Nintendo Switch Tech Speculation discussion

Status
Not open for further replies.
The Fox Con leaks have only gained credibility, the guy nailed the battery size, colored joy cons, L shaped header pipe. The guy had certainly seen the Switch in production. The problem is people were taking the leakers speculation as fact. We had a lot of lost in translation issues.

Laura K Dale also had enough details correct to not be outright dismissed. Sure, she has some bad info, but she also nailed a lot of things.

Eurogamer has been the most reliable, but they are also very careful to not report on anything they aren't fully confident in. I'm not discrediting Eurogamer, but I have not seen them report anything saying a clock speed boost was not done in October. Eurogamer hasn't always been accurate, remember when the 3DS was going to use a Tegra processor? They reported that it would, and obviously didnt. When it comes to leaks, they never all pan out, but some of them do.

Sent from my SM-G360V using Tapatalk

they have not gained credibility, only delusional fanboys are paying attention to that garbage. you should go read the eurogamer article they are confirming those clocks are confirmed for being in the final hardware as well those specs, so why would eurogamer give us an update.
 
At least that gives a clear upgrade path ... glorified NVIDIA Shield TV it is.
Iwata did say NX could be one device or many, so it's certainly possible for Switch to become a few devices of the market demands it.

Sent from my SM-G360V using Tapatalk
 
they have not gained credibility, only delusional fanboys are paying attention to that garbage. you should go read the eurogamer article they are confirming those clocks are confirmed for being in the final hardware as well those specs, so why would eurogamer give us an update.
So he guess all those hardware specifics out of thin air? Speaking of dillusional.......

Sent from my SM-G360V using Tapatalk
 
Enough with the non-tech credibility discussion, please.
That's fine, but we are still dealing with mostly rumors. So what I find to be probable or at least plausible, another member might outright dismiss as false. Unlike Wii U where in game performance seemed to go against any optimistic beliefs surrounding it's specs, Switch seems to be in the opposite position. In game results are outperforming expectations for a low clocked Tegra X1. Nvidia must be wizard's with their tools and API, because developers seem to be praising Switch. Go check out Capcoms recent statements.

Sent from my SM-G360V using Tapatalk
 
because developers seem to be praising Switch. Go check out Capcoms recent statements.

FOR A MOBILE PART. You need to put context into the developers praise.

Previously, Nintendo also had Developers saying the same shit for WiiU, so take with appropriate mountains of salt.
 
That's fine, but we are still dealing with mostly rumors. So what I find to be probable or at least plausible, another member might outright dismiss as false. Unlike Wii U where in game performance seemed to go against any optimistic beliefs surrounding it's specs, Switch seems to be in the opposite position. In game results are outperforming expectations for a low clocked Tegra X1. Nvidia must be wizard's with their tools and API, because developers seem to be praising Switch. Go check out Capcoms recent statements.

Sent from my SM-G360V using Tapatalk

this is just an opinion and not fact, so far most games are in line with what's expected, and with some being disappointing, zelda should diffidently run at 1080p with the DF specs, but that was probably a rushed port.
 
this is just an opinion and not fact, so far most games are in line with what's expected, and with some being disappointing, zelda should diffidently run at 1080p with the DF specs, but tit was probably a rushed port.

brit i was about to post the samething. developers are under nda, and they are pretty much pr talk.
 
this is just an opinion and not fact, so far most games are in line with what's expected, and with some being disappointing, zelda should diffidently run at 1080p with the DF specs, but that was probably a rushed port.
Your stating fact? Sounds like an opinion to me. Assassin's Creed 4 ran in 900p on Xbox One, a last generation port. Cross generation ports never maximize the hardware. After seeing the innards of the Switch all so crammed together, it only made the Switch more impressive to me. A tablet sized portable capable of exceeding previous generation consoles. Good stuff.

What's funny is I actually think your mostly right, but I take offense with your opinion stated as if it were fact.

