• 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!)

Reader mode
eye0h said:
He's gonna post a video from inside the France Nintendo Cube
Click to expand...
giphy.gif
 
karmitt said:
This decision making feels like fuck you money. I wish I had fuck you money.
Click to expand...

Honestly not. I had a good quarter of trading and decided to do something irresponsible. I live in the "not rich but very successful" camp. I do enough daytrading to live a comfortable life and thats it.

Leviraps said:
What could possibly convince you that it’s May? Genuinely curious.

To me it feels like the most logical thing would be a reveal in March or April and a launch in the fall. That way 3rd party devs can show off their games coming out for it in the June Direct/E3/Summer Games Fest circuit, with a release of the console in September/October
Click to expand...

Outside of the amazing research done here by fellow members, Nothing concrete. Its just me taking what they've put out there and connecting dots and combining it with rumor in my investing circles. One guy said May launch and if there was ever a MFer who has a "Fuck you" amounts of money in Nintendo stock and has inside sources -> it would be that guy. Hes a legend with some of his moves which makes me squint my eyes when he has "confidence" of a May launch. Unfortunately none of this can be "validated" as admitting to such would be a crime even though American politicians do that shit daily plus it could simply be a guess but I know a lot of investors who are preparing portfolio's on that "guess"

So this is one of those things you ignore unless it actually happens then you go "WTF..." later. :ROFLMAO:
 
oldpuck said:
What is unlikely is that there is anything new to be seen. Nintendo is unlikely to have added anything substantial in just a few months. Outside chance that some dev feels emboldened to slip some extra details in, considering it’s already semi public
Click to expand...
i mean i know it's real more even more confirmations would be cool
 
ILikeFeet said:
What evidence you have of these claims
Click to expand...

It was 2.72 ms on a 3090 at 1440p pre-RR.

famiboards.com

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

then why the Pokemon games still end up badly? GameFreak has to coordinate with TPC when making games. This is so that TPC can get the anime, merchandise, etc ready. They're on a pretty tight schedule and once that schedule has been set they can't really change it because delaying the anime...
famiboards.com famiboards.com

(One serious question I have in general about hardware RT on mobile hardware is that BVH structures usually eat up RAM so I'm curious to see how this works out with the Switch 2's expected high capacity slow RAM)
 
Last edited:
oldpuck said:
Ahhhh, okay. I suspect this is because just in time frame delivery is more flexible inside a 120hz container. I wonder what a frame pacing graph would say.
Click to expand...

That makes sense. I would like to see an in-depth article on the effect, since it DOES feel odd in a way that the framerate isn't as much of a bottleneck on input latency. In the end though it just makes me more sure that I'd want a 120hz screen on Switch 2 and 3 rather than just the ability of the screen to switch (heh) to 40hz. Especially Switch 3 since I'm becoming more convinced that thing is going to need to be able to play AAA games with path tracing - if the full path tracing revolution comes about, to reduce dev times and install sizes.
 
Pokemaniac said:
Yeah, they're not going to be getting anything of value out of that. LLMs aren't actually intelligent, by design. People need to stop falling into this anthropomorphization trap.
Click to expand...

I don't think anyone mentioned LLMs is going to become self aware or was being anthropomorphized? LLMs are exactly that, just large language models.
 
So Samsung's 3GAP (SF3) process node and Samsung's 4HPC (SF4X) process node are scheduled to start high volume manufacturing (HVM) at 2H 2024. (p. 8)
 
LiC said:
I don't think Nintendo communities were aware of shipping data websites as a possible source of information until that misinformed tweet from Connor, of all people, which you posted here when you got the whole thing started. Based on what we can see now, there likely would have been numerous ways to see things ahead of announcement in years past if we had known what to look for.
Click to expand...
Does this mean we will no longer have to rely on funcles for factory news??
 
LLMs are pretty good, GPT3.5 onward are just shitty directions taken to get hype from the media and investment money from companies.

You need to be able to customize your generative neural network for it to have much value and GPT3.5 onward is drawn to “average” output regardless of what you try to do.

The same is true for a lot of generative art stuff. They just converge to the average of shitty but liked art instead of being able to be customized to something resembling a specific artist, but this is due to choices programmers made designing them and this can be altered in the future.
 
Pokemaniac said:
The point is that they're, at best, questionably appropriate for even the basic tasks people tend to try to use them for.
Click to expand...

I think that's kind of a narrow view on LLMs in general.

LLMs have a pretty wide range of potential applications - one of those is making the NPC dialogue in our games seem more realistic and not limited to a fixed number of responses.
 
ItWasMeantToBe19 said:
It was 2.72 ms on a 3090 at 1440p pre-RR.

famiboards.com

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

then why the Pokemon games still end up badly? GameFreak has to coordinate with TPC when making games. This is so that TPC can get the anime, merchandise, etc ready. They're on a pretty tight schedule and once that schedule has been set they can't really change it because delaying the anime...
famiboards.com famiboards.com

(One serious question I have in general about hardware RT on mobile hardware is that BVH structures usually eat up RAM so I'm curious to see how this works out with the Switch 2's expected high capacity slow RAM)
Click to expand...
this isn't evidence, this is stretching assumptions beyond reasonable limits

and BVH structures are as large as the data held in them. low poly models = smaller bvh
 
lemonfresh said:
my percent chances of confidence of each month for release

January - 0%
February - 5%
March- 30%
April - 40%
May - 45%
June - 50%
July - 50%
August - 65%
September - 75%
October - 80%
November - 90%
December - 40% ( Dont see much December launches)
Click to expand...
I’d swap the September & November percentages, Nintendo is probably going to avoid launching the system on the same month as Pokémon if possible.
 
