Support NeoGAF

[IntentionalPun]

People in this thread seem to think the optimizations for PS4 / XBO won't also have helped this port. They almost certainly did.

A lot of that is optimized for is just the game running on weak hardware, limiting memory use, getting around shit CPU.

The game may be further optimized on Switch 2 considering its running an ARM processor but I wouldn't expect the level of change we had from PS4 launch to the final version.

 

[Buggy Loop]

Let's see how accurate you were

So 100000x better than you but ok

but that's ALL you could muster to reply after I obliterated your reputation of tech?

9-14X.

Reality: 5X

Wow, 5X, you keep clowning

Tell me how Switch 1 to Series S quality street fighter 6 is 5X

Switch 1 Mortal Kombat

Switch 2 SF6

Or Cyberpunk 2077 Phantom Liberty in this very thread

Or Hogwarts Legacy

It's totally 5nm guys! I did some maths

Proven wrong time and time again on this

And nobody has the information on the node, doofus. But that has nothing to do with graphical output, just battery consumption.

If 8nm vs 5nm is the only thing you can muster in months of speculation, compared to the dumbest fucking takes you made in the past months, then that's one of the saddest attempt at a come back I've seen in 21 years on Neogaf.

 

[Gaiff]

Switch 1 was the X1, it was not based on it. Die shot showed it was exactly the same.

Which only strengthen my point, doesn't it?

Switch 2 got a lot more changes from Tegra T234 (clock gating from Lovelace new CPU cluster, file decompression engines etc…) and they were not force to do that either. For all we know the next Tegra successor was on the table (nVIDIA has been working for years on new APUs with ARM CPUs and IGPs for the handheld and PC market).


Guess what… there was no chip like what XSX|S and PS5 received on the market and then lo-and-behold there were (let alone PS5 Pro) but those are semicustom design not designed from scratch, AMD and nVIDIA have modularised blocks they can customise and arrange based on what customers want and are willing to pay for.

But that goes with the presumption that they designed those chips just for them when we know it's not the case. Their roadmap was established years in advance and Sony/Microsoft customized chips already in the production pipeline. That's why we saw eerily similar designs hit the consumer market the same month. In addition, you're also forgetting that NVIDIA doesn't design CPUs outside of their mobile SOCs whereas the consoles picked up custom versions of the Ryzen 3600. Nintendo would have had to either wait, use the T239 CPU anyway, or customize it. In the example you're giving, it would be a unique SoC without a consumer-grade counterpart and the fact that it's Arm and they cannot just use an off-the-shelf CPU further complicates things compared to the two others.

Given the market condition and circumstances, I don't think it's reasonable to expect Nintendo to have NVIDIA design a unique Tegra "T400" with a Lovelace-powered GPU and some unknown CPU. Also, given NVIDIA's history with partners, Nintendo might have done well just sticking to customizing the T234.

https://www.neogaf.com/threads/nvidia-frustrated-by-nintendo-is-working-on-brand-new-soc-for-gaming-handhelds-consoles.1670559/post-269191898

There were some articles and other rumours around this, nVIDIA wanting to get a better crack at mobile gaming and handheld PCs and everything outside of automotive Tegra was being optimised for.
Based on their history, look at CUDA and how much money they spent to make all of their line compatible with it (and then adding RT and Tensor Cores… "The nVIDIA Way" is a good read/listen btw), nVIDIA does not want to fragment their own market in terms of baseline capabilities and wanted Switch 2 to help them push a higher baseline.

Rumors and not seeing where it says they were pushing Nintendo to a different node/arch?

 

[James Sawyer Ford]

So 100000x better than you but ok

but that's ALL you could muster to reply after I obliterated your reputation of tech?

Wow, 5X, you keep clowning

Tell me how Switch 1 to Series S quality street fighter 6 is 5X

Switch 1 Mortal Kombat

Switch 2 SF6

Or Cyberpunk 2077 Phantom Liberty in this very thread

Or Hogwarts Legacy







And nobody has the information on the node, doofus. But that has nothing to do with graphical output, just battery consumption.

