry! Ive been on this site for half my life....
forums.guru3d.com
Last edited:
Guilty_AI
Member
Last edited:
Negotiator101
Member
Who's we? I made no claim, the poster is referring to general performance of this demo. Now he has his own theories, unless you want to do a deep dive into the code your guess is as good as his right?
StateofMajora
Banned
I can appreciate that zen 3 fixed ryzen's ccx issue, and sometimes there were huge differences.
But the first 3 Ryzen were horrible products?
Yeah man, they were so horrible that it forced Intel to add 2 extra cores each generation to their flagship ; 8700k, 9900k, 10900k.
So horrible that production users abandoned Intel, so horrible that even with the lower price gamers shifted from Intel to amd...
So horrible that, amd stock price went up 60x.Bulldozer, that was a horrible product. Ryzen saved us from Intel monopoly.
And don't you know when Amd beat Intel to the ghz point, gave us 64 bit architecture and was ahead of Intel with its athlon 64 architecture?
Heck, even the phenom architecture was good value for money. Phenom 2 was no Intel i7, but it was still a good value option at the time. I had a phenom 2 x4 925, then bumped it to a 1090t x6 later.
------
Anyways, for the record if money was no object I would have chose an 9900k over zen 2 or zen+ obviously. But my 6 core zen+ has served me very well and for $85 new it probably is the best value cpu ever. Now I can't wait for my 5800x3D to arrive
Last edited:
ethomaz
Banned
No.I guess that means
He should. Perhaps he yet well.
avin
If you look others results it is very close to NX… even Alex one is close to NX data.
People are struggling to reach 30fps with that demo even with the best hardware on the market.
My e-pennis is 5fps bigger than NX and DF are just laughable… even more when the bigger e-pennis is still getting most of time sub 30fps… and yes it goes low as 15fps like several videos already showed.
I should not believe anybody that says the demo runs over 30fps most of time… that is misleading the results.
Last edited:
Dust-by-Monday
Member
30fps is the future! I've been saying it from the beginning of the generation. Better hardware means more demanding software.
Last edited:
Midn1ght
Member
winjer
Gold Member
I guess that means
He should. Perhaps he yet well.
avin
If he was a bit humble, he would try to figure out the issue. Be it the demo he packaged or his PC.
He could talk to everyone here, compare results. Try different settings. Try different packaged demos, etc.
Alex was reaching 40-50 fps. Not even close, to what NXGamer was reaching with 14-23 fps.
GameGPU got 43 fps on average on the same 2700X.
I got 30-40 fps. And so did other users here.
ethomaz
Banned
Lies.
“averaging 30fps in static scenes but falling much lower when in movement.”
C’mon you guys are better than that.
Alex and NX results are very close to each other.
Imagine reaching 40-50fps like you said and claiming “we're still far from a locked 30fps”.
You basically is misleading Alex tests.
That is exactly why I made the post… to clear up the narrative.
Last edited:
ethomaz
Banned
Last edited:
winjer
Gold Member
A lot of people here have already proved that NXGamer's results are way lower than anyone else. By a huge margin
The numbers don't lie. So stop making up excuses for him.
ethomaz
Banned
So why did you lie about Alex results?
Midn1ght
Member
ethomaz
Banned
I’m fine… thanks.Are you okay?
winjer
Gold Member
Wtf?
yamaci17
Member
i will do a one to one comparison with nxgamer's video tonight in the exact place he drives his car when he gets 20 fps. i already downloaded his entire video. situation will get ugly but i have to do this, otherwise there's no valid way to get them to understand the difference
if he ends up deny this video, there's nothing that can convince him otherwise and it will my final attempt at doing so
a small demo of what is to come. (3.7 ghz + 3000 cl15 kits)
Last edited:
ethomaz
Banned
winjer
Gold Member
Really? Where?
ethomaz
Banned
So you write thing and forget them? Lol
“Alex was reaching 40-50 fps.”
Bullshit.
Edit - Before you try to move goal post again… the Alex results are documented:
Unreal Engine 5 hands-on: the cost of next generation rendering
Unreal Engine 5 recently emerged from early access, with a full version now available to games creators. Simultaneously…
www.eurogamer.net
And yes it is very similar to NX results… shocking.
Most of time below 30fps with drops way below the target.
PS - I never thought I will have to defend Alex lol he is biased but I never saw him lying about the data… but hey let’s cancel NX data but not Alex when they showed similar results.
Last edited:
PaintTinJr
Member
Judging by the comparison images at a glance, and my own results with someone else's precompiled version on my setup and getting gains of about 5-10fps from lowering the GI level with the result of scenes being much brighter in the distance - with less GI - I don't think your example is using the same level of Lumen GI as nxgamer's. His specular highlights look smaller for a start going by that big one in your image in the road's shadowed section - which might even be light bleeding.i will do a one to one comparison with nxgamer's video tonight in the exact place he drives his car when he gets 20 fps. i already downloaded his entire video. situation will get ugly but i have to do this, otherwise there's no valid way to get them to understand the difference
if he ends up deny this video, there's nothing that can convince him otherwise and it will my final attempt at doing so
a small demo of what is to come. (3.7 ghz + 3000 cl15 kits)
yamaci17
Member
ethomaz
Banned
Are we allowed to compare IQ too?i will do a one to one comparison with nxgamer's video tonight in the exact place he drives his car when he gets 20 fps. i already downloaded his entire video. situation will get ugly but i have to do this, otherwise there's no valid way to get them to understand the difference
if he ends up deny this video, there's nothing that can convince him otherwise and it will my final attempt at doing so
a small demo of what is to come. (3.7 ghz + 3000 cl15 kits)
winjer
Gold Member
So you write thing and forget them? Lol
“Alex was reaching 40-50 fps.”
