By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Gaming - Supposedly, Final WiiU Specs from VGLeaks

Soundwave on 11 September 2012
HappySqurriel said:
Soundwave said:

We don't know what Nintendo paid IBM in R&D so that's an assumption at best. People don't realize "R&D cost" in those year end statements also covers *all* R&D, meaning game development costs as well, that's what R&D is filled under, so when they see X amount of dollars for R&D that does not mean they're spending like a billion dollars developing hardware.

Why else would Nintendo do this? Because it's probably damn cheap to do this. We're talking about something that probably needs to be sold at a reasonable profit for as low as $250 with a giant tablet screen in the mix.

If an enhanced Broadway core x3 brings them to about par with the 360 CPU ... it's a direction I could see Nintendo taking.

These same sources also had more detailed specs released around E3 of this year:

http://www.vgleaks.com/world-premiere-wii-u-specs/


This leak is essentially like saying that the PS4 will use a processor that is made up of a hybrid processor made up of a core i7, 4 Cell processors each with 8 SPEs ... Taking the idea of an overly complicated, expensive and difficult to develop for processor to the extreme.

Not sure why you're just randomly throwing a core i7 in there.

Wouldn't a triple CELL be like 18 cores? I could buy that if Sony was really set on delivering something at a budget price.

 



Around the Network
Crono141 on 11 September 2012

Cell only has 1 core, with 6-8 co-processors that take on specific tasks.



Check out my Youtube Let's Play channel here.

Mr Khan on 11 September 2012
Crono141 said:
Cell only has 1 core, with 6-8 co-processors that take on specific tasks.

That being the whole point of the CELL, of course, to do the work of multicore processors without actually having multiple cores.

While this rumor does fall in line with commentary about the Wii U CPU being low-bandwidth, it does contrast multiple confirmed points (about die-size and the thing being a Power7)



Monster Hunter: pissing me off since 2010.

oni-link on 11 September 2012

http://twitter.com/IBMWatson/status/240241146213842944

One of the many proofs Wii U is using an amped version of Power7 processor. I'm done with this bullshit rumor of 3 broadway cores stuck together. Sony and Microsoft fanboys will still trash the Wii U regardless of fact and logic so it doesn't matter since Wii U is clearly only going to be marginally less powerful than the PS4 or Durango.



Soundwave on 11 September 2012

My personal feeling on this has been that this strategy could work for Nintendo IF:

The system is expandable somehow and can be upgradable cheaply in the future. In 2 years time, the 28nm manufacturing process will be matured, and the cost of things like GDDR5 RAM and a 2TFLOP GPU will be dirt cheap.

If the Wii U had a cheap connector like a Crossfire port on the bottom of the unit, it could be expanded around mid/late 2014 and probably be brought up to par directly with the PS4/720 for people who really need that for probably as low as $99.

I don't think that would be a bad play. Launch in 2012 with a cheap machine, start building that userbase, and then have the option of future proofing as well.

If there's no way to expand the system though, I think Nintendo could be in some trouble here if Sony/MS slap LCD screens onto their next-gen controllers too. Because then Nintendo is stuck the next 3-4 years with a system that doesn't have anything particularily unique about it and is a good deal behind on the horsepower angle too.



Around the Network
Crono141 on 11 September 2012
Soundwave said:

My personal feeling on this has been that this strategy could work for Nintendo IF:

The system is expandable somehow and can be upgradable cheaply in the future. In 2 years time, the 28nm manufacturing process will be matured, and the cost of things like GDDR5 RAM and a 2TFLOP GPU will be dirt cheap.

If the Wii U had a cheap connector like a Crossfire port on the bottom of the unit, it could be expanded around mid/late 2014 and probably be brought up to par directly with the PS4/720 for people who really need that for probably as low as $99.

I don't think that would be a bad play. Launch in 2012 with a cheap machine, start building that userbase, and then have the option of future proofing as well.

