Does this kill the chances of retail units being Pascal-based?
There were rumors that production had already started on retail units.
Me: "Switch will be a Portable Wii U, graphics wise. Readjust your expectations."
Y'all: "BS! It'll cost Nintendo MORE to order old chips and make nvidia produce outdated processors than to just use Pascal! Some Pascal chips get up to [number] gigaflopz so the Switch will be current-gen quality at least when docked! Pascal has already been out for 6 months, totally enough time for implementation! BotW is a Wii U port so of course it won't accurately represent the Switch's power!"
EVERY Nintendo console launch...
I have a few immediate thoughts after reading through the article:
I see a few different scenarios here:
- Firstly, it's worth noting the difference between Maxwell and Pascal is almost entirely down to the manufacturing process. Maxwell was made on 28nm (and in the case of the TX1, 20nm) whereas Pascal is made on 16nm. The actual architectural difference between the two is minimal, and aside from improved color buffer compression, largely irrelevant for a device like the Switch.
- Despite that, the article never makes any mention of the manufacturing process. I find that extremely strange, as it's obviously the defining difference between the two sets of GPUs.
- In fact, the article gets the difference between the two completely the wrong way around, saying "Nintendos box is relatively small, and so it has to fit into the heat profile of a portable device, rather than a set-top box. Thats another reason that explains the older Maxwell technology, as opposed to the Pascals state-of-the-art tech." Pascal is literally a more power efficient version of Maxwell, so the incentive would be the other way around.
- The author says "we expect the Nintendo Switch to be more than 1 teraflop in performance", which is notably higher than even those of us who were expecting Pascal were considering (I literally posted earlier today with a 500-750 Gflop estimate). If this is a Maxwell chip, then that would mean at least 4 SMs (512 "CUDA cores") at 1GHz, as they're not going to be able to push much past that on 28/20nm. This is a much larger GPU than most people would have been expecting.
Basically, if you're to take the article as being accurate, then the only worthwhile takeaway is this quote:
- The Switch SoC uses Maxwell at 20nm, and simply has a much larger GPU than anticipated to account for the performance.
- Nintendo looked at the feature-set planned for Pascal when design started, realised that the new features were largely irrelevant, and decided that they would save time and just use a straight-forward die shrink of Maxwell to 16nm instead. That would technically be a Maxwell GPU, but would be almost completely indistinguishable from Pascal in terms of performance.
- The sources are wrong about Maxwell, the 1 Tflop performance, or both.
A Maxwell Tflop is identical to a Pascal Tflop, and it's largely irrelevant to us whether they achieved that by using a larger Maxwell GPU on 20nm/28nm at a lower clock or a smaller Pascal GPU on 16nm at a higher clock.
It's unclear, what would be interesting to know is if they shrunk maxwell (x1) to 16nm for the 3d transistors, not sure there is really a difference if they did, as maxwell on a newer process could probably be just as stable at 1.5ghz, of course even if pascal is used, it might not be clocked at 1.5ghz when docked anyways, so there isn't much to worry about except battery life.
Hahaha they edited/changed their article.
If there are actually two levels of graphical power depending on if it's docked or not, I'd get the feeling that it'll barely reach Wii U levels when running off battery. From a $249/$299 perspective, that feels accurate to me knowing Nintendo's history with hardware, and I still think it's a great feat for them! The Switch will almost exclusively be a portable for me and my situation and I ain't complaining. Gimme dat Splatoon port and I'm gold.Last I remember, you said it was going to be worse than Wii U.
Considering x1 and pascal are very similar, with the difference being entirely the process node AFAIK, I am not sure you could call it significant. It will be interesting if this is 16nm and they still call it maxwell, it could be 20nm but I don't see how they can only produce 2m by launch if that is the case.The Dev Kits and the Retail Kits are not built by the same people. They wouldn't absolutely have to use the same hardware, though I'd wager it unlikely they would differ significantly.
This means that the Switch doesnt have as much visual horsepower as the PS4 when played on a television, and it may not be able to handle 4K graphics, either. If Nintendo had waited for Pascal, it would have had to push back the launch date of the Switch. Were not so sure if the Switch is weaker than the Xbox One, as the performance may be close.
If there are actually two levels of graphical power depending on if it's docked or not, I'd get the feeling that it'll barely reach Wii U levels when running off battery. From a $249/$299 perspective, that feels accurate to me knowing Nintendo's history with hardware, and I still think it's a great feat for them! The Switch will almost exclusively be a portable for me and my situation and I ain't complaining. Gimme dat Splatoon port and I'm gold.
Well they seem to be able to cram an Xbox One into a Switch, so...I love how the article goes full on concern mode with the whole "It may not be able to output 4k graphics" stuff.
I mean really? REALLY? People were expecting it to do 4k? In that slim, tablet form factor? Personally I'm amazed people even thought it would be as powerful as a PS4. Look at the size of the PS4 slim. Then look at the size of the Switch. Cramming all that into a tiny tablet? People have been at the paint fumes again.
Isn't there a power difference between maxwell and pascal tegra? With the shrink nvidia could have put more transistors in for the same overall power draw as maxwell - or did they just focus on power efficiency with it and keep similar power to maxwell?
Isn't there a power difference between maxwell and pascal tegra? With the shrink nvidia could have put more transistors in for the same overall power draw as maxwell - or did they just focus on power efficiency with it and keep similar power to maxwell?
Pascal uses less shaders at higher frequencies to beat Maxwell. The GTX 980 uses 2048 shaders (16SM) @ just over 1.1ghz to hit 4.6tflops. GTX 1060 uses 1280 shaders @ ~1.7ghz for 4.3 tflops. It does all this on less power (120watts vs 165watts) otherwise the architecture is pretty much identical from a flop to flop perspective.Isn't there a power difference between maxwell and pascal tegra? With the shrink nvidia could have put more transistors in for the same overall power draw as maxwell - or did they just focus on power efficiency with it and keep similar power to maxwell?
So you're going against people who actually know about tech based on... feeling.
nice
I always see a lot of "here's what's technically possible" but not a lot of "here's what's technically possible for a probably-$250 console made by a historically-thrifty company that just spent a crapload on R&D and doesn't care about the race the other two consoles are constantly running."
In the end we have: leaks saying "2-3x Wii U power," leaks that say "pretty much a Wii U," gameplay from a few first-party games that look Wii U-ish in graphical capability, and Nintendo's history with hardware. I think that makes it clear where to safely set expectations.
There has not been a single leak that says pretty much a Wii U. I mean for all the shit this thread is creating they are saying around Xbox One level power.
It's funny how some take the Maxwell part of it's as truth while skipping the ~Xbox One perf part.
I always see a lot of "here's what's technically possible" but not a lot of "here's what's technically possible for a probably-$250 console made by a historically-thrifty company that just spent a crapload on R&D and doesn't care about the race the other two consoles are constantly running."
In the end we have: leaks saying "2-3x Wii U power," leaks that say "pretty much a Wii U," gameplay from a few first-party games that look Wii U-ish in graphical capability, and Nintendo's history with hardware. I think that makes it clear where to safely set expectations.
Seems to be besides the point. Strange garbled news perceived as bad news must mean the worst news and folk wisdom upheld! There've been too many good Switch stories lately. I was beginning to feel something must be creeping up myself, being an anxious Nintendo fan, but this isn't really much of anything without a fuller, clearer picture.There has not been a single leak that says pretty much a Wii U. I mean for all the shit this thread is creating they are saying around Xbox One level power.
Tegra version of Pascal also has proper support for FP16.Pascal uses less shaders at higher frequencies to beat Maxwell. The GTX 980 uses 2048 shaders (16SM) @ just over 1.1ghz to hit 4.6tflops. GTX 1060 uses 1280 shaders @ ~1.7ghz for 4.3 tflops. It does all this on less power (120watts vs 165watts) otherwise the architecture is pretty much identical from a flop to flop perspective.
I think it's safe to say that a sizable percentage of GAF users only read thread titles when posting.
Tegra version of Pascal also has proper support for FP16.
I have a few immediate thoughts after reading through the article:
I see a few different scenarios here:
- Firstly, it's worth noting the difference between Maxwell and Pascal is almost entirely down to the manufacturing process. Maxwell was made on 28nm (and in the case of the TX1, 20nm) whereas Pascal is made on 16nm. The actual architectural difference between the two is minimal, and aside from improved color buffer compression, largely irrelevant for a device like the Switch.
- Despite that, the article never makes any mention of the manufacturing process. I find that extremely strange, as it's obviously the defining difference between the two sets of GPUs.
- In fact, the article gets the difference between the two completely the wrong way around, saying "Nintendos box is relatively small, and so it has to fit into the heat profile of a portable device, rather than a set-top box. Thats another reason that explains the older Maxwell technology, as opposed to the Pascals state-of-the-art tech." Pascal is literally a more power efficient version of Maxwell, so the incentive would be the other way around.
- The author says "we expect the Nintendo Switch to be more than 1 teraflop in performance", which is notably higher than even those of us who were expecting Pascal were considering (I literally posted earlier today with a 500-750 Gflop estimate). If this is a Maxwell chip, then that would mean at least 4 SMs (512 "CUDA cores") at 1GHz, as they're not going to be able to push much past that on 28/20nm. This is a much larger GPU than most people would have been expecting.
Basically, if you're to take the article as being accurate, then the only worthwhile takeaway is this quote:
- The Switch SoC uses Maxwell at 20nm, and simply has a much larger GPU than anticipated to account for the performance.
- Nintendo looked at the feature-set planned for Pascal when design started, realised that the new features were largely irrelevant, and decided that they would save time and just use a straight-forward die shrink of Maxwell to 16nm instead. That would technically be a Maxwell GPU, but would be almost completely indistinguishable from Pascal in terms of performance.
