• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

What we know so far about the Nintendo NX with sources 2.0

Status
Not open for further replies.

udivision

Member
Expect a nintendo system and you won't be let down. Specifics don't matter. The quality of games brought to the system is all that matters in the end. Whether those games are valuable enough for you to buy the system is the core of all of this. Do graphics matter? To some. Do 3rd party matter? To some. Value of the system is all that matters and raw numbers don't mean jack (unless you think good numbers = better games in the future - which isn't a guarantee. Greatness awaits, remember?).

To be fair, "Will the next Nintendo System have good Nintendo Games?" isn't all that interesting to discuss, especially without any specific games. The games we like being there is a given.

The reason why people are interested in specs, "gimmicks", third parties and positioning is because those things are up in the air. Can Nintendo continue to position themselves the way that have in the past two generations? What would they need to attract a specific audience. Is there evidence of them targetting a new market? Are they in talks with third parties? How powerful would their console need to be to compete with the PS4 and Xbox, what does that even mean, and would they even want to? There is potential for the deciding factor for the NX purchase to be more than just the value of the first party.

My purchase of the NX doesn't hinge on any of questions that I just mentioned, but they can provide a decent enough back-and-forth that keep these threads going.
 

Proelite

Member
Figured I'll put in my pessimistic, realistic, and optimistic NX specs again. Created with the help of Wiki, Google, Thaktor, Blu, and my own experiences with Xbox one and PS4 leaks.

And logic.

Pessimistic.
8 core PUMA 1.6 - 2.0 ghz.
8-10 CU GCN 1.2 GPU at 800 - 1000 mgz. 1 teraflop.
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf close to slightly under Xbone. Might have trouble with high bandwidth scenarios such as alpha effects, deferred rendering, etc. Cost of BOM is REALLY low though. Can retailed for $199 and still make a profit. Or pack in an expensive gimmick and retail for $299.

Realistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
12-16 CU Polaris GPU at 1 ghz. 1.6 to 2.0 teraflop.
64 mb of esram. (Takes up less die space than 32mb of esram on 28nm)
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.

Optimistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
20 - 24 CU Polaris GPU at 1.0 ghz. 2.5 - 3 teraflops.
8GB of GDDR5X or combination of LPDDR4 + HBM2.

Real world per 2x ps4.
Break even / slight profit at $399 retail.
 

ozfunghi

Member
Isn't it better that we have someone to shoot down the nonsense rather than letting speculation run wild?

What if the creator of the controller leak never came out and admitted it? Just imagine how things would be, lol.

Every single NX thread would be a lot worse than they already are.

I'm only speaking for myself, but if she actually DOES know this tech stuff, she should be more clear. She is using twitter to form an opinion on a discussion on Gaf, instead of partaking in the discussion herself. That wouldn't be a problem in itself if we could ask her what exactly she means by it, or how we should interpret that info. It's a bit too much hit&run for my liking. She says the story about the AMD mix up is bollocks for a number of reasons, one of which being the fact that both chips are not collocated with each other. For instance, is she saying that the NX chip will be on a different process node than the XBOII chip? Which would indeed mean the NX chip might very well be on a 28nm process. Or maybe she's debunking the existence of an XBOII alltogether (if there is no XBOII, then there is no XBOII chip, and there can't be a mixup, and the chips can't be collocated either... still leaving any process node on the table for NX).

I mean, to me it feels like she doesn't fully understand the implications her statements bring to the table, which in turn, has me wondering if she understands the nature of the subject, and in turn whether or not she is misinterpreting stuff she doesn't fully understand.
 

MuchoMalo

Banned
Jeez. What did she do to you to make you this upset with her? She's letting people know that they should keep their expectations in check - which everyone should. You (not specifically you, i mean the internet at large) don't want to be disappointed because nintendo didn't meet your grandiose unrealistic expectation of them, as I'm sure she's been reading everywhere on the internet. She may not know the specifics but she also knows it's likely not going to be some insane supercomputer like some people might be expecting.

Expect a nintendo system and you won't be let down. Specifics don't matter. The quality of games brought to the system is all that matters in the end. Whether those games are valuable enough for you to buy the system is the core of all of this. Do graphics matter? To some. Do 3rd party matter? To some. Value of the system is all that matters and raw numbers don't mean jack (unless you think good numbers = better games in the future - which isn't a guarantee. Greatness awaits, remember?).

