Support NeoGAF

[ChiefDada]

Ok….but I’m not sure how that relates to what I said.

I was specifically responding to the portion of your comment that I bolded. Everything prior I agreed 100%. You said it's "safe to assume" lack of optimization by the developer in cases where there is significant performance delta. I think that logic is problematic because a developer can optimize to hell and back and there can still be a significant delta that favors the lower TF machine.

I’m talking about forum responses, not DF. I’m also referring to games with significant differences in performance between PS5 and XSX.

Yes, and I am saying that DF is responsible for this flawed logic that many here in this forum use repeatedly.

Secondly, we have to have a consensus for what constitutes as "significant difference". In the context of performance benchmarking, coupled with adopting the flawed methodology of using TF as ultimate measure of performance, any performance benchmark where PS5 outperforms Series X should be considered significant. Why? Because the 12TF XSX is 20% faster than the 10TF PS5 out the gate, based on pure theoretical compute. So any benchmark that shows a PS5 performance advantage indicates that PS5 has overcome a 20% deficit at minimum. In the case of Wild Hearts (since we're discussing in this thread), a 10-12fps PS5 advantage suggest the 10TF PS5 is performing as a 14-15TF console vs the 12TF XSX, or conversely, the 12TF XSX is performing as an 8TF console vs the 10 TF PS5, depending on how you want to organize the numbers/relationship. If you replace both consoles with any two PC cards with the same architecture, ALL tech outlets, DF included, would categorize such results as "significant". I would argue that even when performance is equivalent, that is a significant performance delta in favor of the lower TF console.

 

[Hoddi]

It's not all about loading though, streaming plays a big part of way It's twice as fast.

"So when I talked about the dream of an SSD part of the reason for that 5 gigabytes a second target was to eliminate loads, but also part of the reason for that target was streaming. As in what if the SSD is so fast that as the player is turning around, it's possible to load textures for everything behind the player in that split second.

If you figure that it takes half a second to turn that's 4GB of compressed data you can load, that sounds about right for next gen."

I wouldn't go into details, but there's lots of gameplay mechanics that is now possible with those kind of speeds.

I think that's part of the contention because there are no games that stream data like that yet. And I mean not even close.

I recorded a full playthrough of Rift Apart and it only read ~1.5TB from disk over 8 hours or so. It's much the same deal for any other PS5 exclusive that I've tried.

 

[Topher]

I was specifically responding to the portion of your comment that I bolded. Everything prior I agreed 100%. You said it's "safe to assume" lack of optimization by the developer in cases where there is significant performance delta. I think that logic is problematic because a developer can optimize to hell and back and there can still be a significant delta that favors the lower TF machine.

As I said, the vast majority of games have only slight differences. Games like this are the outliers. So what is different? Not the hardware. The devs and the tech they use are. We can do all the math in the world and the human element is still an unknown variable.

I made similar statements when PS5 was the worse performing system in games like Lego Star Wars. So what are we saying here? Somehow that game's horrible performance on PS5 wasn't about lack of optimization by the developer? No, I don't think my logic here is flawed at all. I will say it can very well be more than just the "lack of optimization". Some engines seem to favor one system versus another. But even then, that is about choices the developer made, not the hardware itself.

However, again... that PS5 SSD speeds would ultimately mean more for devs than for us,

And I'm talking about the aspects that mean something to us. So I'll agree to disagree.

 

[Heisenberg007]

I'm not saying anything is "false". But there has been little difference in practice compared to XSX, which was my original point.

There is; it's just not as pronounced on the user end. You'll have to look at file sizes to understand that PS5's SSD and I/O are creating a difference when compared with XSX in games.

For example, Atomic Heart's loading speed on XSX is 19 seconds. On PS5, it is 13 seconds. The difference is PS5 is 46% faster. But what's less evident is that PS5 also has a 39% smaller file size.

Atomic Heart's file size is 39% smaller on PS5 (48 GB) than XSX (78 GB)

So what's really happening here is that PS5 SSD and I/O are taking a smaller file size (48 GB), decompressing it, and streaming the data in only 13 seconds. Means it works significantly faster than the XSX's SSD and I/O.

Even if the loading speed were the same on PS5 and XSX (say 19 seconds), PS5's SSD and I/O would still be deemed as ~50% faster because it is decompressing data faster from a much smaller (46%) file size.