If 8nm vs 5nm is the only thing you can muster in months of speculation, compared to the dumbest fucking takes you made in the past months, then that's one of the saddest attempt at a come back I've seen in 21 years on Neogaf.

Give it up, it's 8nm. Still keeping the delusional hope alive though

5x is a huge difference from 14x. We have the direct comparisons of switch 1 to switch 2 ports running around 5x better

Here you are arguing with known chip leaker KeplerL2 and got schooled, yet you just laughed at his post.

Pages of baseless speculation, hopes and dreams, all shattered and you still won't admit you were wrong

Why are you quoting some random company and not NVIDIA?

Different designs with different ratios of Logic/SRAM/Analog can have very different densities. Look at H100 vs AD102 as an example.

Been a while since I looked at the hacked NVIDIA data, but IIRC it showed T239 having half the cache for both CPU and GPU portions.

No, but it's quite small on PS5/Xbox.

Clock gating adds minimal area.

It definitely does.

You can remove a lot of stuff. As I said it's already confirmed that T239 doesn't have the NPU and all RAS features. A lot of I/O can be cutdown, encoder/decoder can be simplified due to targeting a fixed resolution.

 

[LordOcidax]

Give it up, it's 8nm. Still keeping the delusional hope alive though

5x is a huge difference from 14x. We have the direct comparisons of switch 1 to switch 2 ports running around 5x better

Switch 1 doesn't perform as well as the PS3 in docked mode, so I believe that Switch 2 in docked mode will probably be shy of the PS4 as well. But even if it does exceed it (through DLSS or whatever), it won't be by much.

 

[James Sawyer Ford]

Fair enough, looks like T234 had some changes from announcement to release.

This doesn't address my argument. Different dies even on the same node with the same architecture can have very different densities due to different ratios of Logic/SRAM/Analog cells or use of different libraries. FYI the most recent rumor from people with NVIDIA sources is that T239 is using 8LPU, which is slightly denser and more power efficient than 8LPP used by gaming Ampere/T234 due to smaller fin size.

I seem to recall T239 GPU only having 2MB of GPU L2 cache from either the NVIDIA hack or Linux open source patches. Same for the Ada clock gating features. No, I'm not gonna bother looking for proof of it. Ask someone at Famiboards and I'm sure they can tell you if I'm right or wrong.

I can't because it's too small.

Not sure about NVIDIA but AMD does that all the time. Intel does it sometimes too.

You're right, this is a waste of time. Just wait 3 months and you'll be disappointed that Nintendo cheaped out on HW like they always do.

 

[James Sawyer Ford]

Digital foundry this week stated it's around ps4 power levels

 

[Buggy Loop]

Give it up, it's 8nm. Still keeping the delusional hope alive though

For 8nm nothing is confirmed, just a fact. It does not change the graphical envelope of the Switch 2, just the power consumption. Nvidia never goes full density on any nodes, be it Samsung or TSMC so the impact is nowhere near the foundry claims in improvements. You peoples have no idea about semiconductors.

Meanwhile, you predicted dock not surpassing base PS4

Just to highlight it again

5x is a huge difference from 14x. We have the direct comparisons of switch 1 to switch 2 ports running around 5x better

5x is out of your ass

Switch 2 will get Star wars outlaws. Doesn't even run on Steam deck. 3.2TFlop docked is exactly 10 times Switch 1 docked. Without even accounting for all the modern architecture. Just computational power.

Here you are arguing with known chip leaker KeplerL2 and got schooled, yet you just laughed at his post

The Kepler's post from 2 years ago? Cute.

 

[LordOcidax]

Digital foundry this week stated it's around ps4 power levels

Dude, you're watching the videos… Cyberpunk looks better than the PS4 version by a country mile You are making the ridiculous at this point.

 

[James Sawyer Ford]

Dude, you're watching the videos… Cyberpunk looks better than the PS4 version by a country mile You are making the ridiculous at this point.