That's the thing, though. She's made her own decisions about what "in check" and "Iwata's philosophy" mean, and it's resulted in huge contradictions that can only be solved by being less vague. For example, she's decided that an Iwata console obviously would not use Polaris, but if that's the case it means that she's either lying about the "good specs" or two of the big things about Iwata's designs (small and low power consumption) has been compromised. It makes me wonder if she actually knows anything at all or if she's just making stuff up to keep people in check according to what she has personally deemed reasonable. As a result, it makes more sense to just toss out everything she's said and go with what's most logical based on what's known. And, frankly, that's Polaris 11, that's 14nm, and that's the same plant as the next Xbox.

Again, all that I'm saying is that the way she's doing it now makes it seem like she cares more about shutting down "unrealistic" expectations than she does about keeping expectations realistic. She's being intentionally vague.
 

ozfunghi

Member
That's the thing, though. She's made her own decisions about what "in check" and "Iwata's philosophy" mean, and it's resulted in huge contradictions that can only be solved by being less vague. For example, she's decided that an Iwata console obviously would not use Polaris, but if that's the case it means that she's either lying about the "good specs" or two of the big things about Iwata's designs (small and low power consumption) has been compromised. It makes me wonder if she actually knows anything at all or if she's just making stuff up to keep people in check according to what she has personally deemed reasonable. As a result, it makes more sense to just toss out everything she's said and go with what's most logical based on what's known. And, frankly, that's Polaris 11, that's 14nm, and that's the same plant as the next Xbox.

Again, all that I'm saying is that the way she's doing it now makes it seem like she cares more about shutting down "unrealistic" expectations than she does about keeping expectations realistic. She's being intentionally vague.

When did she say that about what a Iwata console would be (meaning no polaris)? By that logic, she would mean a small, quiet and energy efficient console on proven tech (CPU/GPU)... and that would mean 14nm over 28nm, especially considering the timeframe.

She also made a comment that the 10k rumors got the "power" wrong. Power as in performance, or power as in consumption?
 

IntelliHeath

As in "Heathcliff"
I have not ben following those threads very closely but... since NX won't be at E3, when are we expecting the reveal ?

Some people are being optimistic with the reveal going to happens in September when I'm being pessimist so I think the reveal isn't going to happens until November.
 

-Horizon-

Member
Figured I'll put in my pessimistic, realistic, and optimistic NX specs again. Created with the help of Wiki, Google, Thaktor, Blu, and my own experiences with Xbox one and PS4 leaks.

And logic.

Pessimistic.
8 core PUMA 1.6 - 2.0 ghz.
8-10 CU GCN 1.2 GPU at 800 - 1000 mgz. 1 teraflop.
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf close to slightly under Xbone. Might have trouble with high bandwidth scenarios such as alpha effects, deferred rendering, etc. Cost of BOM is REALLY low though. Can retailed for $199 and still make a profit. Or pack in an expensive gimmick and retail for $299.

Realistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
12-16 CU Polaris GPU at 1 ghz. 1.6 to 2.0 teraflop.
64 mb of esram. (Takes up less die space than 32mb of esram on 28nm)
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.

Optimistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
20 - 24 CU Polaris GPU at 1.0 ghz. 2.5 - 3 teraflops.
8GB of GDDR5X or combination of LPDDR4 + HBM2.

Real world per 2x ps4.
Break even / slight profit at $399 retail.
If it's cheap, I'd honestly be fine with it not being a powerhouse. Though people will rain fire and brimstone regardless.
 

antonz

Member
No reason to get salty over her Tweet. There is absolutely zero chance that NX parts are in production at this point. So there is honestly no chance that Chips were being made to wrong specifications etc.

They could start production as late as November on actually manufacturing the consoles.
 

Nightbird

Member
gamescom?

Nah, they said they are going to create an own event for it, so we shouldn't expect them to reveal it at either gamescom or TGS. BUT, I'd say around the time of TGS (18.09), a week before/after, would be the earliest realistic date for it.

Yeah, realistically, it should be between September and November.


So... only 5 to 7 months until we see that damn thing. ^^

Yep. I'm putting my money on September. Can't explain it though. I just feel like once you are past September, it's going to be a lot harder getting people's attention.
 
Figured I'll put in my pessimistic, realistic, and optimistic NX specs again. Created with the help of Wiki, Google, Thaktor, Blu, and my own experiences with Xbox one and PS4 leaks.

And logic.

Pessimistic.

Real world perf close to slightly under Xbone. Might have trouble with high bandwidth scenarios such as alpha effects, deferred rendering, etc. Cost of BOM is REALLY low though. Can retailed for $199 and still make a profit. Or pack in an expensive gimmick and retail for $299.

Realistic.
Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.

Optimistic.
Real world per 2x ps4.
Break even / slight profit at $399 retail.

So have fans all casted away the idea of NX hardware sharing a base? Or how do you think they're going to get that in a handheld with the power-inefficient AMD x86 and power-hungry Puma?
 

The_Lump

Banned
Even just on GAF we've had power predictions ranging from suped-up WiiU all the way to PS4K-beating behemoth. So Emily Rogers telling folk to be "realistic" really doesn't tell us anything.
 

EDarkness

Member
Even just on GAF we've had power predictions ranging from suped-up WiiU all the way to PS4K-beating behemoth. So Emily Rogers telling folk to be "realistic" really doesn't tell us anything.

Yeah. I'd go so far as to ask does that mean it won't be a beefed up Wii U, either? What's "realistic"? Her comment doesn't mean anything without something to ground it to "reality".
 

ozfunghi

Member
No reason to get salty over her Tweet. There is absolutely zero chance that NX parts are in production at this point. So there is honestly no chance that Chips were being made to wrong specifications etc.

They could start production as late as November on actually manufacturing the consoles.

Well, that's where the rumor about the mix up came in. I think the premise was that the NX would have been scheduled to release sooner, but because the chips were mixed up, it had to be pushed back. Hence, it "should" have gone in production about now initially.
 

MuchoMalo

Banned
No reason to get salty over her Tweet. There is absolutely zero chance that NX parts are in production at this point. So there is honestly no chance that Chips were being made to wrong specifications etc.

They could start production as late as November on actually manufacturing the consoles.

Actually, it would start earlier if it's 14nm because part of the reason for the ban is most likely due to yields. But yes, this would be too early. The thing is that this also means that she shouldn't even know where it's being fabbed, so why is she claiming that she does? Not even developers have any reason to know the fabrication node, let alone which plant it's at. In other words, she's not being completely honest, which is damaging to her credibility.
 

Vena

Member
A device from the Iwata era, huh?

So, how many gamecubes is that?

In all seriousness, though, that sort of disclosure puts us anywhere from GameCube (start of his career) to WiiU (end of his career). Which tells us nothing.
 

ozfunghi

Member
Actually, it would start earlier if it's 14nm because part of the reason for the ban is most likely due to yields. But yes, this would be too early. The thing is that this also means that she shouldn't even know where it's being fabbed, so why is she claiming that she does? Not even developers have any reason to know the fabrication node, let alone which plant it's at. In other words, she's not being completely honest, which is damaging to her credibility.

I think that's a valid point. If it's bollocks (which i believe it is) then, how would she know (that both NX and XBO chips aren't collocated)?
 

Rodin

Member
Realistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
12-16 CU Polaris GPU at 1 ghz. 1.6 to 2.0 teraflop.
64 mb of esram. (Takes up less die space than 32mb of esram on 28nm)
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.
I'd be more than ok even with the low end of this lol. Just swap Puma+ with 8 A72 cores at 2GHZ for games + 4 A53 at 1.2GHZ for the OS.

I won't complain too much even if the pessimistic scenario happens, but it would still be disappointing.
 

Vena

Member
I think that's a valid point. If it's bollocks (which i believe it is) then, how would she know (that both NX and XBO chips aren't collocated)?

That's easy to answer. "Hey Emily, our chips aren't being made in the same facility as Microsoft's." All of one engineer is required to make this statement, and they would likely know if they are sharing a facility.
 

Hilarion

Member
Figured I'll put in my pessimistic, realistic, and optimistic NX specs again. Created with the help of Wiki, Google, Thaktor, Blu, and my own experiences with Xbox one and PS4 leaks.

And logic.

Pessimistic.
8 core PUMA 1.6 - 2.0 ghz.
8-10 CU GCN 1.2 GPU at 800 - 1000 mgz. 1 teraflop.
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf close to slightly under Xbone. Might have trouble with high bandwidth scenarios such as alpha effects, deferred rendering, etc. Cost of BOM is REALLY low though. Can retailed for $199 and still make a profit. Or pack in an expensive gimmick and retail for $299.

Realistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
12-16 CU Polaris GPU at 1 ghz. 1.6 to 2.0 teraflop.
64 mb of esram. (Takes up less die space than 32mb of esram on 28nm)
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.

Optimistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
20 - 24 CU Polaris GPU at 1.0 ghz. 2.5 - 3 teraflops.
8GB of GDDR5X or combination of LPDDR4 + HBM2.

