• Hey everyone, staff have documented a list of banned content and subject matter that we feel are not consistent with site values, and don't make sense to host discussion of on Famiboards. This list (and the relevant reasoning per item) is viewable here.

StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (Read the staff posts before commenting!)

With the hardware available, it can literally look like this:
Horizon_Forbidden_West_20220204150731_copy.jpg

god-of-war-ragnarok-ps4-performance.jpg

EdQnDdlWAAIV3R-

(PS4 Pro screenshots)

Obviously Zelda will have a different art style, but yeah it has the potential to look damn good.
Crazy that all of the most impressive PS5 games are cross gen titles.

Except Alan wake 2
 
Question, maybe for the more technically inclined, maybe for someone who knows Japanese, the phrase "further-reduced model" was used in reference to Nintendo Switch (Generation 1). Did this mean a reduced model of business leading into and throughout the life of the Switch next, or an admission of an additional minor piece of cost-reduced Nintendo Switch hardware?
 
Brother… the back paddles is meant for the Switch 2 pro controller that cost the same as a switch lite.
ah.gif


But in all serious if the joy cons have back paddles I would be genuinely impressed.
I think Joycons having back paddles would be kinda hard to pull off. Most people's hands are big enough to where a back paddle on the joycon wouldn't be near the tips of their fingers. Back paddles would require a new design of the joycon grip too, where there's no back it to get in the way of said paddles.
 
It's still only a maybe, we don't know what exactly they look like or what they do. They could be touch pads a-la Vita for all we know.
I can't really imagine back touch pads making much sense, though (especially on joy-cons where space is limited), it's probably just buttons. Apparently the details came from third party peripheral makers actively touching the device but not seeing it, so it'd be something they'd notice through touch alone, which touch pads would probably wouldn't be.

Any extra functionality on the back would be fantastic, but I'm gonna go with Occam's Razor and say that simple buttons at the back would be the easiest and cheapest for Nintendo to add to the controller, and with the most actual reliable use-cases.

I think Joycons having back paddles would be kinda hard to pull off. Most people's hands are big enough to where a back paddle on the joycon wouldn't be near the tips of their fingers. Back paddles would require a new design of the joycon grip too, where there's no back it to get in the way of said paddles.
For what it's worth, the joycons are apparently getting a redesign anyway, especially with the new magnetic attachment mechanism, if reports are to be believed.
 
Any chance it could be using Tegra Blackwell?
It's going into cars in 2025 , which takes a couple of years of safety testing, so it's been done for a while; also supports LPDDR5x.
Nope. It's still T239. Nvidia tested (and possibly yes, revised) T239 over the course of 2022, and since then, there has been no more Tegra development (except for Thor which is just now starting to surface in its early stages). While we don't know what they did while testing or revising it, one thing they definitely didn't do is change the GPU architecture or anything else fundamental to the design, because that can't be done in a few months in between tapeouts.
 
How do others feel about those bulletpoints being paraded around?
  • Handheld: Right above PS4* before DLSS
  • Docked: Between PS4 Pro* & Xbox Series S* before DLSS with more modern hardware than the former
  • RAM: Slower than PS5 & XSX|S in
    the speed department, but more capacity than XSS. Should have 10.5-11 GB
    of RAM available to games going by the Switch 1's RAM allotment for its
    OS.
  • Storage: UFS 3.1's max speeds should be a hair under XSX|S (2.1 GB/s vs. 2.4 GB/s), still plenty fast even if not maxed out.Lines up with what Digital Foundry &
    NateDrake heard about decompression techniques & fast load times of
    the BotW tech demo (respectively for each source).
I have always thought trying to say "this is where Switch 2 will be in relative to XSS, PS5, PS4 Pro, etc" is an exercise in futility because the architectures are more than a bit different. As in Switch 2 will excel in areas, but not other areas, compared to XSS, SD, etc.

It's a bit disappointing to see those bulletpoints being pushed/shared in various places (like Twitter for example) more than the shipment data findings have been - because I feel like those kind of assessments can be a bit subjective, I'd rather just stick to the cold hard facts (12GB, 256 UFS 3.1, etc).