That data streaming was advertised as the core benefit of PS5's SSD and I/O. So it's working exactly as it should. Therefore, it's fair to say that PS5's SSD and I/O were not falsely overhyped, and they are delivering as was expected.

 

[sachos]

This happened yet again in this week's DF Direct, although John did come around to say that developers have told him they prefer PS5 API

You remember what Direct/timestamp it was? I would like to hear that. I love DF and watch all their videos but i'll admit i've been kinda disappointed lately in the lack of side by side performance comparisons, be it console vs console or PC with similar settings vs console.

 

[sachos]

So what's really happening here is that PS5 SSD and I/O are taking a smaller file size (48 GB), decompressing it, and streaming the data in only 13 seconds. Means it works significantly faster than the XSX's SSD and I/O.

Even if the loading speed were the same on PS5 and XSX (say 19 seconds), PS5's SSD and I/O would still be deemed as ~50% faster because it is decompressing data faster from a much smaller (46%) file size.

Wow can't believe i haven't thought about it that way before. Can't wait to see what late gen first party games look like on PS5. I'm actually kinda disappointed in Xbox's load times, specially from first party games, haven't seen that many ~2 seconds load times on SX yet like we have on PS5, unless i missed some.

Sorry for double post.

 

[Topher]

There is; it's just not as pronounced on the user end. You'll have to look at file sizes to understand that PS5's SSD and I/O are creating a difference when compared with XSX in games.

For example, Atomic Heart's loading speed on XSX is 19 seconds. On PS5, it is 13 seconds. The difference is PS5 is 46% faster. But what's less evident is that PS5 also has a 39% smaller file size.

Atomic Heart's file size is 39% smaller on PS5 (48 GB) than XSX (78 GB)

So what's really happening here is that PS5 SSD and I/O are taking a smaller file size (48 GB), decompressing it, and streaming the data in only 13 seconds. Means it works significantly faster than the XSX's SSD and I/O.

Even if the loading speed were the same on PS5 and XSX (say 19 seconds), PS5's SSD and I/O would still be deemed as ~50% faster because it is decompressing data faster from a much smaller (46%) file size.

That data streaming was advertised as the core benefit of PS5's SSD and I/O. So it's working exactly as it should. Therefore, it's fair to say that PS5's SSD and I/O were not falsely overhyped, and they are delivering as was expected.

Not saying it isn't working as it should. I'm saying exactly what you said in your first sentence: "it's just not as pronounced on the user end". It was hyped on the user end though. They used Ratchet and Clank to do just that, but what we saw in R&C hasn't been seen since.

 

[Physiognomonics]

Interestingly I noticed some big frame-time spikes (up to 80ms) on XSX in the performance mode (they didn't talk about them in their video). While they say PS5 still gets the performance edge in that mode PS5 doesn't have any of those stutters (in their footage at least) and only get max 33ms frame-time spikes.

It's not a surprise as the PC also get those big frame-time spikes (that can't be hidden by VRR). So PC and Xbox do have I/O stutters at least.

Seems the best way to play this (PC included) would be PS5 + VRR in the perf mode as the framerate drops under 48fps seem really rare. Are people still surprised PS5 seemingly doesn't get any I/O stutters in that game while all the others versions are having those?

 

[supernova8]

Forgetting the whole Series S versus the power of the Series X and PS5, the game just looks like fucking shit. If this is "next-gen" then... I'm good thanks.

 

[Physiognomonics]

There is; it's just not as pronounced on the user end. You'll have to look at file sizes to understand that PS5's SSD and I/O are creating a difference when compared with XSX in games.

For example, Atomic Heart's loading speed on XSX is 19 seconds. On PS5, it is 13 seconds. The difference is PS5 is 46% faster. But what's less evident is that PS5 also has a 39% smaller file size.

Atomic Heart's file size is 39% smaller on PS5 (48 GB) than XSX (78 GB)

So what's really happening here is that PS5 SSD and I/O are taking a smaller file size (48 GB), decompressing it, and streaming the data in only 13 seconds. Means it works significantly faster than the XSX's SSD and I/O.

Even if the loading speed were the same on PS5 and XSX (say 19 seconds), PS5's SSD and I/O would still be deemed as ~50% faster because it is decompressing data faster from a much smaller (46%) file size.

That data streaming was advertised as the core benefit of PS5's SSD and I/O. So it's working exactly as it should. Therefore, it's fair to say that PS5's SSD and I/O were not falsely overhyped, and they are delivering as was expected.

I'd say the difference of size is because data is deduplicated on PS5 only (we have seen a similar outcome in others games like Control). PS5 loading in 13 seconds is 10 seconds more than it should on that platform based on some others games.

 

[Heisenberg007]

I'd say the difference of size is because data is deduplicated on PS5 only (we have seen a similar outcome in others games like Control). PS5 loading in 13 seconds is 10 seconds more than it should on that platform based on some others games.

No, duplication is eliminated on both PS5 and XSX|S because of fast SSDs, so file sizes are smaller to an extent on all consoles.

Beyond that, however, devs can get the PS5 file size even smaller because of additional compression. And PS5's SSD and I/O can decompress data from a highly compressed file package and still load games faster than XSX, which is decompressing data from a relatively less compressed package.

 

[jasondefaoite]

Removing Series S versions wouldn’t just result in a PR hit but a class action lawsuit. They promised a system that could run ALL next-gen content and it’s especially damning when the versions reside within the same download file. Imagine downloading a game and receiving an error that it’s incompatible with your Series S but runs fine on a Series X? The mandate isn’t even a mandate it’s a requirement.

I don't recall if MS broke down the sales %. I'd imagine XSS sold more than the XSX given the price, but even if it was 50/50, there is no way you can cut 50% of your install base.

 

[ToTTenTranz]

On the Series S, the game looks and performs like a PS4 game on a PS4. On the PS5/X, it looks and performs like a PS4 game on a PS4 Pro.


Koei Tecmo / Omega Force are a full generation behind in visuals and performance optimization.
However looking at the massive output of titles this dev house does, I don't think this is a case of "lazy devs". Just look at what they released within the last ~5 years:

- Wild Hearts

- Samurai Warriors 5 2021 (Switch, PS4, XBO, Windows)
- Dynasty Warriors 9 2018 (Switch, PS4, XBO, Windows, Series, PS5)
- Warriors Orochi 4 2018 (Switch, PS4, XBO, Windows)
- Touken Ranbu Warriors 2022 (Switch, Windows)
- Fire Emblem Warriors: Three Hopes 2022 (Switch)
- Persona 5 Strikers 2020 (Switch, PS4, XBO, Windows)
- Dragon Quest Builders 2 2018 (Switch, PS4, XBO, Windows)
- Attack on Titan 2 2018 (Switch, PS4, XBO, Windows)
- Hyrule Warriors: Age of Calamity 2020 (Switch)
- One Piece: Pirate Warriors 4 2020 (Switch)

And they're also launching a game called Fate/Samurai Remnant this year.


They're averaging >2 games a year. Even if this is a 1000-strong developer these guys seem to barely have the eat and sleep, let alone evolving and adapting their custom engine.

 

[yazenov]

Taking sarcastic jabs at the guy is unnecessary, imo.

It was a joke. I suggest removing that stick from your butt.

He doesn't need any help in his defense. If he didn't like the joke he could inform me directly.

 

[Topher]

It was a joke.

 

[MarkMe2525]

Confirmation bias is strong in this thread. Anyone with any sense has had the sentiment that certain engines would favor either Sony's or MS's architecture. We are two years into the generation and I believe that this has shown to be true.

Now we are looking at a game that is a technical shit show on all platforms, but because it is less shit on one of them, this is taken as evidence of it's superior composition? I just don't understand what is going on in here.

 

[Topher]

Confirmation bias is strong in this thread. Anyone with any sense has had the sentiment that certain engines would favor either Sony's or MS's architecture. We are two years into the generation and I believe that this has shown to be true.

Now we are looking at a game that is a technical shit show on all platforms, but because it is less shit on one of them, this is taken as evidence of it's superior composition? I just don't understand what is going on in here.

This is pretty normal for DF warrior-fests where one version outpaces the other significantly.

 

[rodrigolfp]

What`s going on this generation?

- This game: Total mess
- Dead Space Remake : Sub 1080p internal resolution @ 60 fps
- Hogwarts legacy: Also pretty unstable 60 fps
- Evil West: Only 1080p @ 60 fps
- Gotham Knights: No comment
- Halo Infinite: No comment

And the list goes on and on. Also, I've only talked about PS5 and XSX so far. XSS is a technical disaster³.
The PC as a platform isn't much better off either, with its constant "stuttering issues" and generally poor optimization.

At least everyone is patching the stutters on PC later...

 

[MarkMe2525]

This is pretty normal for DF warrior-fests where one version outpaces the other significantly.

I agree with this. I just don't understand how these same posters justify their statements when undoubtedly there will be a game released in the coming weeks that will show the opposite. This goes for both sides of the fence here.

To my eyes, as someone who visits NeoGAF regularly and keeps up with such things, it has been a back and forth all generation. Like clockwork, no matter which console has better performance, the thread is inundated with posters stating, "See, the (insert name of game platform) is obviously the better designed box". It's so bizarre.

 

[solidus12]

Goddamn, I just remember this tweet from Penello, ex Xbox exec.

 

[Topher]

Goddamn, I just remember this tweet from Penello, ex Xbox exec.

Marketing folks say some funny shit

 

[solidus12]

Marketing folks say some funny shit

Indeed, seems like before launch, Xbox people (influencers, gamers, journalists, and execs) were on a crusade against the PS5…

 

[Portugeezer]

So fast that it doesn't even render stuff.

 

[Mr Moose]

Goddamn, I just remember this tweet from Penello, ex Xbox exec.

I remember when the Xbox One was launching and that dude came on here and lied his fucking ass off, he's a clown.

 

[ChiefDada]

You remember what Direct/timestamp it was? I would like to hear that. I love DF and watch all their videos but i'll admit i've been kinda disappointed lately in the lack of side by side performance comparisons, be it console vs console or PC with similar settings vs console.

The DF Direct that comes out later today. Around 49:00 mark.

 

[Lysandros]

I was specifically responding to the portion of your comment that I bolded. Everything prior I agreed 100%. You said it's "safe to assume" lack of optimization by the developer in cases where there is significant performance delta. I think that logic is problematic because a developer can optimize to hell and back and there can still be a significant delta that favors the lower TF machine.



Yes, and I am saying that DF is responsible for this flawed logic that many here in this forum use repeatedly.

Secondly, we have to have a consensus for what constitutes as "significant difference". In the context of performance benchmarking, coupled with adopting the flawed methodology of using TF as ultimate measure of performance, any performance benchmark where PS5 outperforms Series X should be considered significant. Why? Because the 12TF XSX is 20% faster than the 10TF PS5 out the gate, based on pure theoretical compute. So any benchmark that shows a PS5 performance advantage indicates that PS5 has overcome a 20% deficit at minimum. In the case of Wild Hearts (since we're discussing in this thread), a 10-12fps PS5 advantage suggest the 10TF PS5 is performing as a 14-15TF console vs the 12TF XSX, or conversely, the 12TF XSX is performing as an 8TF console vs the 10 TF PS5, depending on how you want to organize the numbers/relationship. If you replace both consoles with any two PC cards with the same architecture, ALL tech outlets, DF included, would categorize such results as "significant". I would argue that even when performance is equivalent, that is a significant performance delta in favor of the lower TF console.

I see where you are coming from and i agree mostly, but your counter argument to teraflop centric approach appears to be somewhat teraflop centric in itself here don't you think?

 

[GymWolf]

So, are the performance still horse shit on pc?

 

[onQ123]

It's a genuine gaming question on a gaming forum. When was the last time a more powerful machine was destroyed in performance by the 'weaker' machine? I would say 360 and ps3 but the 360 wasn't that much weaker than the PS3.

Saturn vs N64? Saturn Vs Ps1?

PS5 isn't weaker it just has less Compute

Goddamn, I just remember this tweet from Penello, ex Xbox exec.

 

[GymWolf]

Goddamn, I just remember this tweet from Penello, ex Xbox exec.

Fun fact: penello in italian language can be roughly translated as "small dick"

 

[ChiefDada]

I see where you are coming from and i agree mostly, but your counter argument to teraflop centric approach appears to be somewhat teraflop centric in itself here don't you think?

Yes, and that's my point. The flawed logic of the TF dictating which platform should have performance edge is, imo, causing Topher and others to say "oh, it must be because of lack of optimization" when in reality the other aspects of a given platform such as clock speeds, memory setup, I/O, more efficient API, etc are much more likely to explain the results. The most interesting conversations are being left on the table. Of course the PS5 isn't "acting like a 14-15TF console", it's acting like the 10TF console it is with higher clock speed and a leaner API suite. Of course the XSX isn't "acting like an 8TF console", it's acting like the 12TF console it is with a more general purpose API, etc. So when we see these differences, we should all want to discover and discuss the other aspects of a given platform that is responsible for the differences. My point is I wish outlets like DF would explore these other platform specific areas in far greater detail when performance deltas like these occur because it is an opportunity to broaden the general public's understanding of how these machines truly work. Instead they erroneously base all their hypothesis on GPU compute, and when that doesn't pan out all we get is a "hmmm that's weird, on to the next one".

 

[onQ123]

Yes, and that's my point. The flawed logic of the TF dictating which platform should have performance edge is, imo, causing Topher and others to say "oh, it must be because of lack of optimization" when in reality the other aspects of a given platform such as clock speeds, memory setup, I/O, more efficient API, etc are much more likely to explain the results. The most interesting conversations are being left on the table. Of course the PS5 isn't "acting like a 14-15TF console", it's acting like the 10TF console it is with higher clock speed and a leaner API suite. Of course the XSX isn't "acting like an 8TF console", it's acting like the 12TF console it is with a more general purpose API, etc. So when we see these differences, we should all want to discover and discuss the other aspects of a given platform that is responsible for the differences. My point is I wish outlets like DF would explore these other platform specific areas in far greater detail when performance deltas like these occur because it is an opportunity to broaden the general public's understanding of how these machines truly work. Instead they erroneously base all their hypothesis on GPU compute, and when that doesn't pan out all we get is a "hmmm that's weird, on to the next one".

The threads & posts I used to make explaining this stuff should have stayed open then we wouldn't have this problem so widespread here but instead no one backed me up & the mods just locked the threads like I was talking crazy.

 

[Bogroll]

Yes, and that's my point. The flawed logic of the TF dictating which platform should have performance edge is, imo, causing Topher and others to say "oh, it must be because of lack of optimization" when in reality the other aspects of a given platform such as clock speeds, memory setup, I/O, more efficient API, etc are much more likely to explain the results. The most interesting conversations are being left on the table. Of course the PS5 isn't "acting like a 14-15TF console", it's acting like the 10TF console it is with higher clock speed and a leaner API suite. Of course the XSX isn't "acting like an 8TF console", it's acting like the 12TF console it is with a more general purpose API, etc. So when we see these differences, we should all want to discover and discuss the other aspects of a given platform that is responsible for the differences. My point is I wish outlets like DF would explore these other platform specific areas in far greater detail when performance deltas like these occur because it is an opportunity to broaden the general public's understanding of how these machines truly work. Instead they erroneously base all their hypothesis on GPU compute, and when that doesn't pan out all we get is a "hmmm that's weird, on to the next one".

No I'd say the Xbox is acting like a 8tf machine in this game. Missing AO and upto 10fps difference at times
(If I remember correctly)

 

[damidu]

So fast that it doesn't even render stuff.

these shills fed stupid takes like this to their base for months up to console launch.
setting expectations for a repeat of
ps4pro vs. onex situation

thats why we are still dealing with these traumatized fanboys whenever reality hits their made-up powaa narrative

 

[Crispy Gamer]

The DF Direct that comes out later today. Around 49:00 mark.

Hmmm listening to John now and he hinted there's more games they're testing that have similar results as the Hogwarts and Calisto analysis

 

[Heisenberg007]

Hmmm listening to John now and he hinted there's more games they're testing that have similar results as the Hogwarts and Calisto analysis

Time stamp a brother!

 

[MarkMe2525]

Hmmm listening to John now and he hinted there's more games they're testing that have similar results as the Hogwarts and Calisto analysis

He gave an interesting answer with many scenarios I haven't thought of before. I personally didn't realize there was a correlation with UE4 engine games generally running slightly better on PS5.

I also didn't factor in the significance of project managers that are assigned to each version of the game that can account for the differences.

Time stamp a brother!

49 minute mark

 

[Heisenberg007]

Hmmm listening to John now and he hinted there's more games they're testing that have similar results as the Hogwarts and Calisto analysis

It seems like they were referring to Atomic Heart and Wild Heart.

 

[Art The Clown]

I’m enjoying the game but it isn’t a beaut that’s for sure

 

[Lysandros]

I don't know if posted before but XSX seems to lack parallax occlusion mapping on rocks here.

 

[Crispy Gamer]

It seems like they were referring to Atomic Heart and Wild Heart.

I think you might be right at any rate, John's explanation made a lot of sense devs just like Sony's development environment more its never been just about the hardware I'll never forget Ted Price's quote about the PS1 when Sony entered the business in how they were the first company to infuse hardware and software together I can't find the video but it always stuck with me. Still love my Series X but it's been extremely disappointing in the performance department this generation I just didn't expect this at all I think it’s time for DF or NX to do a video on it or Microsoft needs to do another "investigation" I refuse to believe it's a hardware issue...

 

[onQ123]

I think you might be right at any rate, John's explanation made a lot of sense devs just like Sony's development environment more its never been just about the hardware I'll never forget Ted Price's quote about the PS1 when Sony entered the business in how they were the first company to infuse hardware and software together I can't find the video but it always stuck with me. Still love my Series X but it's been extremely disappointing in the performance department this generation I just didn't expect this at all I think it’s time for DF or NX to do a video on it or Microsoft needs to do another "investigation" I refuse to believe it's a hardware issue...

It's in MS papers Xbox Series X is 6-8X the GPU power of Xbox One but able to use new tricks to get 10X the pixel count out of it.
I say that to say this GPU hardware has changed a lot over the years somethings took a hit as they made these changes because we moved away from the old way of doing things but some games will rely on the older way to do things & it will run into these bottlenecks.

 

[Pimpbaa]

Arguing about performance in a game that doesn’t look and only does 1080p in performance mode on consoles and performs shitty on PC as well. This is very low resolution hill to die on.

 

[Loxus]

I think you might be right at any rate, John's explanation made a lot of sense devs just like Sony's development environment more its never been just about the hardware I'll never forget Ted Price's quote about the PS1 when Sony entered the business in how they were the first company to infuse hardware and software together I can't find the video but it always stuck with me. Still love my Series X but it's been extremely disappointing in the performance department this generation I just didn't expect this at all I think it’s time for DF or NX to do a video on it or Microsoft needs to do another "investigation" I refuse to believe it's a hardware issue...

After putting some thought into it, there might be a possibility that the PS5's GPU may have a slightly higher IPC than the XBSX's GPU.

While both the PS5 and XBSX are of the same architecture RDNA2, the PS5 is based on RDNA2 and the XBSX integrates full RDNA2.

"First we have a custom AMD GPU based on their RDNA2 technology"

"How Xbox Series X|S Integrates Full AMD RDNA 2 Architecture"

The difference in PS5 being based on RDNA2, could mean different IPC gains as they did their own customizations on top on RDNA2.
They did at least one major revision too, that most llikely improved IPC gains as well.

This plus all the other things going on with the PS5.
Performance = the sum of all parts.

At this point in time, I wouldn't tell someone their wrong if they came up to me and said the PS5 is currently the most powerful console based on performance from current gen games, not to mention it holds the title for best looking games on consoles.

 

[Lysandros]

After putting some thought into it, there might be a possibility that the PS5's GPU may have a slightly higher IPC than the XBSX's GPU.

While both the PS5 and XBSX are of the same architecture RDNA2, the PS5 is based on RDNA2 and the XBSX integrates full RDNA2.

"First we have a custom AMD GPU based on their RDNA2 technology"

"How Xbox Series X|S Integrates Full AMD RDNA 2 Architecture"

The difference in PS5 being based on RDNA2, could mean different IPC gains as they did their own customizations on top on RDNA2.
They did at least one major revision too, that most llikely improved IPC gains as well.

This plus all the other things going on with the PS5.
Performance = the sum of all parts.

At this point in time, I wouldn't tell someone their wrong if they came up to me and said the PS5 is currently the most powerful console based on performance from current gen games, not to mention it holds the title for best looking games on consoles.

Correct me if wrong, shouldn't the more efficient machine have higher (real world) IPC by nature?

 

[Physiognomonics]

I see a little bit of something about these two consoles. i just listen to supposedly experts talk about about them. but i was not paying attention:

time: 9:28.

Very interesting podcast with NXGamer. He has very interesting theories about why consoles games ported on PC are running so poorly. Notably in Forspoken port what he said really make sense and actually explains a lot of things.

 

[Germuso]

So another one?

Meh awful Game however.

 

[yamaci17]

this is practically what happened with 2 gigs gtx 770 and similar cards. they all went to the texture hell. despite adoring the minimum specs for years, you had to play with absolute dogshit textures just because you were 1-1.5 gb shy of required memory budget for textures

devs never bothered creating special set of textures tailored for lower vram, no. otherwise 2 gig owners would stay on their card forever. that simply does not happen.

now it has to happen with series s, as it needs special set of textures if devs target sx/ps5 first for their textures. but as we can see from this game, such textures simply do not exist and game loads dogshit textures once it runs out memory budget

it is not really easy to push a button and scale down textures, no. in most games you play, there are only 1 set of textures fixed, and texture setting usually controls stream budget or it just brutally murders textures to save a couple hundreds meg here and there.

rdr 2, ac valhalla, spiderman, many more. reduce one tick from the maximum texture quality setting and you get dogshit textures. this was a pretty common occurence between 2014-2021 games. you either had enough vram at 1080p for maximum textures that required 3-4 gb vram, or you had to play with dogshit textures, most of the time.

i dont know if series s can provide enough incentive to force devs to create special set of additional textures. that would also hurt Jensen's plans to obsolete all RTX 8 GB cards. if devs really go beyond and create actual nice looking additional subset of textures that does not look like dogshit and is presentable, that would practically make all 6 - 8 GB GPUs to last the whole generation, and thats a thing Jensen wouldn't like either

so I dunnot really whats will happen. we shall see, I guess

you can see with forspoken, all 6-8 gig rtx gpus were obsoleted, proper textures wont even load at 1080p, it is literally unplayable for that userbase (thankfully game sucks). Jensen would actually like that to happen across more and more games, so that massive 8 gb userbase can migrate to his newer cards.

Once again, this is not about the capability of Series S or whether budget is enough or not. if you create special set of textures, series s can play with nice graphics.

same could be said for gtx 770. it was a powerful GPU, yet dogshits textures obsoleted it. if devs wanted to, they could've created special subset of textures that are tailored towards 2 gb budget. they did not have any incentive, since NV actually preyed on this instance to force-migrate massive gtx 600-700 userbase to 900-1000 cards.

there are rare games that had textures tailored for 2 GB cards, such as arkham knight, sekiro and stuff but they were rare. past 2017, it even became rarer for a dev to cater for 2 gb userbase, as it became a niche userbase (hello, Series S too has a niche userbase, lol)

gtx 770 can run arkham knight and rise of tomb raider with gorgeous visuals, since both have special texture packs tailored for 2 gb. but with shadow of tomb raider, you haveo literally destroy and demolish textures just so that game can somehow fit into that buffer. it is what it is.

this shouldn't happen on Series S, or rather, should be avoided. but it is a work for devs that will also have lasting impact on the usability of low VRAM GPUs on PC space. lets see what will they choose.

it will depend on the backlash from series s userbase when this starts to become a more freqent occurence. if the backlash is not enough and majority of series s userbase is content with dogshit textures, you can kiss the possibility of a dev sitting down and creating special subset of textures a goodbye

 

[begotten]

Damn, I forgot this game even released but it seemed like it could have promise. I guess it was shit and MHR killed it?

 

[saintjules]

What`s going on this generation?

- This game: Total mess
- Dead Space Remake : Sub 1080p internal resolution @ 60 fps
- Hogwarts legacy: Also pretty unstable 60 fps
- Evil West: Only 1080p @ 60 fps
- Gotham Knights: No comment
- Halo Infinite: No comment

And the list goes on and on. Also, I've only talked about PS5 and XSX so far. XSS is a technical disaster³.
The PC as a platform isn't much better off either, with its constant "stuttering issues" and generally poor optimization.

Makes you wonder, right? I always thought it was a timeframe thing with investors looking to make their money back asap.

 

[STARSBarry]

Damn, I forgot this game even released but it seemed like it could have promise. I guess it was shit and MHR killed it?

It's groundwork is solid, but it's god awful preformance killed it. Depending on how they handle this it will either slowly come into its own or shot in the back with buckshot into the trench with everything else that does not immediately pan out for EA.