If there's no way to expand the system though, I think Nintendo could be in some trouble here if Sony/MS slap LCD screens onto their next-gen controllers too. Because then Nintendo is stuck the next 3-4 years with a system that doesn't have anything particularily unique about it and is a good deal behind on the horsepower angle too.

In theory this is a good idea, but we know that in practice it just doesn't work.  For reference, look to Sega's 32X and CD addon.  It fragments your user base, and fragments your developer base.  Bad bad idea.



Check out my Youtube Let's Play channel here.

Pokemonbrawlvg on 11 September 2012

Regardless of how powerful the Wii U is, people have already set their minds on whenever or not to buy it. Not sure how the tech specs affects on the purchase of the console itself. I guess some people are like that. I can't really bring anything into this discussion (mostly because I have no idea what any of this means), but I can bring in some interesting posts found on the internet. Like this one on NeoGaf:

I Stalk Alone: 814

I would guess that on a 3:1 clock ratio to broadway, the Espresso cores would be around 2.187GHz each which might be the maximum clock ceiling the chip is able to reach and probably consuming around 8 Watts. It is possible however that it is only 2:1 ratio at 1.458Ghz which three of these cores will perform about 10% faster than the Xenon if it ran 5 instructions per cycle. We still don't know if there is indeed one main core with more cache.

xenon 2 instructions @ 3.2Ghz x3 cores = 19200 in order instructions 

Espresso 5 instructions (similar to power PC G5) @ 1.458GHz x3 cores = 21870 out of order instructions

even at 1.458Ghz the 3 Espressso cores will beat the Xenon and running at only 4 watts.

The Xenon used a lot of one core for sound and the Wii U has a DSP at 120Mhz which will also help. Isn't there also an arm co cpu? Even at this speed it will perform about 1.5x Xenon.

If however it is clocked at a 3:1 ratio on a 729Mhz bus

Espresso 5 instructions (similar to power PC G5) @ 2.187GHz x3 cores = 32805 out of order instructions

+DSP + the arm co cpu will probably perform about 2.5x the Xenon CPU

For reference three of these cores are faster then Xenon.

The GPU is what we would really like to know as well.

If the GPU has 640 alu's and is clocked at 607.5MHz then we are probably looking at over 768GFLOPS in the same vain as a HD7750 which is 819GFLOPS, but performs faster than a 4870 at 1.2TFLOPS. More realistically, the gpu can be clocked at around 486Mhz which would be around 622.1GFLOPS but still perform faster then the 1000GFLOP HD 4850 @500MHz @110w TDP. For reference, the Radeon E6760 is 576GFLOPS @35w TDP also outperforms the HD4850 which is around 5x faster than the Xenos in real world scenarios.


The WiiU also has 32MB Edram which will help with AA especially on 720p with 4xAA looking good. 1080 will also be possible but probably no AA.

The ram being 1024MB as of now alloted for games is over 2x that of the xbox 360 which also used its ram for the OS. The Wii U OS might have 512MB or more currently dedicated to it. This will mean multitasking while playing games is definitely possible.


So we are still basically unknown on some numbers so we don't know but is it possible we are looking at maybe two scenarios with the first one being likely but the second one still possible?

Espresso Tri core clocked at 1.458Ghz or 2.187Ghz (i hope it is the latter)
"Enhanced Broadway" similar to PowerPC 476FP architecture.
3MB L2 Cache 
core 0: 512 KB
core 1: 2048 KB
core 2: 512 KB
OoOE 
5 instructions per cycle (unknown)
45nm @ 4-8w TDP?


GPU 
32MB Edram 4x AA 720p or 1080p no AA
1024 MB Video DDR3 (2GB total) or GDDR5 (1.5GB total)
486Mhz or 607.5MHz (HD4850 performance or HD4870 performance)
640 ALU 
Open GL 4.3 

The low end will outperform the XBOX 360 probably 2.5x and the high number will be 4.5x
if it was low end I would see $249 with no pack in game but on High End I would see $299 without including the pack in game.


