Current Generation Games Analysis Technical Discussion [2024] [XBSX|S, PS5, PC]

[Boss]

Read a recent article about rage-baiting. Deliberately creating content designed to rile viewers into posting rage, because that gets engagement which pushes adverts and make money. So now you have people creating personas just to be obnoxious and generate lucrative hate clicks. This almost certainly has feedback loops that shift the gestalt.

You can make an argument that the content is rage bait but the end net result is positive. The issue is being dragged into the limelight so that actual and proper discussion can be had regarding performance. If you read the comments on the video, there are people commenting that they didn't know how to verbalize the issues they were noticing but, they appreciate that someone is shining a light on this issue. Now are their methods the best? Nope. However imo, the end justifies the means. In an era of shrinking gains from Gpu generation to gpu generation and rapidly rising gpu costs, resources have to be used efficiently. With the 5090 rumored to blast 600watts of power just to play games, it would be a shame if we get more UE5 blunders like remnant 2 at 45fps at 4k on the 4090.

 

[IQandHDR]

UE5 can be hit or miss as I have seen it in games.
The litmus test for me will be "The Witcher 4" against RED Engine ("The Witcher 3"/"Cyperpunk 2077").
It had better look better AND perform better or I will consider it (UE5) an failure
(Any shader stutter and I will be very vocal)

 

[arandomguy]

Yeah we have a few examples.

The CineBench score of the 14900K is 1795, the GPU score of the 4080S is 27790! The 4080S is 15x faster than 14900K.
In Blender, the 14900K scores 442, the 4080S scores 9012, the 4080S is 20x faster than 14900K.

Puget Systems Laptop vs Desktop Performance Comparison for Content Creation

With the launch of the new Puget Systems 17" laptop, there are a lot of performance questions we want to address. Today, we will be examining the performance of this new mobile workstation compared to a more traditional desktop.

www.pugetsystems.com

In another set of examples, we can compare accelerated ray tracing on vs off.

Spoiler: RT on/off

In Octane, the 4090 is 2.2x faster with it on.
In Blender, the 3090 is 3x faster than 6900XT due to better ray tracing acceleration.

NVIDIA GeForce RTX 4080 GPU Rendering Performance Review

The second GPU to grace NVIDIA's Ada Lovelace GeForce lineup has landed: RTX 4080. Like the previously-released RTX 4090, the RTX 4080 continues to show us that the Ada Lovelace generation offers huge upticks in rendering performance over the previous-gen Ampere.

techgage.com

I'm going to throw something in here about these rendering benchmarks in that if they are (and I believe most of them are) testing the speed per sample or for a fixed amount of samples something to keep in mind is that the end result of a single sample is not identical unless the rendering path is indentical. But the kicker here is the rendering path is not identical if the underlying framework isn't. This means 1 sample on the CPU vs GPU isn't the same end result, even 1 sample on say CUDA vs Optix isn't the same result.

That's a synthetic benchmark but as we've just learned, that doesn't necessarily translate to real-world gains. I've seen GPU acceleration in Blender myself, which is why I imagined render times for movies had come right down, so it's a surprise it hasn't for reasons I hadn't appreciated, particularly the RAM one. But that's a certain class of rendering. Below that there'll be animated TV series and advertisements and student films and indie shorts that aren't 500 GB a frame. These are the real-world examples we can look to to see what's being rendered and how and in what timeframes on GPUs, if there's any info on them out there.

So I'm really hoping we can find a Netflix series or animated short somewhere with render times. Maybe student projects? ¯\_(ツ)_/¯

I believe would be to hard to generalize due the varying scope of the projects.

Another thing is with movies, and offline renders in general, is that while you can just prep everything and have it spit out a single end frame in practice (especially for the higher end productions) this isn't how it works. From what I've read it is that certain frames can take hundreds of hours in the all encompassing sense on the highest end productions for a single end frame. But this would include things such as simulation and really multiple layers that are composited together.