Bullshit.
Edit - Before you try to move goal post again… the Alex results are documented:
Unreal Engine 5 hands-on: the cost of next generation rendering
Unreal Engine 5 recently emerged from early access, with a full version now available to games creators. Simultaneously…
And yes it is very similar to NX results… shocking.
Most of time below 30fps with drops way below the target.
OK, so you are talking about that article. I was talking about the video, where he spent almost the whole time benching the 10900K, where he got 40-50 fps. This is not a lie.
Just because you were talking about one CPU he barely touched in the video, and I was talking about the CPU he used in the whole video, doesn't mean I was lying.
But Alex gets much better results than NXGamer, be it driving or static.
But if you want just the 3600 that he talks in that article, then it's an average of 30 fps in static scenes and around 25 while driving.
Here is the proof.
And we have video of him driving around. As you can see it's in the mid 20s.
Now compare to NXgamers results.
In static scenes h gets 20-23 fps.
And in movement, he goes to the mid teens. This is very far from being the same result as Alex.
He even stated that he got as low as 14 fps in his 2700X.
Last edited:
ethomaz
Banned
Man, really?OK, so you are talking about that article. I was talking about the video, where he spent almost the whole time benching the 10900K, where he got 40-50 fps. This is not a lie.
Just because you were talking about one CPU he barely touched in the video, and I was talking about the CPU he used in the whole video, doesn't mean I was lying.
But Alex gets much better results than NXGamer, be it driving or static.
But if you want just the 3600 that he talks in that article, then it's an average of 30 fps in static scenes and around 25 while driving.
Here is the proof.
And we have video of him driving around. As you can see it's in the mid 20s.
Now compare to NXgamers results.
In static scenes h gets 20-23 fps.
And in movement, he goes to the mid teens. This is very far from being the same result as Alex.
Are you listening the video?
He was clear about running in 720p.
“A Core i9 10900K at 5.0GHz locked paired with an RTX 3090 can see a performance average of around 44fps in a fairly taxing scene - and that's at native 720p rendering, attempting to remove the GPU as a limiting factor to performance.”
He got sub30fps average at 1080p.
You did not know vem read the top part of the screen lol
Last edited:
yamaci17
Member
GI2 vs GI3
Imgsli
imgsli.com
1) GI3 works and makes a difference over GI2
2) I get similar 38-39 FPS on both
if you say my GI3 is not matching his GI3 (because he himself says console uses GI3 and uses GI3 on his video), then there's nothing i can say anymore. it seems like we've gone from facts to %800 pixel peeping to make arguments ouf of this discussion?
winjer
Gold Member
it's a CPU bound scenario. None of them are reaching 100% GPU utilization.
Look at Alex's results. His average while driving and stressing the system is higher than NXGamer's average while standing.
At this point it's you who is lying, basically comparing Alex's lowest frame rate, with NXgamers highest frame rate, and saying they are the same.
And then you have the nerve to insult people who clearly show you the evidence.
You are truly dishonest,.
ethomaz
Banned
It is no what he said.
But keep the 40-50fps bullshit… people here already know you are not being honest.
Imagine trying to pass much lower than 30fps as 26fps lol
Last edited:
Shmunter
Member
Backtracking already, lol. Now look at driving framerates deep into traffic, never mind collisions.OK, so you are talking about that article. I was talking about the video, where he spent almost the whole time benching the 10900K, where he got 40-50 fps. This is not a lie.
Just because you were talking about one CPU he barely touched in the video, and I was talking about the CPU he used in the whole video, doesn't mean I was lying.
But Alex gets much better results than NXGamer, be it driving or static.
But if you want just the 3600 that he talks in that article, then it's an average of 30 fps in static scenes and around 25 while driving.
Here is the proof.
And we have video of him driving around. As you can see it's in the mid 20s.
Now compare to NXgamers results.
In static scenes h gets 20-23 fps.
And in movement, he goes to the mid teens. This is very far from being the same result as Alex.
He even stated that he got as low as 14 fps in his 2700X.
So many hurt defenders out skewing reality together, repeating each other’s misinformation like a circle jerk of parrots is amusing, trying to take out NXGmers for simply representing findings, that’s how derranged this all is.
This sort of pile-on behaviour is something I would support bans on.
Last edited:
winjer
Gold Member
You keep insisting that I lied, just because we were talking about different CPUs that Alex tested.
And you keep ignoring all the evidence, people have showed you, from other testers. And even from users in this forum.
I already showed you exactly where the 40-50 fps came from. And I already showed you the real numbers from Alex's 3600x.
What backtracking.
Compare results in the same scene, and NXGamers results are always significantly lower.
Last edited:
ethomaz
Banned
It is very weird people trying to cancel NX while accept Alex with the same results.
Hypocrisy.
And pictures with different IQ because the settings doesn’t work equal to everybody unless you compile it yourself… Alex explained that in B3D how it was difficult to make the file settings proper work.
So we need to start to check IQ from these supposed 50fps screens.
Hypocrisy.
And pictures with different IQ because the settings doesn’t work equal to everybody unless you compile it yourself… Alex explained that in B3D how it was difficult to make the file settings proper work.
So we need to start to check IQ from these supposed 50fps screens.
Last edited: