Support NeoGAF

[BDGAME]

So I guess what i'm asking is - hypothetically - can the switch do at least fantastic ports of last gen (xbox 360/ps3) titles, or what? Like when Bluepoint Games pulled off that magical port of Titanfall on the Xbox 360, hypothetically (not necessarily realistically), could the switch do a flawless version of Destiny 1? Titanfall? Diablo 3?

Or are the big 3rd party games of last gen held back once again by limitations of Nintendo hardware? As I said, reading through Gaf is confusing. I get that the devs aren't necessarily biting due to the unproven install base, but is there a concern that the hardware is incapable of doing those old AAA ports?

Yes, it can.
We already have proofs of that. Let's see some games infos:

Minercraft: bigger maps on switch with a bigger draw distance. The producers say the switch version can handle a bigger resolution too.

Dragon Quest Heros II 1080p on switch vs 720p on ps3 (and that's a bad port)

Cars 3 support 4 players split screen on switch vs 2 players on ps3 and Wii U.

Skyrim: all screens from the switch version are from the remaster version.

Sine Mora EX see this video: https://youtu.be/SHaGfKnPJAU

Last generation most games runs at a 720p or less. Switch can handle these games with more details, better shaders and 1080p resolution.

 

[orioto]

wow so you actually belive there is magic port button in games

i thought it was obvious that the gifs were memes

Any port needs time, people and costs money. I'm saying the downgrade part of it isn't a big enough struggle to say "it's not doable cause of that".

In the end it's always a matter of will.

 

[ViolentP]

I think we will know what the Switch is capable of once we see what the Switch is capable of.

 

[Skittzo0413]

There are certainly some overkill loadouts that bring things to a crawl. That is totally understandable because things in the game can get hectic. It seems as if some effects however are not as refined in this version of the game. Still my favorite game on the device.

My second favorite game on the device after Zelda. Honestly, it's fun as hell to get such a ridiculous loadout that you wind up grinding the framreate to a halt. It reinforces the idea that you can basically break the game, which is neat.

But yeah I don't think we should be judging it's power based on Binding of Isaac.

 

[ViolentP]

My second favorite game on the device after Zelda. Honestly, it's fun as hell to get such a ridiculous loadout that you wind up grinding the framreate to a halt. It reinforces the idea that you can basically break the game, which is neat.

But yeah I don't think we should be judging it's power based on Binding of Isaac.

That was kind of my original point. While Bomberman has a lot more going on, it hits 60fps whereas an older, simpler game like Isaac has hitching problems. Lends to the fact that we are still in the first generation of software and while the specs are written on paper, the Switch's capability will ultimately be dictated by its greatest performer.

 

[oSoLucky]

Year one of PS4/XBO pretty much the whole forum was about users wishing game X to come to PS4/XBO. It was fine then to be excited about your new console and wanting as much support/games on it as possible...PS2/PS3 games getting HD Remasters and new version on PS4 wasnt frowned up on.

I believe it a both sides issue were some troll just started to bring up Switch in every discussion announcement threads, to rile up some community members, creating the situation we have right now were people just feel incredible offended if the System is even brought up.

Well lets hope it slows down after e3...and people realize what we can expect from the system. Right now we are still in the Honemoon phase.

It should slow down soon, and this isn't nearly as bad as the Sony vs MS shit slinging at launch was. Nintendo fans seem a bit more feisty than a lot of the others though. This fall will likely go right back to the launch window bs with the Scorpio release and DF threads will start to become graveyards once again. In the down months there's the PC vs console(usually just PC vs PS/XB1). A seemingly never-ending cycle.

I wonder if 3rd parties do come en masse, if they will bother optimizing for both modes. There are some that seem fine with releasing whatever just to check a box.

 

[10k]

Wii U>>Switch Undocked>>>Switch Docked>>>>>>>>Xbox One>>>>PS4

 

[Drek]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Analogy is probably the best way to understand a system's current horsepower, and ideally with as few additional fixed variables (developers, release window, etc.) as possible.

Thankfully we have a great example with Breath of the Wild. Both made by the same Nintendo development team. Both released at basically the same time. I'd probably give a "pulled from my buttocks" frames per second "average" (i.e. what you'll experience by just playing the game, not stress testing specific areas) on each respective version (Wii U, handheld Switch, docked Switch, later two post patch) at 24 FPS, 29 FPS, 27 FPS.

Now if we just play along and say those are roughly accurate average estimates we can calculate pixel throughput per second, as follows:

Wii U - 720p or 921,600 pixels per frame x 24 frames per second = 22.118M pixels per second.

Switch Handheld - 921,600 pixels per frame x 29 frames per seond = 26.726M pixels per second, a roughly 20% increase in pixel throughput.

Switch Docked - 900p so 1,440,000 pixels per frame x 27 frames per second = 38.88M pixels per second, a roughly 45% increase over the handheld mode, and a 75% increase over the Wii U.

For comparison the PS3 v. PS4 versions of The Last of Us were:
PS3 - 720p (921,600 pixels) per frame x approximately 27 frames per second = 24.883M

PS4 - 1080p (2.1M pixels) per frame x 60 frames per second = 126M

So a roughly 500% increase in pixel throughput from an end of life game ported to new hardware.

*DISCLAIMER* Not all frames are rendered equally. This does not work on a game v. game comparison. These examples used here are to show the difference in performance between the same game from one generation of hardware versus it's next generation release.

So with that out of the way, you can see where a claim of the PS4 being 5 times more powerful than the PS3 is something you can meaningfully back up with math.

The Switch versus the Wii U however shows that the system is much closer to it's predecessor than what we typically associate a generational transition with in the console space.

That is a direct product of the fact that the Switch is not actually the Wii U's direct successor. It is more the successor to the 3DS but design in such a way so as to allow Nintendo to merge their two markets (handheld and console) onto a single platform. Some in this thread have stated that Switch games will perpetually be held back by the handheld mode demand. While technically true, games will be designed based on an acceptable performance threshold in handheld mode, that isn't really holding the games back in my opinion. To me the Switch is much more like buying the latest Nintendo handheld and getting the Switch "Pro" for free via docked mode. This is even more relevant than the PS4 to PS4 Pro or XB1 to Scorpio upgrade as handhelds are traditionally not designed to be capable of outputting a native resolution worthy of being displayed on a TV screen. The Switch solves that out of the box.

I once said on this forum that to me the Wii U wasn't under powered as it delivered all the horsepower needed to deliver Nintendo's artistic vision. Zelda: BotW has made me recant that statement completely. If that is indicative of Nintendo's new level of output going forward I seriously question if the Switch is up to the task. Despite that however it is still the single best hardware offering Nintendo has presented to us since the SNES. No gimmicks like the Wii U tablet or Wii motion controls. Not missing the boat on key hardware features like they did with both the N64 and GC media choices. Not asking contemporary hardware MSRP for clearly aged tech like with the NDS or 3DS.

I loved most of those systems, but the successes were the result of software carrying otherwise questionable hardware across the finish line and the failures were almost always directly tied to their poor hardware choices. The Switch avoids any of the later that I can readily identify and therefore opens the door for the software to carry it.

It is the most powerful gaming dedicated handheld ever made, think of it like that and not as a quarter step into the current generation that it would be as a console, because doing anything else misses the real value in it's design.

 

[bomblord1]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Analogy is probably the best way to understand a system's current horsepower, and ideally with as few additional fixed variables (developers, release window, etc.) as possible.

Thankfully we have a great example with Breath of the Wild. Both made by the same Nintendo development team. Both released at basically the same time. I'd probably give a "pulled from my buttocks" frames per second "average" (i.e. what you'll experience by just playing the game, not stress testing specific areas) on each respective version (Wii U, handheld Switch, docked Switch, later two post patch) at 24 FPS, 29 FPS, 27 FPS.

Now if we just play along and say those are roughly accurate average estimates we can calculate pixel throughput per second, as follows:

Wii U - 720p or 921,600 pixels per frame x 24 frames per second = 22.118M pixels per second.

Switch Handheld - 921,600 pixels per frame x 29 frames per seond = 26.726M pixels per second, a roughly 20% increase in pixel throughput.

Switch Docked - 900p so 1,440,000 pixels per frame x 27 frames per second = 38.88M pixels per second, a roughly 45% increase over the handheld mode, and a 75% increase over the Wii U.

For comparison the PS3 v. PS4 versions of The Last of Us were:
PS3 - 720p (921,600 pixels) per frame x approximately 27 frames per second = 24.883M

PS4 - 1080p (2.1M pixels) per frame x 60 frames per second = 126M

So a roughly 500% increase in pixel throughput from an end of life game ported to new hardware.

*DISCLAIMER* Not all frames are rendered equally. This does not work on a game v. game comparison. These examples used here are to show the difference in performance between the same game from one generation of hardware versus it's next generation release.

So with that out of the way, you can see where a claim of the PS4 being 5 times more powerful than the PS3 is something you can meaningfully back up with math.

The Switch versus the Wii U however shows that the system is much closer to it's predecessor than what we typically associate a generational transition with in the console space.

That is a direct product of the fact that the Switch is not actually the Wii U's direct successor. It is more the successor to the 3DS but design in such a way so as to allow Nintendo to merge their two markets (handheld and console) onto a single platform. Some in this thread have stated that Switch games will perpetually be held back by the handheld mode demand. While technically true, games will be designed based on an acceptable performance threshold in handheld mode, that isn't really holding the games back in my opinion. To me the Switch is much more like buying the latest Nintendo handheld and getting the Switch "Pro" for free via docked mode. This is even more relevant than the PS4 to PS4 Pro or XB1 to Scorpio upgrade as handhelds are traditionally not designed to be capable of outputting a native resolution worthy of being displayed on a TV screen. The Switch solves that out of the box.

I once said on this forum that to me the Wii U wasn't under powered as it delivered all the horsepower needed to deliver Nintendo's artistic vision. Zelda: BotW has made me recant that statement completely. If that is indicative of Nintendo's new level of output going forward I seriously question if the Switch is up to the task. Despite that however it is still the single best hardware offering Nintendo has presented to us since the SNES. No gimmicks like the Wii U tablet or Wii motion controls. Not missing the boat on key hardware features like they did with both the N64 and GC media choices. Not asking contemporary hardware MSRP for clearly aged tech like with the NDS or 3DS.

I loved most of those systems, but the successes were the result of software carrying otherwise questionable hardware across the finish line and the failures were almost always directly tied to their poor hardware choices. The Switch avoids any of the later that I can readily identify and therefore opens the door for the software to carry it.

It is the most powerful gaming dedicated handheld ever made, think of it like that and not as a quarter step into the current generation that it would be as a console, because doing anything else misses the real value in it's design.

Now do Lego City undercover

 

[lwilliams3]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Analogy is probably the best way to understand a system's current horsepower, and ideally with as few additional fixed variables (developers, release window, etc.) as possible.

Thankfully we have a great example with Breath of the Wild. Both made by the same Nintendo development team. Both released at basically the same time. I'd probably give a "pulled from my buttocks" frames per second "average" (i.e. what you'll experience by just playing the game, not stress testing specific areas) on each respective version (Wii U, handheld Switch, docked Switch, later two post patch) at 24 FPS, 29 FPS, 27 FPS.

Now if we just play along and say those are roughly accurate average estimates we can calculate pixel throughput per second, as follows:

Wii U - 720p or 921,600 pixels per frame x 24 frames per second = 22.118M pixels per second.

Switch Handheld - 921,600 pixels per frame x 29 frames per seond = 26.726M pixels per second, a roughly 20% increase in pixel throughput.

Switch Docked - 900p so 1,440,000 pixels per frame x 27 frames per second = 38.88M pixels per second, a roughly 45% increase over the handheld mode, and a 75% increase over the Wii U.

For comparison the PS3 v. PS4 versions of The Last of Us were:
PS3 - 720p (921,600 pixels) per frame x approximately 27 frames per second = 24.883M

PS4 - 1080p (2.1M pixels) per frame x 60 frames per second = 126M

So a roughly 500% increase in pixel throughput from an end of life game ported to new hardware.

*DISCLAIMER* Not all frames are rendered equally. This does not work on a game v. game comparison. These examples used here are to show the difference in performance between the same game from one generation of hardware versus it's next generation release.

So with that out of the way, you can see where a claim of the PS4 being 5 times more powerful than the PS3 is something you can meaningfully back up with math.

The Switch versus the Wii U however shows that the system is much closer to it's predecessor than what we typically associate a generational transition with in the console space.

That is a direct product of the fact that the Switch is not actually the Wii U's direct successor. It is more the successor to the 3DS but design in such a way so as to allow Nintendo to merge their two markets (handheld and console) onto a single platform. Some in this thread have stated that Switch games will perpetually be held back by the handheld mode demand. While technically true, games will be designed based on an acceptable performance threshold in handheld mode, that isn't really holding the games back in my opinion. To me the Switch is much more like buying the latest Nintendo handheld and getting the Switch "Pro" for free via docked mode. This is even more relevant than the PS4 to PS4 Pro or XB1 to Scorpio upgrade as handhelds are traditionally not designed to be capable of outputting a native resolution worthy of being displayed on a TV screen. The Switch solves that out of the box.

I once said on this forum that to me the Wii U wasn't under powered as it delivered all the horsepower needed to deliver Nintendo's artistic vision. Zelda: BotW has made me recant that statement completely. If that is indicative of Nintendo's new level of output going forward I seriously question if the Switch is up to the task. Despite that however it is still the single best hardware offering Nintendo has presented to us since the SNES. No gimmicks like the Wii U tablet or Wii motion controls. Not missing the boat on key hardware features like they did with both the N64 and GC media choices. Not asking contemporary hardware MSRP for clearly aged tech like with the NDS or 3DS.

I loved most of those systems, but the successes were the result of software carrying otherwise questionable hardware across the finish line and the failures were almost always directly tied to their poor hardware choices. The Switch avoids any of the later that I can readily identify and therefore opens the door for the software to carry it.

It is the most powerful gaming dedicated handheld ever made, think of it like that and not as a quarter step into the current generation that it would be as a console, because doing anything else misses the real value in it's design.

First off, nice long post. It was a good read.

There as a few things that I have a different opinion of. In terms of power, the Switch as a portable is like the successor of a theorical successor of the 3DS. The successor of the 3DS would have likely been much closer to the PSVita in power.

As a console, I also wouldn't consider it as a "quarter step into the current generation" due to its graphical features, RAM quantity, and software tools being closer to parity with the other consoles. It just has much less raw power that is still considered remarkable for its form factor.

 

[Nostremitus]

How 'bout the power to move you?

Underappreciated comment.


This is my understanding...

Handheld low power mode still edges out Wii U.

Docked mode is a bit below Xbox One, but has half-precision so could theoretically punch above it's weight in specific situations.
Sort of how the PS4Pro got a 30% performance increase in Mass Effect Andromeda when Bioware switched from fp32 to fp16 for applicable processes.



If this type of increase is common for FP16 use, then the Switch should have little trouble with the majority of games capable of running on stock Xbox One.

 

[Costia]

So.. After a few comments in this thread my revised version looks like this:

XBOX360 - 240 GFLOPS - 48 vector processors * 10FP/cycle per processor * 500Mhz
WIIU - probably 170 GFLOPS - 8 clusters * 20 cores/cluster *550Mhz * 2 (20 cores/cluster seems to be a speculation, might be 32 or a lot less likely 40)
XBOX ONE - 1300 GFLOPS - 768 cores at 853Mhz *2
PS4 - 1800 GFLOPS - 1152 cores *800 Mhz *2

Nintendo switch docked: ~400 GFLOPS - 256 cores * 768Mhz * 2 (there was a rumor of ~900Mhz, but I didn't find a good source, so sticking with 768)
Nintendo switch undocked: ~200 GFLOPS - 256 cores * 384Mhz * 2

So, better than a WiiU when portable, and around 1/3 of xb1 when docked.

Notes:
Removed ps3 because cell is strange, hard to compare directly.
The memory BW of the switch is a lower than the XBONE. ~x3 slower for main memory, +xbone has 32MB of ~x8 faster memory.
XB1 has 8 x86 cores @1.7Ghz (~6 available for games), while Switch has 4 ARM probably at 1Ghz (3 available for games) cores. So probably overall x2-3 times slower?
This is a rough estimate. It doesn't mean that a game that runs on xb1 can run on switch at x0.6 resolution (0.6^2=0.36 -> 1/3 of total pixels).

 

[Nostremitus]

So.. After a few comments in this thread my revised version looks like this:

XBOX360 - 240 GFLOPS - 48 vector processors * 10FP/cycle per processor * 500Mhz
WIIU - probably 170 GFLOPS - 8 clusters * 20 cores/cluster *550Mhz * 2 (20 cores/cluster seems to be a speculation, might be 32 or a lot less likely 40)
XBOX ONE - 1300 GFLOPS - 768 cores at 853Mhz *2
PS4 - 1800 GFLOPS - 1152 cores *800 Mhz *2

Nintendo switch docked: ~400 GFLOPS - 256 cores * 768Mhz * 2 (there was a rumor of ~900Mhz, but I didn't find a good source, so sticking with 768)
Nintendo switch undocked: ~200 GFLOPS - 256 cores * 384Mhz * 2

So, better than a WiiU when portable, and around 1/3 of xb1 when docked.

Notes:
Removed ps3 because cell is strange, hard to compare directly.
The memory BW of the switch is a lower than the XBONE. ~x3 slower for main memory, +xbone has 32MB of ~x8 faster memory.
XB1 has 8 x86 cores @1.7Ghz (~6 available for games), while Switch has 4 ARM probably at 1Ghz (3 available for games) cores. So probably overall x2-3 times slower?
This is a rough estimate. It doesn't mean that a game that runs on xb1 can run on switch at x0.6 resolution (0.6^2=0.36 -> 1/3 of total pixels).

Nvidia FLOPs and AMD FLOPs are not 1:1,

Switch has big.LITTLE 8core Arm (4 A57s and 4 A53s), but we don't know how the LITTLE cores are used since we know it isn't set up to switch between them in an eitherr configuration but OS is a good guess.
But even then, I doubt the OS would need all four of the A53 cores.

 

[Brad Grenz]

Nvidia FLOPs and AMD FLOPs are not 1:1,

Switch has big.LITTLE 8core Arm (4 A57s and 4 A53s), but we don't know how the LITTLE cores are used since we know it isn't set up to switch between them in an eitherr configuration but OS is a good guess.
But even then, I doubt the OS would need all four of the A53 cores.

The big and little cores can't be used simultaneously at all so we know the A53s are basically disabled in Switch completely.

 

[D_prOdigy]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Analogy is probably the best way to understand a system's current horsepower, and ideally with as few additional fixed variables (developers, release window, etc.) as possible.

Thankfully we have a great example with Breath of the Wild. Both made by the same Nintendo development team. Both released at basically the same time. I'd probably give a "pulled from my buttocks" frames per second "average" (i.e. what you'll experience by just playing the game, not stress testing specific areas) on each respective version (Wii U, handheld Switch, docked Switch, later two post patch) at 24 FPS, 29 FPS, 27 FPS.

Now if we just play along and say those are roughly accurate average estimates we can calculate pixel throughput per second, as follows:

Wii U - 720p or 921,600 pixels per frame x 24 frames per second = 22.118M pixels per second.

Switch Handheld - 921,600 pixels per frame x 29 frames per seond = 26.726M pixels per second, a roughly 20% increase in pixel throughput.

Switch Docked - 900p so 1,440,000 pixels per frame x 27 frames per second = 38.88M pixels per second, a roughly 45% increase over the handheld mode, and a 75% increase over the Wii U.