Yeah, it's unfortunate. We got some great info that may give us a better sense of where switch 2 lands, but we're still missing some critical details so people shouldn't take what was written there and celebrate too soon.

THAT BEING SAID...

Would these estimates work if Switch 2 used at the very least, the same clockspeeds as the switch?
 
I keep seeing all over the internet today "Switch 2 will be around PS4 level in handheld mode"

Significantly faster CPU, significantly more RAM, significantly faster RAM, significantly faster I/O, GPU architecture and feature sets that are multiple generations newer, etc.

I don't see how these comparisons are made. It's going to be much more capable than a PS4 in handheld mode.
 
Honestly, while all the power speculation is neat (and it might actually exceed expectations, maybe, 1080p DLSS upscaled with 12GB RAM is pretty good), the most exciting reports are that the controllers might actually have back paddles, if only just one on each side (four total would be ideal). The fact that it's taken so long for any of the big three to adopt back paddles as standard is mind-boggling.
Back Paddles are patented by SCUF. Microsoft pays them to use it on Elite controller. I doubt Nintendo would pay some dollars to be able to use it on Switch, but I may be wrong.
 
I would like to know why everyone is so confident with this new info. Why should we trust you in this era of fake news everywhere?

The latest findings that have been determined by some through following shipment data is done very meticulously, continuously questioning what's found and the conversation that sparks from that. I think some even have to paid a subscription to track all the data.
Moreover, it's done by people who have been consistently contributing discussion to the thread and unless someone felt like trolling, I'll believe what's being found as from what I've read and trying to speculate; it lines up with what I think is feasible to end up in the console.

You don't have to believe tbh. Like no joke, if you want to overestimate or underestimate based on any speculation that's made here, at the end of the day the product will be like it is at release. Nintendo's decisions are not relevant to any speculations here. So even if there are conclusions made on fake info, it just impacts the online discussion here at Fami or elsewhere, which depending on your mental state may be a rip in your universe or just another thing that happened.

Moreover, what you will find is, probably the most cohesive collection and documentation about the speculation of a new console that you can go through. Part of that is especially due to the NVIDIA leak and there's been a fair share of "outside" speculation being brought in here, which continuously caused discussion with different conclusions.
In the end it can all be different from the reality the overall thread may be concluding on.
 
I would like to know why everyone is so confident with this new info. Why should we trust you in this era of fake news everywhere?
these are coming from publicly available shipment documents that people are deciphering part numbers and matching them to manufacturers identifications for parts, not some insider who may or may not be a real one, or random screen snip from random website leaking news.
 
I can't really imagine back touch pads making much sense, though (especially on joy-cons where space is limited), it's probably just buttons. Apparently the details came from third party peripheral makers actively touching the device but not seeing it, so it'd be something they'd notice through touch alone, which touch pads would probably wouldn't be.

Any extra functionality on the back would be fantastic, but I'm gonna go with Occam's Razor and say that simple buttons at the back would be the easiest and cheapest for Nintendo to add to the controller, and with the most actual reliable use-cases.
First I just want to say this isn't the case, touchpads would absolutely have more various usecases and are pretty reliable tech, they could even, yes, act as buttons.

Second, some of these are arguments FOR them to be touch pads - limited space means no touchpad on the front. A touchpad would also be thinner than a button, not thicker.

Third, ¿por qué no los dos? It could be a capacitive pressure sensitive button without necessarily taking up more space than a button alone.
 
😯 how do you make these, thats amazing

Just a simple python script 😂 , I just copied your data to a .csv file with tab separated columns.
I didn't spend much time on it lol, but I'll share some sloppy code..