Real world per 2x ps4.
Break even / slight profit at $399 retail.

I'm surprised even your pessimistic scenario has a system with 8 gigs of RAM and a teraflop GPU. Your pessimistic system is the XBone's kid brother, but that system would still be leaps and bounds more powerful than the Wii U (the difference between 180 gigaflops and 1 teraflop, or from 2 gigs of RAM to 8, is self-evidently substantial). So you're putting your low-end expectation on a system 4-6x the WiiU and going up from there?

Honestly, your "Realistic" is my "Optimistic" and your "Optimistic" is my "No way."
 

ozfunghi

Member
That's easy to answer. "Hey Emily, our chips aren't being made in the same facility as Microsoft's." All of one engineer is required to make this statement, and they would likely know if they are sharing a facility.

Maybe. Then why doesn't she just say that?
 

aBarreras

Member
I'm surprised even your pessimistic scenario has a system with 8 gigs of RAM and a teraflop GPU. Your pessimistic system is the XBone's kid brother, but that system would still be leaps and bounds more powerful than the Wii U (the difference between 180 gigaflops and 1 teraflop, or from 2 gigs of RAM to 8, is self-evidently substantial). So you're putting your low-end expectation on a system 4-6x the WiiU and going up from there?

Honestly, your "Realistic" is my "Optimistic" and your "Optimistic" is my "No way."

how many times is the Wii U more powerful thatn te Wii? maybe is not that unrealistic, i dont know
 
Figured I'll put in my pessimistic, realistic, and optimistic NX specs again. Created with the help of Wiki, Google, Thaktor, Blu, and my own experiences with Xbox one and PS4 leaks.

And logic.

Pessimistic.
8 core PUMA 1.6 - 2.0 ghz.
8-10 CU GCN 1.2 GPU at 800 - 1000 mgz. 1 teraflop.
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf close to slightly under Xbone. Might have trouble with high bandwidth scenarios such as alpha effects, deferred rendering, etc. Cost of BOM is REALLY low though. Can retailed for $199 and still make a profit. Or pack in an expensive gimmick and retail for $299.

Realistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
12-16 CU Polaris GPU at 1 ghz. 1.6 to 2.0 teraflop.
64 mb of esram. (Takes up less die space than 32mb of esram on 28nm)
8-12 GB LPDDR4 on 128 bit bus. 68 gb/s. (Equivalent to 93 gb/s on pre GCN 1.2 devices)

Real world perf 25-50% greater than PS4.
Can swap the esram for 256 bit bus and twice the LPDDR4 chips at half density if that's a cheaper route. Break even / slight profit at $299 retail.

Optimistic.
14nm
8 core Puma+ 1.6 - 2.0 ghz.
20 - 24 CU Polaris GPU at 1.0 ghz. 2.5 - 3 teraflops.
8GB of GDDR5X or combination of LPDDR4 + HBM2.

Real world per 2x ps4.
Break even / slight profit at $399 retail.

I think something you guys need to keep in mind when making these predictions is that the costs aren't solely attributable to the board set-up. It also factors in the shell casing, cooling fan, included HDD, and assembly into an actual console, not to mention the cost for the disc drive (if it's using discs; in Nintendo's case also factor in the licensing costs for blu-ray).

For the actual system box (including packaging) you also have to factor in the controller cost (even if that can be subsidized) and all of that needs to fit within a given MSRP range of, let's say $299 in this case.

Not exactly sure how much reseller distribution and stocking factors into MSRP costs, but I figure that also goes into the price, tho not by too much (retailers don't make profit on selling these systems AFAIK).

I'm only speaking for myself, but if she actually DOES know this tech stuff, she should be more clear. She is using twitter to form an opinion on a discussion on Gaf, instead of partaking in the discussion herself. That wouldn't be a problem in itself if we could ask her what exactly she means by it, or how we should interpret that info. It's a bit too much hit&run for my liking. She says the story about the AMD mix up is bollocks for a number of reasons, one of which being the fact that both chips are not collocated with each other. For instance, is she saying that the NX chip will be on a different process node than the XBOII chip? Which would indeed mean the NX chip might very well be on a 28nm process. Or maybe she's debunking the existence of an XBOII alltogether (if there is no XBOII, then there is no XBOII chip, and there can't be a mixup, and the chips can't be collocated either... still leaving any process node on the table for NX).

I mean, to me it feels like she doesn't fully understand the implications her statements bring to the table, which in turn, has me wondering if she understands the nature of the subject, and in turn whether or not she is misinterpreting stuff she doesn't fully understand.

If the chip mixup stuff is fake, then what's the real reason NX is skipping E3 and for its delay until "around March 2017"? At least the AMD chip stuff sounds logical and is an acceptable reason for the delay of the system's launch, and yes, it was delayed; with Wii U turning out what it's becoming, there's zero chance they were intending for a 2017 release.

Seriously, if the AMD chip mixup is false (which I doubt), what's a true reason for the non-E3 presence and 2017 release that isn't mired in stupidity or poor management decisions?

No hostility directed towards you in particular; I'm simply jumping off from here and thinking aloud.
 

Vena

Member
Maybe. Then why doesn't she just say that?

Does she have to? I think its fairly obvious one of her (and probably other leakers/insiders) contacts is an engineer at Redmond.

I'm surprised even your pessimistic scenario has a system with 8 gigs of RAM and a teraflop GPU. Your pessimistic system is the XBone's kid brother, but that system would still be leaps and bounds more powerful than the Wii U (the difference between 180 gigaflops and 1 teraflop, or from 2 gigs of RAM to 8, is self-evidently substantial). So you're putting your low-end expectation on a system 4-6x the WiiU and going up from there?

Honestly, your "Realistic" is my "Optimistic" and your "Optimistic" is my "No way."

WiiU isn't 180 GFLOPS.
 
Unveiling the NX makes the most sense from July to September. And no way would Nintendo care enough to reveal the system at Gamescom or TGS. They don't even show up at TGS lol.

The best thing would be to unveil it prior to those events and then they can take the NX to those events to get people to play it hands on, but who knows what Nintendo's plans are. Of course it will have demos at retailers.

They'll probably reveal it in November or December since they're so unpredictable.
 
D

Deleted member 465307

Unconfirmed Member
Unveiling the NX makes the most sense from July to September. And no way would Nintendo care enough to reveal the system at Gamescom or TGS. They don't even show up at TGS lol.

The best thing would be to unveil it prior to those events and then they can take the NX to those events to get people to play it hands on, but who knows what Nintendo's plans are. Of course it will have demos at retailers.

They'll probably reveal it in November or December since they're so unpredictable.

My expectations are for a reveal between August and October, but I think your range is perhaps better since October is already the beginning of the holiday marketing machine and release rush. I could also see a reveal happen in the dead of December or early January.

However, I think, ideally, we'd get a reveal before the Sonic event in July (I want an NX game announced there!) and maybe some kind of showcase at TGS or a separate event around that time (since Nintendo at TGS is rare).
 

bomblord1

Banned
I'm expecting the system to be based on around mobile tech which has me thinking a high end mobile CPU/GPU combo probably over clocked because the console will have the thermal headroom. Highest flops I'm expecting is comfortably between the Xbox one and PS4. I would love to see a PSNeo competitor though.

Handheld same hardware probably less cores and underclocked maybe a different GPU to make the shared architecture work. Basically little and big brother.
 
My expectations are for a reveal between August and October, but I think your range is perhaps better since October is already the beginning of the holiday marketing machine and release rush. I could also see a reveal happen in the dead of December or early January.

However, I think, ideally, we'd get a reveal before the Sonic event in July (I want an NX game announced there!) and maybe some kind of showcase at TGS or a separate event around that time (since Nintendo at TGS is rare).

Ideally, I think the best time is prior to holiday season and have the demos at Gamescom and more importantly TGS. Also, there better be a commercial during the Super Bowl. Ok they missed the holiday, but you do NOT miss a chance to advertise the NX at the Super Bowl considering it launches a month after the Super Bowl. There's like over 100 million people tuning into it and that's the perfect advertising opportunity they have.
 
Does she have to? I think its fairly obvious one of her (and probably other leakers/insiders) contacts is an engineer at Redmond.



WiiU isn't 180 GFLOPS.

You do happen to realize that if the AMD rumors are false, then things are probably actually quite bad or worst behind-the-scenes, right? At least a chip mixup validates a reason for the delay and (kind of) skipping E3 or Gamescon (if nothing's really ready to demo by next month, an extra month won't make a difference) in a way that falls completely outside of Nintendo's control.

Otherwise the onus is completely on them; they're choosing to forgo E3 and/or Gamescon for their own event b/c software that should be further along than it is, happens to not be as far along as it should. Which would suggest software development got started later than it should have, very weird considering Wii U's situation has been known for a few years by now.