A country mile? No it doesn't. Looks like I was spot on, actually

"If it does look better, not by much"

 

[James Sawyer Ford]

5x is out of your ass

Why isn't BOTW and ToTK running at much higher than 1440p then?

Should be able to push 4k native if what you say is true, but it doesn't

 

[LordOcidax]

A country mile? No it doesn't. Looks like I was spot on, actually

"If it does look better, not by much"

 

[Buggy Loop]

Why isn't BOTW and ToTK running at much higher than 1440p then?

Should be able to push 4k native if what you say is true, but it doesn't

Its a patch, not built from the ground up and by all means it seems Switch 1 games are emulated.

By your definition, why is Phantom Liberty 0p 0fps on PS4 Pro?

 

[Buggy Loop]

A country mile? No it doesn't. Looks like I was spot on, actually

"If it does look better, not by much"

Delusional, holy shit



The base fucking game, not even dog town.

You're completely out of your league. GTFO.

 

[Represent.]

Looks like trash and there's no reason to play this game on such a weak platform.

 

[James Sawyer Ford]

Delusional, holy shit



The base fucking game, not even dog town.

You're completely out of your league. GTFO.

The game has been patched since then, why are you using extremely old footage? It was well known the game was a mess at launch across a huge number of platforms and not optimized at all

 

[naguanatak]

Looks like trash and there's no reason to play this game on such a weak platform.

There are different reasons, even though some people can't seem to see them

 

[FUBARx89]

Weird question. But the handheld mode has a 120hz screen. So 40fps is okay.

But what about docked mode? Are they counting on people having a 120hz TV?

 

[Buggy Loop]

The game has been patched since then, why are you using extremely old footage? It was well known the game was a mess at launch across a huge number of platforms and not optimized at all

Basically like patch 1.2 improvements, which is a tad better 30 fps but cutoff of graphics to get it done

What a progress! Much more stable! The cutoffs were worth it!

Spoiler: Phantom Liberty on PS4

 

[LordOcidax]

Weird question. But the handheld mode has a 120hz screen. So 40fps is okay.

But what about docked mode? Are they counting on people having a 120hz TV?

Obvious…

 

[James Sawyer Ford]

phantom Liberty isn't on ps4 for the same reasons that rdr isn't on Wii U, it's not worth the effort since there's no roi

 

[Hyet]

phantom Liberty isn't on ps4 for the same reasons that rdr isn't on Wii U, it's not worth the effort since there's no roi

Actually CDPR stated why the expansion did not release on PS4 and XBONE

Due to some technical challenges, this change isn't available on the previous generation of consoles," the update blog reads. To get around these challenges in future patches, CDPR is halting work on previous-gen consoles: Patch 1.6 will be the last for Xbox One and PS4

The fact that the expansion runs on NS2 means the platform doesn't present the challeges previous systems did.
They also released a ton of free updates on PS4 that went beyond fixing critical errors, so ROI is hardly an element in this.

 

[James Sawyer Ford]

Actually CDPR stated why the expansion did not release on PS4 and XBONE

The fact that the expansion runs on NS2 means the platform doesn't present the challeges previous systems did.
They also released a ton of free updates on PS4 that went beyond fixing critical errors, so ROI is hardly an element in this.

Technical challenges is buzzword for "we don't think we'll get an ROI porting it"

Technical challenges are overcome by resources ($$)

 

[LordOcidax]

Actually CDPR stated why the expansion did not release on PS4 and XBONE

The fact that the expansion runs on NS2 means the platform doesn't present the challeges previous systems did.
They also released a ton of free updates on PS4 that went beyond fixing critical errors, so ROI is hardly an element in this.

James: They most be lying…

Fake edit: Beaten

 

[Buggy Loop]

Technical challenges is buzzword for "we don't think we'll get an ROI porting it"

Technical challenges are overcome by resources ($$)

CDPR : its technical challenge

James : They didn't want to support their 2nd biggest selling platform because... ROI.

James' source

Whatever happens to CDPR wanting to get its reputation back and wipe up the situation with older gen consoles? Naw. Not in James' world. Selling a high price DLC, why would CDPR want that?

Meanwhile CDPR makes full fledge ultimate edition in 7 weeks on Switch 2, somehow with no technical hurdle. ROI? They use the most expensive cartridge for absolutely no reasons when they could sell you a digital keychain on cart.

You got it James. That's the reason why barely holding 30 fps to mid 20 fps base game didn't receive the even more demanding DLC. ROI...

 

[Hyet]

Technical challenges is buzzword for "we don't think we'll get an ROI porting it"

Updating the game for free: good ROI
Selling an expansion for 20 bucks: bad ROI, better say it's because consoles won't run it.

Sounds like real business

 

[Buggy Loop]

Updating the game for free: good ROI
Selling an expansion for 20 bucks: bad ROI, better say it's because consoles can run it.

Sounds like real business

Selling $30 DLC to 2nd highest userbase and give us a good name back with peoples who were disappointed with launch situation, naw, not worth the under 7 weeks of effort (7 weeks here is full port so far into a more playable state, a patch on PS4 is not a full port).

Those reasons James come up with

Even the ex IRAQ minister of defense would blush

 

[James Sawyer Ford]

Updating the game for free: good ROI
Selling an expansion for 20 bucks: bad ROI, better say it's because consoles won't run it.

Sounds like real business

They have the data on platform active users, the entire audience was on ps5 by the time phantom Liberty released

Yes, real business

 

[TheMan]

looks fine i guess. Just hard to get excited about a game that's been out for years adn that I already bought twice.

 

[LordOcidax]

They have the data on platform active users, the entire audience was on ps5 by the time phantom Liberty released

Yes, real business

 

[rodrigolfp]

Gunplay footage

Gameplay from bad players with controllers are always hilariously sad to see.

 

[Bitstream]

Gameplay from bad players with controllers are always hilariously sad to see.

Have they shown anyone running the game with the joycon using gyro or mouse controls?

 

[rodrigolfp]

Doesn't even run on Steam deck.

Since when??

Have they shown anyone running the game with the joycon using gyro or mouse controls?

 

[FUBARx89]

Obvious…

But the peasents on here cry on that 40fps game modes are pointless because no-one owns a 120hz TV.

 

[Buggy Loop]

Since when??

I mean, yes you can start it, there's no form of playable here though



Need ultra performance FSR 3 on already low resolution baseline frame gen cannot save that of course

It hitches like crazy

 

[Mozza]

Digital foundry this week stated it's around ps4 power levels

Yes a two hour video about the Switch 2's specs. When they had no actual confirmed specs. Sounds totally reasonable.

 

[Hyet]

They have the data on platform active users, the entire audience was on ps5 by the time phantom Liberty released

They had the data, and stated that the reasons that drove their decision were technical.

You have no data, yet contradict what the holders of the data say, because, acording to you, the data that they keep private and you have not seen, supports your argument instead of theirs.

For me the real ROI is reading your answers. Please keep It up.

 

[Mozza]

These diminishing returns came home to roost. The masses will think this is a protable PS5 Pro.

 

[cinnamonandgravy]

Weird question. But the handheld mode has a 120hz screen. So 40fps is okay.

But what about docked mode? Are they counting on people having a 120hz TV?

yeah
go get one if you dont have one. 120hz is big improvement over 60.

but even if you have a 60hz tv, 40fps feels a lot better than 30fps.

 

[Gaiff]

They have the data on platform active users, the entire audience was on ps5 by the time phantom Liberty released

Yes, real business

What's the point of a discussion if you're just gonna ignore proof in favor of a narrative you made up in your head?

 

[kevboard]

even if you have a 60hz tv, 40fps feels a lot better than 30fps.

feels better but looks awful.
and many people go for looks rather than playability.