Python:
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

df = pd.read_csv("switch.txt",sep="\t", header=None, names=["TFLOPS", 
                                                "CLOCKS",
                                                "QuestionMark",
                                                 "SEC5", "SEC8"])
plt.style.use('ggplot')

fig,axs = plt.subplots(1)

axs.plot(df.QuestionMark, df.TFLOPS, 'o-')
axs.plot(df.SEC5, df.TFLOPS, 'o-')
axs.plot(df.SEC8, df.TFLOPS, 'o-')

for x, y, text in zip(df.QuestionMark-2.5, df.TFLOPS, df.CLOCKS):
    plt.text(x, y, text, alpha=0.3)

axs.set_xlim(0, 35)
axs.set_ylabel("TFLOPs (FP32)")
axs.set_xlabel("Estimated Power consumption [W]")

plt.legend(['???', 'SEC5', 'SEC8'])

plt.suptitle("Estimated Power Consumption T239 [12SM GPU]", y=0.99)
plt.title("Marker labels: GPU Clock Speed [GHz] \nData taken from: 7hatDeadCat & Thraktor", fontsize=8)

plt.grid(True)

plt.savefig("t239-estimated-powerFreq-curve.png")

plt.show()
 
Yea that's exactly what I was expecting. Still hoping it'll be $350 max 😬
If the inflation landscape remains the same (and also weak yen) I would expect something like $430, €399 and ¥39000, considering that US has less problems with the price than Europe (and Nintendo would asbolutely avoid having more than ¥40000 in Japan imo)
 
I keep seeing all over the internet today "Switch 2 will be around PS4 level in handheld mode"

Significantly faster CPU, significantly more RAM, significantly faster RAM, significantly faster I/O, GPU architecture and feature sets that are multiple generations newer, etc.

I don't see how these comparisons are made. It's going to be much more capable than a PS4 in handheld mode.
Clocks will probably be lower in portable mode due to battery and heat.
 
0
Just a simple python script 😂 , I just copied your data to a .csv file with tab separated columns.
I didn't spend much time on it lol, but I'll share some sloppy code..

Python:
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

df = pd.read_csv("switch.txt",sep="\t", header=None, names=["TFLOPS",
                                                "CLOCKS",
                                                "QuestionMark",
                                                 "SEC5", "SEC8"])
plt.style.use('ggplot')

fig,axs = plt.subplots(1)

axs.plot(df.QuestionMark, df.TFLOPS, 'o-')
axs.plot(df.SEC5, df.TFLOPS, 'o-')
axs.plot(df.SEC8, df.TFLOPS, 'o-')

for x, y, text in zip(df.QuestionMark-2.5, df.TFLOPS, df.CLOCKS):
    plt.text(x, y, text, alpha=0.3)

axs.set_xlim(0, 35)
axs.set_ylabel("TFLOPs (FP32)")
axs.set_xlabel("Estimated Power consumption [W]")

plt.legend(['???', 'SEC5', 'SEC8'])

plt.suptitle("Estimated Power Consumption T239 [12SM GPU]", y=0.99)
plt.title("Marker labels: GPU Clock Speed [GHz] \nData taken from: 7hatDeadCat & Thraktor", fontsize=8)

plt.grid(True)

plt.savefig("t239-estimated-powerFreq-curve.png")

plt.show()
Matplotlib my beloved. That's my shit
 
Meant to ask earlier - where did you find numbers for the first curve (the one labeled ??? in the graph)? Was it from Thraktor's writeup? Wondering what the first curve is but I assume you're not sure thus the ??? label?
actually no thats entirely my doing. thraktor estimated the SEC5nm formula as 73% of 4132, so i wondered what 73% of 3016 would be, which gives me 2201. its a little more than half the power consumption of SEC8nm, which is just a silly lil "what if" for something lower than SEC5nm. i didnt know what to call it and i dont want to say "this is my TSMC4nm" cus eh its nowhere near as well thought out as thraktors work
 
With the hardware available, it can literally look like this:
Horizon_Forbidden_West_20220204150731_copy.jpg

god-of-war-ragnarok-ps4-performance.jpg

EdQnDdlWAAIV3R-

(PS4 Pro screenshots)

Obviously Zelda will have a different art style, but yeah it has the potential to look damn good.

Would be nice to have more confirmation of switch 2 being PS4 Pro level in just raw power alone, so then I could watch videos of Horizon on PS4 Pro in anticipation of the switch 2 instead of just watching Horizon on base PS4 :sneaky:
 
Back Paddles are patented by SCUF. Microsoft pays them to use it on Elite controller. I doubt Nintendo would pay some dollars to be able to use it on Switch, but I may be wrong.
Apparently not back buttons, though - hence the third-party controllers conspicuously not using the SCUF-style paddles, and also the Steam Deck.