For comparison the PS3 v. PS4 versions of The Last of Us were:
PS3 - 720p (921,600 pixels) per frame x approximately 27 frames per second = 24.883M

PS4 - 1080p (2.1M pixels) per frame x 60 frames per second = 126M

So a roughly 500% increase in pixel throughput from an end of life game ported to new hardware.

*DISCLAIMER* Not all frames are rendered equally. This does not work on a game v. game comparison. These examples used here are to show the difference in performance between the same game from one generation of hardware versus it's next generation release.

So with that out of the way, you can see where a claim of the PS4 being 5 times more powerful than the PS3 is something you can meaningfully back up with math.

The Switch versus the Wii U however shows that the system is much closer to it's predecessor than what we typically associate a generational transition with in the console space.

That is a direct product of the fact that the Switch is not actually the Wii U's direct successor. It is more the successor to the 3DS but design in such a way so as to allow Nintendo to merge their two markets (handheld and console) onto a single platform. Some in this thread have stated that Switch games will perpetually be held back by the handheld mode demand. While technically true, games will be designed based on an acceptable performance threshold in handheld mode, that isn't really holding the games back in my opinion. To me the Switch is much more like buying the latest Nintendo handheld and getting the Switch "Pro" for free via docked mode. This is even more relevant than the PS4 to PS4 Pro or XB1 to Scorpio upgrade as handhelds are traditionally not designed to be capable of outputting a native resolution worthy of being displayed on a TV screen. The Switch solves that out of the box.

I once said on this forum that to me the Wii U wasn't under powered as it delivered all the horsepower needed to deliver Nintendo's artistic vision. Zelda: BotW has made me recant that statement completely. If that is indicative of Nintendo's new level of output going forward I seriously question if the Switch is up to the task. Despite that however it is still the single best hardware offering Nintendo has presented to us since the SNES. No gimmicks like the Wii U tablet or Wii motion controls. Not missing the boat on key hardware features like they did with both the N64 and GC media choices. Not asking contemporary hardware MSRP for clearly aged tech like with the NDS or 3DS.

I loved most of those systems, but the successes were the result of software carrying otherwise questionable hardware across the finish line and the failures were almost always directly tied to their poor hardware choices. The Switch avoids any of the later that I can readily identify and therefore opens the door for the software to carry it.

It is the most powerful gaming dedicated handheld ever made, think of it like that and not as a quarter step into the current generation that it would be as a console, because doing anything else misses the real value in it's design.

make a DBZ joke or stop wasting everyone's time

Spoiler

great post, thanks

 

[llien]

Nvidia FLOPs and AMD FLOPs are not 1:1.

I'm afraid people don't realize what stands behind that statement.
That hardly applies to consoles, on top of there being nothing "unbalanced" about 7850/7870.

I recall 2 PS4 games are dominating the recent "amazing screenshot" thread. Not even PS4 Pro.
That's roughly 750Ti levels.

 

[xochipiltontli]

I really hope Nintendo does something about Splatoon 2's resolution. The fact that the game was subres in undocked mode and 720p with no AA when docked is really sad and makes no sense when other Switch games seem to suggest that the system's stronger than Wii U undocked.

 

[killatopak]

Not enough for a console, good enough for a handheld.

 

[Oblivion]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Using the PS2 as an example of inferior hardware outputting comparable results with its rival consoles is completely inappropriate for this specific discussion. Yes, PS2 had the weakest of the big 3's hardware, but at its absolute worst, it was never a generational gap between it and either the GC or the Xbox. Nobody was ever going to confuse a competently developed PS2 game for a PS1 game. Hell, even in a few, limited aspects like fill rate, I believe PS2 actually had the edge over the other two.

 

[Nostremitus]

The big and little cores can't be used simultaneously at all so we know the A53s are basically disabled in Switch completely.

I used to think that too, but it hasn't been true since 2013.
Phones just don't use them simultaneously because of their use case. There's really nothing stopping Nintendo from using a scheduler that makes use of the cores.

In the earlier big.LITTLE software models (cluster migration and CPU migration), the software switched between cores and could not switch all cores on simultaneously. In the more recent software model, Global Task Scheduling, software can enable all cores to be active at once because the OS is aware of the big and LITTLE cores in the system and is in direct control of thread allocation among the available cores. With Global Task Scheduling the OS power management mechanisms will continue to idle unused cores in the same way it does in standard multi-core systems today.

-Brian Jeff (Product Marketing Director at ARM


https://community.arm.com/processors/b/blog/posts/ten-things-to-know-about-big-little
From what I understand, it all depends on the OS scheduler, Android's scheduler is only either/or.
The Switch being a closed system games console with a custom OS means it has a custom scheduler, it's not a phone. The only reason we only ever discuss the big cores on Switch is because those are the only ones Nintendo mentioned devs have access to in the documentation. We really have no clue what's going on with the little ones.
They could be disabled, but there's no real reason for them to be or for the silicon to be wasted.

 

[Costia]

They could be disabled, but there's no real reason for them to be or for the silicone to be wasted.

Thermal is a reason for them to be disabled.
The silicon is "wasted" to save battery.

 

[Nostremitus]

Thermal is a reason for them to be disabled.
The silicon is "wasted" to save battery.