Read my original story on Fictionpress (Shinigami Twin): http://www.fictionpress.com/s/2996503/1/Shinigami-Twin 

As well as my other one (Hell's Punishment): http://www.fictionpress.com/s/3085054/1/Hell-s-Punishment

Nintendo Network ID: kingofe3

Soundwave on 11 September 2012
Crono141 said:
Soundwave said:

My personal feeling on this has been that this strategy could work for Nintendo IF:

The system is expandable somehow and can be upgradable cheaply in the future. In 2 years time, the 28nm manufacturing process will be matured, and the cost of things like GDDR5 RAM and a 2TFLOP GPU will be dirt cheap.

If the Wii U had a cheap connector like a Crossfire port on the bottom of the unit, it could be expanded around mid/late 2014 and probably be brought up to par directly with the PS4/720 for people who really need that for probably as low as $99.

I don't think that would be a bad play. Launch in 2012 with a cheap machine, start building that userbase, and then have the option of future proofing as well.

If there's no way to expand the system though, I think Nintendo could be in some trouble here if Sony/MS slap LCD screens onto their next-gen controllers too. Because then Nintendo is stuck the next 3-4 years with a system that doesn't have anything particularily unique about it and is a good deal behind on the horsepower angle too.

In theory this is a good idea, but we know that in practice it just doesn't work.  For reference, look to Sega's 32X and CD addon.  It fragments your user base, and fragments your developer base.  Bad bad idea.

Completely different eras, in the early/mid-90s most game consumers had to rely on mommy and daddy to buy any and all video game hardware they got. Things have changed ... a lot.

The 32X and Sega CD also had very few good games and cost more than the base hardware itself. 32X was $150 alone I believe on top of the $99 Genesis, Sega CD was even more expensive.

A more modern comparable would be the Kinect add-on for the XBox 360 or Nintendo's RAM pak for the N64.

If Nintendo could introduce something affordable, like say for $99.99-$149.99 with a game bundled (Wii Fit principal) ... if I'm a Wii U owner, I'm more apt to stay with Nintendo rather than paying $400-$500 again to get a 720/PS4 for "next gen" gaming.

I think this could work today. The more scalable nature of third party games today also favors this today. Also obviously today with the rise of Crossfire/SLi ports in the PC industry, support for multiple processing units is no longer anything all that difficult or expensive.

It's a long shot, but I'd be very curious to see if the bottom of the Wii U has an Expansion Slot on there like the SNES/N64/GCN did. If Nintendo designed the Wii U in this way, I think it could be potentially brilliant.



Crono141 on 11 September 2012
Soundwave said:
Crono141 said:
Soundwave said:

My personal feeling on this has been that this strategy could work for Nintendo IF:

The system is expandable somehow and can be upgradable cheaply in the future. In 2 years time, the 28nm manufacturing process will be matured, and the cost of things like GDDR5 RAM and a 2TFLOP GPU will be dirt cheap.

If the Wii U had a cheap connector like a Crossfire port on the bottom of the unit, it could be expanded around mid/late 2014 and probably be brought up to par directly with the PS4/720 for people who really need that for probably as low as $99.

I don't think that would be a bad play. Launch in 2012 with a cheap machine, start building that userbase, and then have the option of future proofing as well.

If there's no way to expand the system though, I think Nintendo could be in some trouble here if Sony/MS slap LCD screens onto their next-gen controllers too. Because then Nintendo is stuck the next 3-4 years with a system that doesn't have anything particularily unique about it and is a good deal behind on the horsepower angle too.

In theory this is a good idea, but we know that in practice it just doesn't work.  For reference, look to Sega's 32X and CD addon.  It fragments your user base, and fragments your developer base.  Bad bad idea.

Completely different eras, in the early/mid-90s most game consumers had to rely on mommy and daddy to buy any and all video game hardware they got. Things have changed ... a lot.

The 32X and Sega CD also had very few good games and cost more than the base hardware itself. 32X was $150 alone I believe on top of the $99 Genesis, Sega CD was even more expensive.

A more modern comparable would be the Kinect add-on for the XBox 360 or Nintendo's RAM pak for the N64.