That means the decision to skip the holiday season (and the sales that'd come from such) is of Nintendo's own volition, which makes zero sense b/c the only two easily recognizable game devices that had releases outside of the holiday season (Saturn, 3DS) both struggled out of the gate and botched those releases.

Even aside from that, if by some chance (however slim) they got the idea after seeing Sony's 1st party games doing well outside of the cluttered holiday season, it's still a non-optimal choice, because Nintendo's games don't usually have a problem selling during the holiday period, and even the Wii U did very well during its first holiday season (which happened to coincide with its launch).

So it still doesn't make sense. And I've yet to hear one reason why these measures are being taken by Nintendo for the NX aside from the AMD chip issues that sounds logical or reasonable, and/or outside of Nintendo's realm of control.
 

Vena

Member
You do happen to realize that if the AMD rumors are false, then things are probably actually quite bad or worst behind-the-scenes, right? At least a chip mixup validates a reason for the delay and (kind of) skipping E3 or Gamescon (if nothing's really ready to demo by next month, an extra month won't make a difference) in a way that falls completely outside of Nintendo's control.

No. You and Malo need to divorse your imaginations from reality. There are many other reasons for a potential delay that still have to do with fabrication and hardware.
 

MuchoMalo

Banned
A device from the Iwata era, huh?

So, how many gamecubes is that?

In all seriousness, though, that sort of disclosure puts us anywhere from GameCube to WiiU. Which tells us nothing.

My point exactly, though I guess GameCube isn't Iwata-era. So, what is? Well, let's look at DS, Wii, 3DS, and Wii U. What do they all have in common?

  1. Weak for the time of release
  2. (except Wii U) Low BOM
  3. Outdated hardware (this could have possibly excluded 3DS if is used Tegra as rumored)
  4. Low power consumption
  5. Small
  6. Quiet and unobtrusive
  7. (excluding Wii) At least one fabrication node behind the competition
  8. (Wii and Wii U) Rehashed Gecko CPU
  9. Backwards compatibility
  10. GPU tech at least 4 years old
  11. More RAM than expected for power level

Now, here's the issue: It's impossible to hit all of these while having "good specs" unless your definition of that is simply "better than Wii U," in which case she was being extremely misleading. It wouldn't really shock me though, since she seems to have either given up on third-party support or thinks that we shouldn't care. From there, I think these are the most logical "Iwata era" designs. These factor in the lack of an optical drive.

Config 1: Backwards compatible (does not use an APU but instead an SoP like Wii U)
  • 8-core version of the Wii U CPU, shrunk to 32nm or 28nm, @1.5-1.6GHz
  • 28nm GCN 1.0 Oland-based (6-8CUs) GPU @~800MHz with 32MB eSRAM
  • 4-6GB LPDDR3 or LPDDR4 RAM
  • Flash-based storage only
  • GPU customized to work perfectly with Wii U games
  • 100% BC
  • Power consumption on-par with Wii U
  • Would be only the second Nintendo console to be a node behind
  • Contradicts: "good specs", LCGeek, WSJ's "industry leading chips"
  • Cannot scale down to a handheld due to CPU


Config 2: 28nm low-power consumption (under 60W most likely even with an HDD)
  • 8-core ARM A53 or A57 @1.75-2.2GHz (Note: A53 would contradict LCGeek)
  • Pre-Polaris GCN based on Cape Verde (8-10CU) @800MHz or less and a 256-bit bus
  • 8GB LPDDR3 or LPDDR4 (no eSRAM to keep the size and cost of the APU down)
  • Might have an HDD, might not
  • Can possibly scale down to a handheld by cutting down CPU, GPU, and RAM if LPDDRx
  • Contradicts: Iwata points 8, 9, 11, possibly 10, possibly "good specs", possibly "industry leading chips", possibly LCGeek

Config 3: 28nm PS4-parity
  1. 8-core ARM A57 or A72 @1.75-2.2GHz
  2. Pre-Polaris GCN with 16-18CUs @800-900MHz with 256-bit bus
  3. 8GB GDDR5 or LPDDR4
  4. Might have an HDD, might not
  5. Can possibly scale down to a handheld by cutting down CPU, GPU, and RAM if LPDDRx
  6. Contradicts: Iwata points 4, 8, 9, 11 either 5 or 6, possibly 10, possibly "industry leading chips"