A53 cores though? They don't use or produce much heat. That's their purpose, it's the reason for the LITTLE part of big.LITTLE. They exist to be low power, low heat chips to use instead of the big ones for tasks that don't need the hungrier cores.

Makes even less sense for Docked mode.

 

[Costia]

A53 cores though? They don't use or produce much heat. That's their purpose, it's the reason for the LITTLE part of big.LITTLE. They exist to be low power, low heat chips to use instead of the big ones for tasks that don't need the hungrier cores.
Makes even less sense for Docked mode.

http://www.anandtech.com/show/8718/the-samsung-galaxy-note-4-exynos-review/4
Looks like it can get to 0.8W. For smartphones the entire thermal budget is ~5W. So if you want to run the big cores and the GPU, you will probably get worse performance per watt with the a53s enabled (and worse overall, since the bottleneck will be heat dissipation).
I guess you can turn them on when docked.

Edit: looks like the CPU and the GPU on the switch are underclocked, even in docked mode. One possible explanation is that they are already hitting thermal limits at those freq.

 

[Calm Mind]

Convenience

 

[Nostremitus]

http://www.anandtech.com/show/8718/the-samsung-galaxy-note-4-exynos-review/4
Looks like it can get to 0.8W. For smartphones the entire thermal budget is ~5W. So if you want to run the big cores and the GPU, you will probably get worse performance per watt with the a53s enabled (and worse overall, since the bottleneck will be heat dissipation).
I guess you can turn them on when docked.

Just north of 0.5W at Switch's CPU clock speed. Hell, keeping the A53's on could actually explain why Nintendo clocked the Switch lower than the Nvidia Shield or Pixel C but still needs active cooling.

 

[sneas78]

I think we will know what the Switch is capable of once we see what the Switch is capable of.

 

[Costia]

Just north of 0.5W at Switch's CPU clock speed. Hell, keeping the A53's on could actually explain why Nintendo clocked the Switch lower than the Nvidia Shield or Pixel C but still needs active cooling.

Pixel C probably wont be able to run at full speed for a long time (according to wikipedia ,X1 TDP is 15w) , the switch has to.
You could underclock the a57 to turn on the a53. You will get more overall performance at the expense of single threaded performance. But running at 1GHz instead of 1.9Ghz - doesn't look like it would have been worth it if that's the reason, especially that single threaded performance seems to be important to games.

Edit: anyway, my point is that the performance is probably limited by thermal and battery life, so i wouldn't expect any extra power to be unlocked later on. Except maybe the OS reserved core, but i highly doubt it, there is no kinect to turn off on the switch.

 

[PartTimeWarrior]

It's the most powerful system ever made. So powerful that third party developers are too afraid to make games for it

 

[Nostremitus]

Pixel C probably wont be able to run at full speed for a long time (according to wikipedia ,X1 TDP is 15w) , the switch has to.
You could underclock the a57 to turn on the a53. You will get more overall performance at the expense of single threaded performance. But running at 1GHz instead of 1.9Ghz - doesn't look like it would have been worth it if that's the reason, especially that single threaded performance seems to be important to games.

Unless you're running the OS on the A53s so devs can have access to all four A57s.
The A57s wouldn't be at 1.9 either way, though, to prevent thermal throttling.

 

[Toadthemushroom]

A53 cores though? They don't use or produce much heat. That's their purpose, it's the reason for the LITTLE part of big.LITTLE. They exist to be low power, low heat chips to use instead of the big ones for tasks that don't need the hungrier cores.

Makes even less sense for Docked mode.

They still haven't enabled the home button light on the Joy-Con R (which would be visible while Switch is docked), so I'm guessing they have so plans to use the A53s for stuff that the 3DS's co-processors did.

 

[GinoFelino]

The actual power of the Switch is its portability

Also, having a ton of Nintendo games and awesome Japanese ones

 

[Spukc]

About two fiddy

 

[ShadowFox08]

I really hope Nintendo does something about Splatoon 2's resolution. The fact that the game was subres in undocked mode and 720p with no AA when docked is really sad and makes no sense when other Switch games seem to suggest that the system's stronger than Wii U undocked.

You do realize that resolution can be sacrificed for graphical fidelity(better textures, shadows, higher poly count, lighting, draw distance, rtc) or a smoother performance(in fps), right? That's what we got from snake pass and Minecraft. Not still how well optimized Splatoon 2 is though, but my point still stands. I wouldn't mind devs doing 720p docked and 540p handheld for games like cod...

 

[Brofist]

I've heard it's like Gohan.

Came in for the DB power level analogies. Not disappointed.

 

[requiem02]

How 'bout the power to move you?

I haven't read through the whole thread, but I'd just like to give this Tenacious D reference the love it deserves. Props.

 

[Conan-san]

Well given it runs tetris without imploding I imagine it's like two PS4s taped together.

 

[ramparter]

I've heard it's like Gohan.

But Gohan is the goat db character.

 

[ffdgh]

But Gohan is the goat db character.

Kid Gohan or adult Gohan? There's a difference you know.