Where did you hear this? I was under the impression that the new buttons (if the accessory manufacturer leaks were even real) were extra side buttons.
The Mobapad report quite specifically says there's a back button and an extra button on the front on each joy-con, complete with images showing specific locations.



First I just want to say this isn't the case, touchpads would absolutely have more various usecases and are pretty reliable tech, they could even, yes, act as buttons.

Second, some of these are arguments FOR them to be touch pads - limited space means no touchpad on the front. A touchpad would also be thinner than a button, not thicker.

Third, ¿por qué no los dos? It could be a capacitive pressure sensitive button without necessarily taking up more space than a button alone.
To be fair, it'd have to be well-designed to act as both a trackpad and a button. You don't want to press it accidentally, for example. I wasn't really a fan of the Vita's back pad, for example, but we'll see. But current reporting heavily suggests they'll be back buttons.
 
Just a simple python script 😂 , I just copied your data to a .csv file with tab separated columns.
I didn't spend much time on it lol, but I'll share some sloppy code..

Python:
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

df = pd.read_csv("switch.txt",sep="\t", header=None, names=["TFLOPS",
                                                "CLOCKS",
                                                "QuestionMark",
                                                 "SEC5", "SEC8"])
plt.style.use('ggplot')

fig,axs = plt.subplots(1)

axs.plot(df.QuestionMark, df.TFLOPS, 'o-')
axs.plot(df.SEC5, df.TFLOPS, 'o-')
axs.plot(df.SEC8, df.TFLOPS, 'o-')

for x, y, text in zip(df.QuestionMark-2.5, df.TFLOPS, df.CLOCKS):
    plt.text(x, y, text, alpha=0.3)

axs.set_xlim(0, 35)
axs.set_ylabel("TFLOPs (FP32)")
axs.set_xlabel("Estimated Power consumption [W]")

plt.legend(['???', 'SEC5', 'SEC8'])

plt.suptitle("Estimated Power Consumption T239 [12SM GPU]", y=0.99)
plt.title("Marker labels: GPU Clock Speed [GHz] \nData taken from: 7hatDeadCat & Thraktor", fontsize=8)

plt.grid(True)

plt.savefig("t239-estimated-powerFreq-curve.png")

plt.show()
o_O yeh i aint smart enough to do that, thank you! i love the graph
 
0
If the inflation landscape remains the same (and also weak yen) I would expect something like $430, €399 and ¥39000, considering that US has less problems with the price than Europe (and Nintendo would asbolutely avoid having more than ¥40000 in Japan imo)
Due to the inclusion of taxes in European pricing, and exclusion from American pricing, I expect the price to be €399/$399 or €449/$449.

Also ¥39000 is ≈ $250, there's no way there's a 200USD discount for Japan. While the economy is different, I could see as much as ¥69000 in Japan, but I think ¥59000 is more likely.

There's no likely path for Nintendo to make anything but a miserable loss every unit trying to sell the device at ¥39000 when production and shipping is broadly paid for in USD.

¥39000 yen is currently €230 - the price of a Nintendo Switch Lite in Europe.
 
Last edited:
Question, maybe for the more technically inclined, maybe for someone who knows Japanese, the phrase "further-reduced model" was used in reference to Nintendo Switch (Generation 1). Did this mean a reduced model of business leading into and throughout the life of the Switch next, or an admission of an additional minor piece of cost-reduced Nintendo Switch hardware?
The only wording I see similar to this is "later-reduced model" in Q7 of the google translation of the Q&A. If that's what you're referring to, it's a mistranslation and should say successor model.
 
The sequential read of the SSD is apparently 2100Mb/s which is only around 300mb/s slower than the Series X's. I think it'll be fine overall, don't worry too much about it.

It doesn't even need that, the UE5 demo only needed 10MB/s per frame. So an 880MB/s microSD Express will be fine for basically any modern game, let alone the internal storage.
 
Please read this staff post before posting.

Furthermore, according to this follow-up post, all off-topic chat will be moderated.
Last edited:


Back
Top Bottom