If Nintendo could introduce something affordable, like say for $99.99-$149.99 with a game bundled (Wii Fit principal) ... if I'm a Wii U owner, I'm more apt to stay with Nintendo rather than paying $400-$500 again to get a 720/PS4 for "next gen" gaming.

I think this could work today. The more scalable nature of third party games today also favors this today.

I still think its a bad idea.  How many different games are there for WiiFit board?  How many blockbuster Kinect games are there?  Games that require addons just don't sell well, and Nintendo knows it.  That's why the wii controller was the motion controller.  That's why the Upad is the WiiU controller.  Nintendo knows that if they don't have these things as part of the install base; that unless every WiiU owner has a Upad, then developers won't develop for it.

Games that require addons never sell well.



Check out my Youtube Let's Play channel here.

Soundwave on 11 September 2012
Crono141 said:
Soundwave said:
Crono141 said:
Soundwave said:

My personal feeling on this has been that this strategy could work for Nintendo IF:

The system is expandable somehow and can be upgradable cheaply in the future. In 2 years time, the 28nm manufacturing process will be matured, and the cost of things like GDDR5 RAM and a 2TFLOP GPU will be dirt cheap.

If the Wii U had a cheap connector like a Crossfire port on the bottom of the unit, it could be expanded around mid/late 2014 and probably be brought up to par directly with the PS4/720 for people who really need that for probably as low as $99.

I don't think that would be a bad play. Launch in 2012 with a cheap machine, start building that userbase, and then have the option of future proofing as well.

If there's no way to expand the system though, I think Nintendo could be in some trouble here if Sony/MS slap LCD screens onto their next-gen controllers too. Because then Nintendo is stuck the next 3-4 years with a system that doesn't have anything particularily unique about it and is a good deal behind on the horsepower angle too.

In theory this is a good idea, but we know that in practice it just doesn't work.  For reference, look to Sega's 32X and CD addon.  It fragments your user base, and fragments your developer base.  Bad bad idea.

Completely different eras, in the early/mid-90s most game consumers had to rely on mommy and daddy to buy any and all video game hardware they got. Things have changed ... a lot.

The 32X and Sega CD also had very few good games and cost more than the base hardware itself. 32X was $150 alone I believe on top of the $99 Genesis, Sega CD was even more expensive.

A more modern comparable would be the Kinect add-on for the XBox 360 or Nintendo's RAM pak for the N64.

If Nintendo could introduce something affordable, like say for $99.99-$149.99 with a game bundled (Wii Fit principal) ... if I'm a Wii U owner, I'm more apt to stay with Nintendo rather than paying $400-$500 again to get a 720/PS4 for "next gen" gaming.

I think this could work today. The more scalable nature of third party games today also favors this today.

I still think its a bad idea.  How many different games are there for WiiFit board?  How many blockbuster Kinect games are there?  Games that require addons just don't sell well, and Nintendo knows it.  That's why the wii controller was the motion controller.  That's why the Upad is the WiiU controller.  Nintendo knows that if they don't have these things as part of the install base; that unless every WiiU owner has a Upad, then developers won't develop for it.

Games that require addons never sell well.


The difference here though is it's not an either or concept. When you buy a PC game like Battlfield 3 on PC it supports many different hardware configurations ... which on the high end are wildly removed from the lowest end, no?

But it's still all on one disc.

I think this principal could be applied to the Wii U. Lets say you buy a game like Watch Dogs on Wii U. The base version looks like the PS3 game. If you have the add-on, it runs like the PS4 version. The assets are already there because developers make them for the PS4/PC/720 anyway, so it's not like they have to be created again from scratch.

As is, the Wii U I don't really see it as a five year product otherwise. It doesn't have as strong of a "hook" for casuals as the Wii did. If MS is really looking into to adding a touchscreen to their next-gen controller ... Nintendo really is hooped IMO. Because then they're stuck with a system that doesn't have anything particularily special about it and gets compromised third party ports if it gets them at all.