 

[Calm Mind]

Kid Gohan or adult Gohan? There's a difference you know.

Kid Gohan most definitely. Adult Gohan is the Wii U.

 

[Kurt]

The thing is. Most people here talking in terms of specs instead of visual. If we would compare scorpion with xbox 1, its 6times the power or so. Yet the visuals are only a little improvement. Even the games between ps4 and ps4 pro are like minor differents. The same thing between xbox one and switch. Yeah there are differents, but just minor things.

If you are really care anout 720p vs 1080p and minor things like that, than you should buy a pc instead...

 

[Costia]

The thing is. Most people here talking in terms of specs instead of visual. If we would compare scorpion with xbox 1, its 6times the power or so. Yet the visuals are only a little improvement. Even the games between ps4 and ps4 pro are like minor differents. The same thing between xbox one and switch. Yeah there are differents, but just minor things.
If you are really care anout 720p vs 1080p and minor things like that, than you should buy a pc instead...

If you are expecting XB1 games on switch at a lower resolution/less effects/etc you will be disappointed.
XB1 was the base minimum spec those games were made for.
So while scaling up it goes from dynamic resolution @30fps to 4k@30, scaling down it will go from dynamic resolution@30 to dynamic resolution@10fps - not playable.
You should be looking at the CPU as well. XB1->scorpio isn't x6, its ~x1.3. While xb1 vs switch is probably ~x3.
So no, the difference between XB1 and switch is not "minor things".

 

[oti]

It's an infinite source of salt.

 

[opricnik]

Based off of ports we've seen already on Switch VS PS4: Theoretically one scenario could ideally look like this(in terms of displaying power differences) for a AAA third party multiplat ports, if devs optimized the Switch version:

PS4: 1080p, 60fps
Xbone: 900p, 60fps
Switch: 720p, 60fps with nearly identical graphical fidelity but some minor downgrades to any combination of lighting, shadows, polygons, anti-aliasing, particle effects and/or textures vs Xbone and PS4

This is just one combination. They could have the same resolution, but half the frame rate for some games as well as a downgrade in some other effects(if devs choose to do so) also.

I don't know about demanding AAA 3rd party games like call of duty though, especially single player. I haven't taken the bandwidth to account, since that is Switch's biggest weakpoint. But I'm really really curious to see how it would work. All of this is just my hunch and based off of empirical evidence from what we've seen so far. So I'm thinking PS4 is at least 3x as powerful as Switch, and Xbone is 2x as powerful, after you count all the under the hood optimizations from working with a much newer and more efficient nvidia hardware vs older AMD architecture+mixed precision mode(not saying its a god send or anything) on unreal 4 engines.




ehhh current gohan was kind of implied to be on par with SSB4 Goku from the first 3 punches at least. PS4 Pro being goku with kaioken x10 is fairly accurate in GPU performance at least. lol



If there's anything I learned from the wii generation era for third party ports on wii, anything is possible. The Wii is easily like 20x weaker than the 360 in GPU power, but we managed to get really good ports of the call of duty games, despite the resolution downgrade, frame rate cute in half(to 30fps), and graphic fidelity of course(polygons, textures, lighting, shadows, particle effects, etc). The core gameplay was identical the 360/ps3 to a T minus a few missing features and the cut frame rate.

So that 360 game ported from a PS4 game you're talking about is definitely doable, but it will be scaled down significantly. It's more of a matter of if devs would make profit from the game in the first place to have it ported if anything.

Thing is PS4 barely 900p these days and xbox one at 720p.
so what will be switch 576p docked 360p not docked?

 

[opricnik]

The thing is. Most people here talking in terms of specs instead of visual. If we would compare scorpion with xbox 1, its 6times the power or so. Yet the visuals are only a little improvement. Even the games between ps4 and ps4 pro are like minor differents. The same thing between xbox one and switch. Yeah there are differents, but just minor things.

If you are really care anout 720p vs 1080p and minor things like that, than you should buy a pc instead...

lol

"minor"

nintendo fans live in their bubble right.

"little" , " minor" right...

 

[blu]

I used to think that too, but it hasn't been true since 2013.
Phones just don't use them simultaneously because of their use case. There's really nothing stopping Nintendo from using a scheduler that makes use of the cores.


-Brian Jeff (Product Marketing Director at ARM


https://community.arm.com/processors/b/blog/posts/ten-things-to-know-about-big-little
From what I understand, it all depends on the OS scheduler, Android's scheduler is only either/or.
The Switch being a closed system games console with a custom OS means it has a custom scheduler, it's not a phone. The only reason we only ever discuss the big cores on Switch is because those are the only ones Nintendo mentioned devs have access to in the documentation. We really have no clue what's going on with the little ones.
They could be disabled, but there's no real reason for them to be or for the silicon to be wasted.

TX1 can't do HMP - there's a hardware (CCI) limitation, ergo the LITTLE cores are useless in the switch, unless something can run entirely off LITTLE cores.

 

[Skittzo0413]

Just north of 0.5W at Switch's CPU clock speed. Hell, keeping the A53's on could actually explain why Nintendo clocked the Switch lower than the Nvidia Shield or Pixel C but still needs active cooling.