I played God of War 2018 in performance mode back on PS4, which was very often in the 40fps range... looked stuttery as hell, but the alternative was the input lag laden quality mode, which was unplayable to me (120ms-ish input lag... yikes)

 

[Mozza]

feels better but looks awful.
and many people go for looks rather than playability.

I played God of War 2018 in performance mode back on PS4, which was very often in the 40fps range... looked stuttery as hell, but the alternative was the input lag laden quality mode, which was unplayable to me (120ms-ish input lag... yikes)

The Cyberpunk footage looks pretty smooth to me.

 

[Shifty1897]

I'm just here to cry about there being a Switch 2 Cyberpunk version but not a PS5 Pro version

 

[James Sawyer Ford]

They had the data, and stated that the reasons that drove their decision were technical.

You have no data, yet contradict what the holders of the data say, because, acording to you, the data that they keep private and you have not seen, supports your argument instead of theirs.

For me the real ROI is reading your answers. Please keep It up.

What's the point of a discussion if you're just gonna ignore proof in favor of a narrative you made up in your head?

Let me guess. You guys are also naive enough to believe everything that comes out of a politician or executives mouth too, right?

They said it, therefore clearly it's all true and also no other factors played a role.

Think about it logically, if they said "we aren't releasing on ps4 because it's not worth the return on investment" that's going to cause an inflammatory response as opposed to just hand waiving it away under the guise of technical reasons.

 

[Gaiff]

Let me guess. You guys are also naive enough to believe everything that comes out of a politician or executives mouth too, right?

They said it, therefore clearly it's all true and also no other factors played a role.

Think about it logically, if they said "we aren't releasing on ps4 because it's not worth the return on investment" that's going to cause an inflammatory response as opposed to just hand waiving it away under the guise of technical reasons.

If you have strong supporting evidence this wasn't due to technical issues, post it. Just saying "I don't believe him" isn't a counter-argument.

Phantom Liberty is even more demanding than the base game and it was a disaster on last-gen consoles. Both could also be true.

 

[James Sawyer Ford]

If you have strong supporting evidence this wasn't due to technical issues, post it. Just saying "I don't believe him" isn't a counter-argument.

Phantom Liberty is even more demanding than the base game and it was a disaster on last-gen consoles. Both could also be true.

There's no evidence either way. Where is your proof that ROI didn't play a role? Nothing in the statement says it didn't. He just highlighted one aspect (technical challenges)

 

[Hyet]

Think about it logically, if they said "we aren't releasing on ps4 because it's not worth the return on investment"

Bad news for you is that never happened. What they actually said is they weren't doing it for technical reasons.

that's going to cause an inflammatory response

According to your own argument, this wouldn't happen because the active player data would show nobody cares about those platforms.

Also:

if they said "we aren't releasing on ps4 because it's not worth the return on investment"

If they had said that, would you believe that? Let me guess. You are also naive enough to believe everything that comes out of a politician or executives mouth too, right?

 

[S0ULZB0URNE]

Not bad for a 7 week build.
How long did it take to get running right on PS4 and PS5?

 

[LordOcidax]

There's no evidence either way. Where is your proof that ROI didn't play a role? Nothing in the statement says it didn't. He just highlighted one aspect (technical challenges)

The only obvious thing here is that you are avoiding to accept that the PS4 is not capable to run Phantom Liberty because that will ruin your whole argument about the Switch 2 graphics power… Just take the L and…

 

[Gaiff]

There's no evidence either way.

Dev word is evidence.

Where is your proof that ROI didn't play a role?

You made the claim. You back it with substance

Nothing in the statement says it didn't. He just highlighted one aspect (technical challenges)

And he didn't say they didn't release it because the PS4 is black and they don't like black consoles. Clearly, if we can all believe what we want because he didn't say the opposite, this can go on forever.

Point is we know for a fact technical issues was a reason. That the ROI was too low doesn't change that. The argument is about the performance relative to the Switch 2, which doesn't seem to have technical hurdles, so the ROI is irrelevant.