Sent from my SM-G360V using Tapatalk
 
Last edited:
Your stating fact? Sounds like an opinion to me. Assassin's Creed 4 ran in 900p on Xbox One, a last generation port. Cross generation ports never maximize the hardware. After seeing the innards of the Switch all so crammed together, it only made the Switch more impressive to me. A tablet sized portable capable of exceeding previous generation consoles. Good stuff.

Sent from my SM-G360V using Tapatalk

we already knew the shield tv could to that, there are 360/ps3 games that run 1080p and some even with better frame rate then 360/ps3 so we know it has the power. so i don't know why you're surprised considering, switch doesn't have to run android, is coded to the metal, and has Nividia latest dev tools especially designed to take advantage of switch.
 
this is just an opinion and not fact, so far most games are in line with what's expected, and with some being disappointing, zelda should diffidently run at 1080p with the DF specs, but that was probably a rushed port.

Zelda has a lot of foliage, potentially stressing memory BW, and the WiiU can't hold a steady 30 fps at 720p.

Nintendo may have chosen a stable 30 fps at 900p over a stuttering BW choked 1080p.

BW is likely the one area where NX doesn't have an advantage over WiiU. NX does not appear to have much BW at all.
 
we already knew the shield tv could to that, there are 360/ps3 games that run 1080p and some even with better frame rate then 360/ps3 so we know it has the power. so i don't know why you're surprised considering, switch doesn't have to run android, is coded to the metal, and has Nividia latest dev tools especially designed to take advantage of switch.
I guess it's because for so long we assumed Shield TV ran at max clocks, and apprarently it does actually throttle based on the test done by the user at Gaf. So a Tegra X1 with the CPU cores clocked at 1Ghz, and the 256 Cuda cores clocked at 768Mhz, I find the results pretty impressive. Switch is also far more dense than Shield TV, making its thermal limitations much more prevalent.

Sent from my SM-G360V using Tapatalk
 
Some more stuff to add to the Foxconn leaker's list of specs, from that teardown. L-shaped heatpipe, for example.
 
Someone on Reddit posted a comparison of the new 2017 Shield TV and Switch's Tegra SoC
Both have a portion of their serial code similar. 1632A2 PARR39.COP and -A2 at the end.

Hopefully Chipworks or someone will delid these and take a die shot to compare.

uLF2CUq.png
 
Some more Tegra X1 part numbers:

Shield TV 2015: TM670D-A1
Pixel C: TM671D-A2

AFAIK the markings in the middle have to do with where/when it was fabricated and not the part number.

Both Shield TV revisions and Pixel C use the same 2x K4F2E30 Samsung DRAM. I'm not completely sure but I think it's the same on the Switch.

These Tegra X1s are probably all close to the same, and in particular the Shield TV gen 2 part and the Switch would have to be nearly identical to support the same decoupling capacitor layout on the package. It's typical for high volume OEMs like Nintendo to get their own part number and it could mean some tuned binning and configuration fusing but probably very, very close to the same die.

So it probably doesn't have much better perf/W than the original Shield TV. Which had 4x2GHz CPU/1GHz GPU and consumed over 19W under a heavy gaming load (http://www.anandtech.com/show/9289/the-nvidia-shield-android-tv-review/9). All this is without a display. Nintendo's usage estimates along with the battery rating (~16Wh) gives power consumption for Switch at about 2.5W to 6.5W in portable mode. Minus the display it'd be more like around 1.5W to 4.5W or maybe 5W. I just can't possibly see it running at 4x1.7GHz CPU/921MHz GPU and hitting that power consumption window. On the other hand, Eurogamer's 4x1020MHz CPU/307MHz GPU spec looks about right.

We know for sure that it does clock the GPU higher when docked. We can argue about whether or not Eurogamer is right and the CPU clock remains fixed. Personally I think Nintendo would be making the right decision to keep the CPU locked at the same peak clock speeds between mobile and docked mode, but I've made my case on this already and I don't expect anyone to be convinced now if they disagree.
 