 

[Remij]

UE5 can be hit or miss as I have seen it in games.
The litmus test for me will be "The Witcher 4" against RED Engine ("The Witcher 3"/"Cyperpunk 2077").
It had better look better AND perform better or I will consider it (UE5) an failure
(Any shader stutter and I will be very vocal)

Judging from their tech talks, regardless of whether The Witcher 4 has some stutters here or there, we can rest assured that they're taking it very serious and are well aware of it. It's literally one of the first things they mention, and go on to explain how they're planning to mitigate the issue, regarding traversal stuttering specifically. Shader compilation I'm much less worried about to be honest. They'll do a good job of it I'm sure. I'm also fairly certain it will blow the doors off of what came in TW3 and CP2077. It will be a generational jump for sure. Have some faith

Watch it if you haven't

 

[see colon]

UE5 can be hit or miss as I have seen it in games.
The litmus test for me will be "The Witcher 4" against RED Engine ("The Witcher 3"/"Cyperpunk 2077").
It had better look better AND perform better or I will consider it (UE5) an failure
(Any shader stutter and I will be very vocal)

Considering the state that Witcher 3 and CP2077 launched in, and their performance at the time, I don't know how I would even be able to compare the new releases. Witcher 3 had a bunch of bugs, and a bunch of stuttering at launch. Especially on non-nVidia GPUs. And it had it's highest graphics options locked behind Gameworks nonsense that further drove down performance. Was there shader stutter? Not sure I could have even notice.

 

[Remij]

It really sucks that Steam went ahead and allows devs to make branches private now, effectively killing our ability to know if a game is being patched or worked on through SteamDB.

Also sounds like Ubisoft is pressuring Valve to either remove concurrent player info from being accessible by SteamDB, or at least being able to hide it.

If that happens, SteamDB basically becomes useless. I hope that Valve doesn't consider doing the latter.. because for as many developers that would complain about it (because their games flop) it also undoubtedly helps some games become even bigger hits.. such as Helldivers 2 and Palworld.. not to mention Black Myth. All of the positive word of mouth generated by strong CCUs feeds back into itself, and the smart publishers know how to capitalize off of it.

 

[Charlietus]

It really sucks that Steam went ahead and allows devs to make branches private now, effectively killing our ability to know if a game is being patched or worked on through SteamDB.

Also sounds like Ubisoft is pressuring Valve to either remove concurrent player info from being accessible by SteamDB, or at least being able to hide it.

If that happens, SteamDB basically becomes useless. I hope that Valve doesn't consider doing the latter.. because for as many developers that would complain about it (because their games flop) it also undoubtedly helps some games become even bigger hits.. such as Helldivers 2 and Palworld.. not to mention Black Myth. All of the positive word of mouth generated by strong CCUs feeds back into itself, and the smart publishers know how to capitalize off of it.

Honestly, removing some bullets from the guns of ragebait YouTubers can only be a good thing. If a game flops on steam it has a negative effect on all the other platforms. And why do we need this data anyway? Maybe for multiplayer games but for single player games it should be a choice of the publisher.

 

[IQandHDR]

Honestly, removing some bullets from the guns of ragebait YouTubers can only be a good thing. If a game flops on steam it has a negative effect on all the other platforms. And why do we need this data anyway? Maybe for multiplayer games but for single player games it should be a choice of the publisher.

Removing data because you have a single issue with the use of the data sounds very single minded.

 

[Charlietus]

Removing data because you have a single issue with the use of the data sounds very single minded.

The data objectively gets used maliciously, so yes, I take issue with it.

 

[Shifty Geezer]

The data objectively gets used maliciously, so yes, I take issue with it.

It also gets used non-maliciously. We get fewer and fewer insights these days. The "sales thread" used to be full of monthly discussion and trend watching. Now it's kinda empty with companies choosing to invent new metrics like MAUs.

Data is just data. If we want to hide info because people won't use it properly, we should just ban the internet.

 

[IQandHDR]

For me, what eg. Ubisoft is doing is trying to hide their blunders, because investors have taken notice of their failings.
It only serves those failed policies (and those who are responsible for those bad decisions) to hide the data, not the general public.

Censorship is inherently bad, no other way around it.

What is next on the ban list?
GPU sales?
Windows sales?
Console sales?
Stock prices?
Currency value?

Slippery slope is slippery slope.

 

[Charlietus]

It also gets used non-maliciously. We get fewer and fewer insights these days. The "sales thread" used to be full of monthly discussion and trend watching. Now it's kinda empty with companies choosing to invent new metrics like MAUs.

Data is just data. If we want to hide info because people won't use it properly, we should just ban the internet.

But why not give the choice to the creators? During the days of NPD and clear data, there weren't YouTubers, tik tokers and others making a literal career out of spreading negativity and misery.

The times have changed for the worst.

 

[Shifty Geezer]

But why not give the choice to the creators?