Config 4: Low-power 14nm Special
  1. 8-core ARM A72 @1.75-2.2GHz
  2. Polaris 11-based GPU with 16-18CUs @850-1000MHz with custom 256-bit bus to avoid using eSRAM, reducing the size of the APU and increasing yields
  3. 8-12GB LPDDR4
  4. Might have an HDD, might not
  5. Can easily scale down to a handheld by cutting down CPU, GPU, and RAM if LPDDRx
  6. Slightly faster than PS4
  7. Should consume under 65W with an HDD
  8. 14nmFF yields could explain the delay, but a chip like this shouldn't be much bigger than a Polaris 10 GPU die so I'm not sure.
  9. Contradicts: Iwata points 8-10, Emily's statements on specs other than "good specs"

I spent too much time on this, but yeah, these are the best I can think of in terms of ballparks. I don't even remember why I did this. Whatever. I think it sums things up nicely anyway. Tell me if I missed anything. I believe that all of these should be able to be sold for $300 or less in March 2017 while not posting a loss. I also believe that Emily automatically took "we are not selling at a loss" to mean "we plan to have a healthy profit margin," which I doubt is the case if exchange rates had to be mentioned.
 
No. You and Malo need to divorse your imaginations from reality. There are many other reasons for a potential delay that still have to do with fabrication and hardware.

Like what? Polaris isn't suffering any delays and that'll be in products around or before September. So if it's using Polaris that in itself shouldn't be causing any hiccups. If it's using carts and waiting on something from Macronix, I can't imagine they'd of waited so long as to have Macronix figure something out only towards the start of the holiday season if not later. So the only other possible thing hardware-wise is that the AMD rumors are true, but obviously Nintendo would want to downplay that publicly since it's not a good look.

As for the other things? Like I said, if the software isn't ready, then development should have began sooner. If Iwata's passing was a major setback, they should have had contingency plans in place to minimize those setbacks even in the worst-case scenario. It's also possible that their resources were just stretched too thin between 3DS, Wii U and NX last year and maybe the year before, and that affected NX software development. Even there things land back at Nintendo's doorstep.

Just delaying because you want the software to be "perfect" doesn't quite add up when that means missing the biggest holiday shopping season of the year and releasing during one of the most quiet periods instead (only summer routinely beats that dubious claim). And whatever hardware fab issues they're having should have zero bearing on 1st party NX software from being present at E3, because that software (if it's as far along as it should have been) can run off of PCs simulating the NX hardware setup.

And just choosing to ignore E3 for your own special reveal event also makes no sense, because that just means one less chance to get an impression out there, and essentially means less time for spreading the word and promotion. Hell, even that causes problems because now you increase the chances for consumer confusion by condensing all hype and marketing into a smaller crunch period, especially if there's a lot of new stuff to learn about that product.

So until there's some concrete clarification I'll buy the AMD rumors as being true, because anything else makes Nintendo look ridiculously more unprepared than I'd like to believe.
 

Vena

Member
The Wii had a more potent device designed in parallel. It was dropped for the Wii and mote to keep the price down.

Which was the right decision.
 

MuchoMalo

Banned
The Wii had a more potent device designed in parallel. It was dropped for the Wii and mote to keep the price down.

Which was the right decision.

Yeah, I'm aware of this, which is another point against Emily's assumptions of what an Iwata console is. Wii U was downgraded too.

Like what? Polaris isn't suffering any delays and that'll be in products around or before September. So if it's using Polaris that in itself shouldn't be causing any hiccups. If it's using carts and waiting on something from Macronix, I can't imagine they'd of waited so long as to have Macronix figure something out only towards the start of the holiday season if not later. So the only other possible thing hardware-wise is that the AMD rumors are true, but obviously Nintendo would want to downplay that publicly since it's not a good look.

As for the other things? Like I said, if the software isn't ready, then development should have began sooner. If Iwata's passing was a major setback, they should have had contingency plans in place to minimize those setbacks even in the worst-case scenario. It's also possible that their resources were just stretched too thin between 3DS, Wii U and NX last year and maybe the year before, and that affected NX software development. Even there things land back at Nintendo's doorstep.

Just delaying because you want the software to be "perfect" doesn't quite add up when that means missing the biggest holiday shopping season of the year and releasing during one of the most quiet periods instead (only summer routinely beats that dubious claim). And whatever hardware fab issues they're having should have zero bearing on 1st party NX software from being present at E3, because that software (if it's as far along as it should have been) can run off of PCs simulating the NX hardware setup.