Some more Tegra X1 part numbers:

Shield TV 2015: TM670D-A1
Pixel C: TM671D-A2

AFAIK the markings in the middle have to do with where/when it was fabricated and not the part number.

Both Shield TV revisions and Pixel C use the same 2x K4F2E30 Samsung DRAM. I'm not completely sure but I think it's the same on the Switch.

These Tegra X1s are probably all close to the same, and in particular the Shield TV gen 2 part and the Switch would have to be nearly identical to support the same decoupling capacitor layout on the package. It's typical for high volume OEMs like Nintendo to get their own part number and it could mean some tuned binning and configuration fusing but probably very, very close to the same die.

So it probably doesn't have much better perf/W than the original Shield TV. Which had 4x2GHz CPU/1GHz GPU and consumed over 19W under a heavy gaming load (http://www.anandtech.com/show/9289/the-nvidia-shield-android-tv-review/9). All this is without a display. Nintendo's usage estimates along with the battery rating (~16Wh) gives power consumption for Switch at about 2.5W to 6.5W in portable mode. Minus the display it'd be more like around 1.5W to 4.5W or maybe 5W. I just can't possibly see it running at 4x1.7GHz CPU/921MHz GPU and hitting that power consumption window. On the other hand, Eurogamer's 4x1020MHz CPU/307MHz GPU spec looks about right.

We know for sure that it does clock the GPU higher when docked. We can argue about whether or not Eurogamer is right and the CPU clock remains fixed. Personally I think Nintendo would be making the right decision to keep the CPU locked at the same peak clock speeds between mobile and docked mode, but I've made my case on this already and I don't expect anyone to be convinced now if they disagree.

"Originally Posted by Eurogamer

Documentation supplied to developers along with the table above ends with this stark message: "The information in this table is the final specification for the combinations of performance configurations and performance modes that applications will be able to use at launch."

unless we have solid proof contracting a reliable source like eurogamer, with such hard evidence and sources. i don't why this is up for debate anymore, they are on point. now the guy that has been preaching, The foxcon leak over everything else is saying these pics are from devkit, not final hardware, lmao
 
Zelda has a lot of foliage, potentially stressing memory BW, and the WiiU can't hold a steady 30 fps at 720p.

Nintendo may have chosen a stable 30 fps at 900p over a stuttering BW choked 1080p.

BW is likely the one area where NX doesn't have an advantage over WiiU. NX does not appear to have much BW at all.
That's not how it works, nvidia bandwidth > amd (at least prior to vega) due to different memory compression.

It's also worth noting the draw distance is improved on switch as well, with other improvements/changes here and there. It'll be interesting to see if the draw distance, or something other than resolution is cut back in handheld mode. If that's the case then that would explain why it's only 900p docked.
 
Capcom recently talked about Nintendo providing an optional dev kit that includes a GPU capable of emulating Switch on the PC. Something about making it possible to do all the work on PC, which makes things much faster. This would make the Fox Con leak about the additional GPU plausible, but it's strictly for development kits.

I think the Eurogamer clock speeds are 90 percent likely at this point. I leave open the possibility of a clock speed boost in October because we have seen late in development changes to clock speeds before on various platforms.



Sent from my SM-G360V using Tapatalk
 
The PSP had a clockspeed boost from 222mhz to 333mhz a couple of years into its life. Maybe Nintendo are leaving themselves a bit of wiggle room and, when they have a newer iteration with better battery life, they'll upclock the base model, at least when docked.
 
The model numbers on the LPDDR4 memory chips seem to point to 2GB 32bit chips. Doesn't it seen off to use two chips when they could have gotten a single 4GB 64bit chip?

Sent from my SM-G360V using Tapatalk
 
Status
Not open for further replies.
Back
Top