If you allow people to be selective, the data gets polluted. You can only have all or nothing if the data is going to be complete and meaningful. So, either you decide data is good, and we have it, all of it, and make sense of it, or you decide data isn't at which point you have some of it, none of it, an incomplete picture and possibly a misleading picture.

Let's imagine gaming is in a dive. Developers and publishers are seeing numbers dropping across the board. There's a good argument to be made to improve tax breaks and kick-start things, or something. However, all the devs are hiding their data because they feel it's a negative, would cause internet raging or whatever. Except, there are a few big games going from strength to strength. They show their numbers because it's good PR and they are parading how big their games are. A researcher looking for information on the state of the games industry is only going to see the wins broadcast.

There's a fundamental philosophy one has to subscribe to if one is to use data properly - "The truth will set you free." If you don't believe that and feel information has to be controlled, manipulated, obfuscated, for the best outcome, then you don't want data at all and just want selective messaging.

The times have changed for the worst.

And going along with that will only make it worse. If things are ever to improve, we need examples of correct data, correct analysis, correct reactions. If people never see that and only see polarised opinions and misinformation, society will just reinforce that trajectory.

 

[IQandHDR]

But why not give the choice to the creators? During the days of NPD and clear data, there weren't YouTubers, tik tokers and others making a literal career out of spreading negativity and misery.

The times have changed for the worst.

Then use the data to show that they are wrong?
It is the same data, right?
Only catch is that if you cannot do that, your interpation of the data might be flawed

 

[homerdog]

But why not give the choice to the creators? During the days of NPD and clear data, there weren't YouTubers, tik tokers and others making a literal career out of spreading negativity and misery.

The times have changed for the worst.

If there are bad actors you should focus on them. Leave the data out of it.

Ubisoft is big mad since this data makes it harder to mislead investors and creditors

 

[Charlietus]

To respond to all of you at once, you bring up some good points. It's just that from my point of view, we would lose data that's valuable and transparent, but it would throw a wrench in the future arguments of bad actors. So there are positive and negatives.

But in my view, the positive of accessing the data gets completely obscured by the negativity it causes. And unless we get in to social media censorship there is no way to combat this rampant negativity.

 

[techuse]

To respond to all of you at once, you bring up some good points. It's just that from my point of view, we would lose data that's valuable and transparent, but it would throw a wrench in the future arguments of bad actors. So there are positive and negatives.

But in my view, the positive of accessing the data gets completely obscured by the negativity it causes. And unless we get in to social media censorship there is no way to combat this rampant negativity.

Could you provide an example of where this negativity caused undue harm?

 

[Charlietus]

Could you provide an example of where this negativity caused undue harm?

The first thing that comes to mind is any late port to PC. We know from the insomniac leaks that those kind of ports are really, really cheap, but everytime one comes out and it doesn't lit the charts on fire the game gets labelled as a flop.

Other examples are multiplatform games that do low numbers on PC and higher numbers on console and, as a last example, sequels to games that do lower numbers than the predecessors that receive the treatment without knowing the budget.

YouTube is full of those people posting those videos with thumbnails of steam charts trying to bring games down, and it's so tiring.

 

[iroboto]

Could you provide an example of where this negativity caused undue harm?

I think it’s common place where investors are purposely trying to nose dive companies because they shorted their stock.

They are most definitely trying their best to cause other investors to lose confidence in the business anyway possible.

Futures imo, has ruined the market. Instead of keeping people honest, they now have direct interest to nose dive or overhype stocks for their own benefit.

 

[techuse]

The first thing that comes to mind is any late port to PC. We know from the insomniac leaks that those kind of ports are really, really cheap, but everytime one comes out and it doesn't lit the charts on fire the game gets labelled as a flop.

Other examples are multiplatform games that do low numbers on PC and higher numbers on console and, as a last example, sequels to games that do lower numbers than the predecessors that receive the treatment without knowing the budget.

YouTube is full of those people posting those videos with thumbnails of steam charts trying to bring games down, and it's so tiring.

How does any of this cause harm?

I think it’s common place where investors are purposely trying to nose dive companies because they shorted their stock.

They are most definitely trying their best to cause other investors to lose confidence in the business anyway possible.

Futures imo, has ruined the market. Instead of keeping people honest, they now have direct interest to nose dive or overhype stocks for their own benefit.

I really doubt SteamDB withholding some info will change this.