And just choosing to ignore E3 for your own special reveal event also makes no sense, because that just means one less chance to get an impression out there, and essentially means less time for spreading the word and promotion. Hell, even that causes problems because now you increase the chances for consumer confusion by condensing all hype and marketing into a smaller crunch period, especially if there's a lot of new stuff to learn about that product.

So until there's some concrete clarification I'll buy the AMD rumors as being true, because anything else makes Nintendo look ridiculously more unprepared than I'd like to believe.

An APU is a more complex design than a GPU, as well as being larger. That means lower yields. On top of this, Nintendo probably doesn't have the most favorable track record given Wii U and might be lower-priority than the others using the node this year (Apple, Samsung, Sony, Microsoft, and Qualcomm, to name a few).

Also note that, historically, the 360 seems to be the only console in the past 20+ years to use a node that's been in use for less than 12 months.

PS: Polaris launches either this month or next, not September.
 

Ogodei

Member
Like what? Polaris isn't suffering any delays and that'll be in products around or before September. So if it's using Polaris that in itself shouldn't be causing any hiccups. If it's using carts and waiting on something from Macronix, I can't imagine they'd of waited so long as to have Macronix figure something out only towards the start of the holiday season if not later. So the only other possible thing hardware-wise is that the AMD rumors are true, but obviously Nintendo would want to downplay that publicly since it's not a good look.

As for the other things? Like I said, if the software isn't ready, then development should have began sooner. If Iwata's passing was a major setback, they should have had contingency plans in place to minimize those setbacks even in the worst-case scenario. It's also possible that their resources were just stretched too thin between 3DS, Wii U and NX last year and maybe the year before, and that affected NX software development. Even there things land back at Nintendo's doorstep.

Just delaying because you want the software to be "perfect" doesn't quite add up when that means missing the biggest holiday shopping season of the year and releasing during one of the most quiet periods instead (only summer routinely beats that dubious claim). And whatever hardware fab issues they're having should have zero bearing on 1st party NX software from being present at E3, because that software (if it's as far along as it should have been) can run off of PCs simulating the NX hardware setup.

And just choosing to ignore E3 for your own special reveal event also makes no sense, because that just means one less chance to get an impression out there, and essentially means less time for spreading the word and promotion. Hell, even that causes problems because now you increase the chances for consumer confusion by condensing all hype and marketing into a smaller crunch period, especially if there's a lot of new stuff to learn about that product.

So until there's some concrete clarification I'll buy the AMD rumors as being true, because anything else makes Nintendo look ridiculously more unprepared than I'd like to believe.

Just because products may be launching in September doesn't mean that this is right in time for Nintendo to roll out a full platform (especially if it is going to be console plus handheld, requiring twice the silicon).
 

Vena

Member
Yeah, I'm aware of this, which is another point against Emily's assumptions of what an Iwata console is. Wii U was downgraded too.

No, its fairly consistent for Iwata's placement in the history of Nintendo. He never shied from power or high-end tech, but he'd simply put innovative ideas or in some cases fads ahead of it.

The GameCube is in a time when he was already high-up in creative and control positions of Nintendo, and he had his hands in its development. That was a powerful machine but it was smart, sleek, and small, and its performance was measured in real figures not the BS of Sony or Microsoft at the time... which hasn't abated.

The Wii and its parallel brother were a balance of innovation and competing power. Innovation won on a very good call, but even then there was a strong older brother. Again, Iwata never shied from power but he went with sleek, affordable, and innovative. They could have launched a 400$ Wii360... probably best they didn't.

3DS would have sported a brand-spanking-new chip right off the production run had Tegra done what was needed and what had been pitched (burn those bridges). It ended up weaker but not intentionally so, and ultimately just had its "3D fad" as a selling point aside from "Nintendo handheld".

The WiiU followed sleek and an idea on innovation, it just followed a bad idea. Even then, the hardware it has is good given the time and design ethos, just ask Blu (unless I am misremembering).

Still, all of these systems were designed to be affordable and profitable (gamecube was cutting it close), and the WiiU wouldn't have been a loss-leader if not for the value change in Yen (and the expensive screen controller).

Moreover, given other information we have on hand that I'd say is more than believable, its that Nintendo is looking to make sure that UE4 (not mobile) runs flawlessly on this machine. That gives you a floor on performance right there, and we also know where the CPU generally weighs in at on vague benchmark metrics.

Here's your happy floor for UE4, or, like Sony and Microsoft, you can go lower much to their chagrin: https://wiki.unrealengine.com/Recommended_Hardware
 
Status
Not open for further replies.
Top Bottom