- The sources are wrong about Maxwell, the 1 Tflop performance, or both.
A Maxwell Tflop is identical to a Pascal Tflop, and it's largely irrelevant to us whether they achieved that by using a larger Maxwell GPU on 20nm/28nm at a lower clock or a smaller Pascal GPU on 16nm at a higher clock.
Wow.The Tegra X1 (Maxwell) was the first Nvidia Chip to support the new FP16 setup. Even if Nintendo were using a chip derived from the X1, they would still see FP16 benefits.
Okay, why exactly this thread has 19 pages ?
Because people want, expect, or fear that a certain vision of Nintendo be realized.So if this article is true (and OMG how badly written it is) the switch is more powerful than a lot of people here expected. I mean, the consensus was clearly below XboxOne.
But why all the morons saying "lol nintendo being cheap", "lol every nintendo new console", "they're doomed lmfao"?
I have a few immediate thoughts after reading through the article:
I see a few different scenarios here:
- Firstly, it's worth noting the difference between Maxwell and Pascal is almost entirely down to the manufacturing process. Maxwell was made on 28nm (and in the case of the TX1, 20nm) whereas Pascal is made on 16nm. The actual architectural difference between the two is minimal, and aside from improved color buffer compression, largely irrelevant for a device like the Switch.
- Despite that, the article never makes any mention of the manufacturing process. I find that extremely strange, as it's obviously the defining difference between the two sets of GPUs.
- In fact, the article gets the difference between the two completely the wrong way around, saying "Nintendos box is relatively small, and so it has to fit into the heat profile of a portable device, rather than a set-top box. Thats another reason that explains the older Maxwell technology, as opposed to the Pascals state-of-the-art tech." Pascal is literally a more power efficient version of Maxwell, so the incentive would be the other way around.
- The author says "we expect the Nintendo Switch to be more than 1 teraflop in performance", which is notably higher than even those of us who were expecting Pascal were considering (I literally posted earlier today with a 500-750 Gflop estimate). If this is a Maxwell chip, then that would mean at least 4 SMs (512 "CUDA cores") at 1GHz, as they're not going to be able to push much past that on 28/20nm. This is a much larger GPU than most people would have been expecting.
Basically, if you're to take the article as being accurate, then the only worthwhile takeaway is this quote:
- The Switch SoC uses Maxwell at 20nm, and simply has a much larger GPU than anticipated to account for the performance.
- Nintendo looked at the feature-set planned for Pascal when design started, realised that the new features were largely irrelevant, and decided that they would save time and just use a straight-forward die shrink of Maxwell to 16nm instead. That would technically be a Maxwell GPU, but would be almost completely indistinguishable from Pascal in terms of performance.
- The sources are wrong about Maxwell, the 1 Tflop performance, or both.
A Maxwell Tflop is identical to a Pascal Tflop, and it's largely irrelevant to us whether they achieved that by using a larger Maxwell GPU on 20nm/28nm at a lower clock or a smaller Pascal GPU on 16nm at a higher clock.
Because people want, expect, or fear that a certain vision of Nintendo be realized.
Is it? are people expecting better triple a third party support this time?
Me: "Switch will be a Portable Wii U, graphics wise. Readjust your expectations."
Y'all: "BS! It'll cost Nintendo MORE to order old chips and make nvidia produce outdated processors than to just use Pascal! Some Pascal chips get up to [number] gigaflopz so the Switch will be current-gen quality at least when docked! Pascal has already been out for 6 months, totally enough time for implementation! BotW is a Wii U port so of course it won't accurately represent the Switch's power!"
EVERY Nintendo console launch...
A huge percentage of GAF useres wants to see Nintendo go third-party...
Its kinda of pointless debating about specs right now since we are at a month before the big announcement. They question will ultimately be "who will be on board"? We have names (of companies) but no titles to announce (as of yet)...well, several have already gave us the slip, and the fact that most of the rumors have been positive so far speaks good things.
Me: "Switch will be a Portable Wii U, graphics wise. Readjust your expectations."
Y'all: "BS! It'll cost Nintendo MORE to order old chips and make nvidia produce outdated processors than to just use Pascal! Some Pascal chips get up to [number] gigaflopz so the Switch will be current-gen quality at least when docked! Pascal has already been out for 6 months, totally enough time for implementation! BotW is a Wii U port so of course it won't accurately represent the Switch's power!"
EVERY Nintendo console launch...
So not only do these idiots somehow said that Pascal would be hotter and larger than previous chips (except the opposite is true), they also said the Switch is going to be at least 1TF anyway?
Seriously, this is idiotic. I think we can safely throw this rumor into the trash pile.
This should be in the OP to make the first page even more embarrassing.
A huge percentage of GAF useres wants to see Nintendo go third-party...
Its kinda of pointless debating about specs right now since we are at a month before the big announcement. They question will ultimately be "who will be on board"? We have names (of companies) but no titles to announce (as of yet)...well, several have already gave us the slip, and the fact that most of the rumors have been positive so far speaks good things.