User MDave here actually discovered the Shield TV throttles to similar levels as the Switch, though usually it throttles either the CPU or GPU and not both at once to the same levels, though I think it can do that. Pixel C also throttles quite a bit.

 

[Nostremitus]

TX1 can't do HMP - there's a hardware (CCI) limitation, ergo the LITTLE cores are useless in the switch, unless something can run entirely off LITTLE cores.

Oh, well there you go. What makes the X1 different from other big.LITTLE ARM SOCs in that regard? Could the "custom"

Spoiler

lol

X1 in Switch have been modified to change this? Or has it been confirmed that all of the engineering hours Nvidia put into the SOC were strictly software related?

 

[LinkSlayer64]

Raw numbers have lied about real performance since the 8-bit days, so that won't do you any good.

It also is not a constant in terms of the output we see. The PS2 is an excellent example of how inferior hardware can produce comparable results due to far more development priority. Developers really worked to maximize everything the PS2 could do and found a lot of advanced tricks to that end. Meanwhile the Vita and Wii U are likely two systems that punched well below their actual weight due to lack of life cycle and developer interest.

Analogy is probably the best way to understand a system's current horsepower, and ideally with as few additional fixed variables (developers, release window, etc.) as possible.

Thankfully we have a great example with Breath of the Wild. Both made by the same Nintendo development team. Both released at basically the same time. I'd probably give a "pulled from my buttocks" frames per second "average" (i.e. what you'll experience by just playing the game, not stress testing specific areas) on each respective version (Wii U, handheld Switch, docked Switch, later two post patch) at 24 FPS, 29 FPS, 27 FPS.

Now if we just play along and say those are roughly accurate average estimates we can calculate pixel throughput per second, as follows:

Wii U - 720p or 921,600 pixels per frame x 24 frames per second = 22.118M pixels per second.

Switch Handheld - 921,600 pixels per frame x 29 frames per seond = 26.726M pixels per second, a roughly 20% increase in pixel throughput.

Switch Docked - 900p so 1,440,000 pixels per frame x 27 frames per second = 38.88M pixels per second, a roughly 45% increase over the handheld mode, and a 75% increase over the Wii U.

For comparison the PS3 v. PS4 versions of The Last of Us were:
PS3 - 720p (921,600 pixels) per frame x approximately 27 frames per second = 24.883M

PS4 - 1080p (2.1M pixels) per frame x 60 frames per second = 126M

So a roughly 500% increase in pixel throughput from an end of life game ported to new hardware.

*DISCLAIMER* Not all frames are rendered equally. This does not work on a game v. game comparison. These examples used here are to show the difference in performance between the same game from one generation of hardware versus it's next generation release.

So with that out of the way, you can see where a claim of the PS4 being 5 times more powerful than the PS3 is something you can meaningfully back up with math.

The Switch versus the Wii U however shows that the system is much closer to it's predecessor than what we typically associate a generational transition with in the console space.

That is a direct product of the fact that the Switch is not actually the Wii U's direct successor. It is more the successor to the 3DS but design in such a way so as to allow Nintendo to merge their two markets (handheld and console) onto a single platform. Some in this thread have stated that Switch games will perpetually be held back by the handheld mode demand. While technically true, games will be designed based on an acceptable performance threshold in handheld mode, that isn't really holding the games back in my opinion. To me the Switch is much more like buying the latest Nintendo handheld and getting the Switch "Pro" for free via docked mode. This is even more relevant than the PS4 to PS4 Pro or XB1 to Scorpio upgrade as handhelds are traditionally not designed to be capable of outputting a native resolution worthy of being displayed on a TV screen. The Switch solves that out of the box.

I once said on this forum that to me the Wii U wasn't under powered as it delivered all the horsepower needed to deliver Nintendo's artistic vision. Zelda: BotW has made me recant that statement completely. If that is indicative of Nintendo's new level of output going forward I seriously question if the Switch is up to the task. Despite that however it is still the single best hardware offering Nintendo has presented to us since the SNES. No gimmicks like the Wii U tablet or Wii motion controls. Not missing the boat on key hardware features like they did with both the N64 and GC media choices. Not asking contemporary hardware MSRP for clearly aged tech like with the NDS or 3DS.

I loved most of those systems, but the successes were the result of software carrying otherwise questionable hardware across the finish line and the failures were almost always directly tied to their poor hardware choices. The Switch avoids any of the later that I can readily identify and therefore opens the door for the software to carry it.

It is the most powerful gaming dedicated handheld ever made, think of it like that and not as a quarter step into the current generation that it would be as a console, because doing anything else misses the real value in it's design.

Why not just use Mario Kart 8DX?
59fps 720p -> 60fps 1080p

 

[mclem]

I also desperately need this answer.

Can someone add Switch to this image so that I can understand?

Fundamentally incompatible.

Spoiler

The switch doesn't use Cell

 

[Nostremitus]

.

Spoiler

The switch doesn't use Cell

Neither does that image.

Edit: Damn, top of the page...