0
darthdiablo said:
I think that's kind of a narrow view on LLMs in general.

LLMs have a pretty wide range of potential applications - one of those is making the NPC dialogue in our games seem more realistic and not limited to a fixed number of responses.
Click to expand...
That sounds like a good recipe for having your NPCs say something really offensive.

The problem with LLMs is that basically the only thing they're actually good at (which is incidentally what they're actually trained to do) is generating gibberish filler text. Credit where credit is due, they are really good at doing that. Too good, in fact, to the point where it's being mistaken for intelligence and they're getting put in places where they absolutely do not belong, like user support or web search. Presenting LLM generated text to the user without a human editing it somewhere in the process is almost always a bad idea, because it's almost certainly going to generate something that's undesirable by some measure for some input without your knowledge. And that's not getting into being an outside network dependency which is both expensive and a potential point of failure.
 
Literally the only thing GPT3.5+ is amazing at is never using slurs or offensive language, what is this argument.

Putting an LLM in the game directly would eat up so many resources to make no sense, but every text heavy game within ten years will be using generative text to create more text so NPCs say more stuff. It’s just that GPT3.5+ is a step in the wrong direction compared to GPT2.

I would expect generative machine learning to be creating most art assets and animations (based on core animations and assets created by the artists) in games within 10-15 years as well.
 
Pokemaniac said:
That sounds like a good recipe for having your NPCs say something really offensive.

The problem with LLMs is that basically the only thing they're actually good at (which is incidentally what they're actually trained to do) is generating gibberish filler text. Credit where credit is due, they are really good at doing that. Too good, in fact, to the point where it's being mistaken for intelligence and they're getting put in places where they absolutely do not belong, like user support or web search. Presenting LLM generated text to the user without a human editing it somewhere in the process is almost always a bad idea, because it's almost certainly going to generate something that's undesirable by some measure for some input without your knowledge. And that's not getting into being an outside network dependency which is both expensive and a potential point of failure.
Click to expand...
LLMs is like every tool ever. Bad input = bad output. Use tool incorrectly = disastrous results. LLM like every other technology, will mature over time.

I just find it odd you brought up anthropomorphizing LLM when nobody did any such thing. I'll drop the discussion, sorry I brought it up.
 
darthdiablo said:
LLMs is like every tool ever. Bad input = bad output. Use tool incorrectly = disastrous results. LLM like every other technology, will mature over time.

I just find it odd you brought up anthropomorphizing LLM when nobody did any such thing. I'll drop the discussion, sorry I brought it up.
Click to expand...
To try to better describe my position: the uses of LLMs described in that article all involve an implicit level of anthropomorphization because an LLM would only be an appropriate tool for those tasks if it actually were intelligent.

It's why I sort of tend to bristle at mentions of LLMs in general. They're very niche tools that are managing to worm their way into extremely inappropriate places because people are anthropomorphizing them without realizing they're doing it. It's a very human bias that you have to stay vigilant of when discussing ML tools. The field often being referred to as "AI" probably doesn't help with that.
 
LiC said:
I guess I would just phrase it as, we can't draw any conclusions from this data. My biased perspective says there's evidence of nearness in there, but I think that the worst case for an objective look is just indeterminate, not behind schedule wrt the Switch's March launch. Because I think that this is what it would look like if MP were starting very soon, but I also think this is what it would look like if MP were a few months off (which is pretty much the whole range it could be, and the difference between an H1 and H2 launch).
Click to expand...
Yes 100% agree. In retrospect I wasn’t as clear as I meant. “Nintendo is a month behind the Switch schedule” isn’t a fact but one story I could make with the data.

I tend to assume H1, and have for non-public reasons for a couple months now. Just saying thay H1 isn’t a lock.
 
0
gnetc said:
I mean I find it suspicious how quiet it actually is, honestly I feel like it’s more likely with there being basically no chatter.
Click to expand...
Nintendo want you focused on Mario Wonder just now then Wario then Mario RPG. My bet is Wonder will push an absolutely enormous amount of Switch units this winter especially the OLED.

I personally think people expecting a reveal of Switch 2 this year are in for disappointment.

For me ->

Spring release -

Hype announcement - late Jan 2024.
Full reveal and line up - Feb 2024.
Released - March 2024.

OR

Winter release -

Hype announcement - late September 2024.
Full reveal and line up - October 2024.
Released - November 2024.
 
"Manage your expectations, don't set yourselves up for disappointment"
p038m442.jpg
 
Samuspet said:
For me ->

Spring release -

Hype announcement - late Jan 2024.
Full reveal and line up - Feb 2024.
Released - March 2024.

OR

Winter release -

Hype announcement - late September 2024.
Full reveal and line up - October 2024.
Released - November 2024.
Click to expand...
These are some of the most incredibly tight reveal to launch windows I've ever seen.
I agree that time is running out for people wanting it to release in the spring. We basically need to hear something this week or next week.
 
CeramicPigeon said:
These are some of the most incredibly tight reveal to launch windows I've ever seen.
I agree that time is running out for people wanting it to release in the spring. We basically need to hear something this week or next week.
Click to expand...
I think they could get away with an April or May release if they announce it early next year. Spring is doable with a similar announcement-to-launch timeframe to the Switch is they do it in January. Early January to May is about that 4 month wait the Switch saw.
 
0
Neoxon said:
VGC actually has multiple sources within the gaming industry, so the generic “sources” are mainly to protect the people who provide information to Andy & Co. for the article. The likes of them & Eurogamer are far above your average Twitter rumor mongerer.

I’m sure Andy has seen the concrete evidence himself, or was told by enough sources to be confident enough to report this. And in Eurogamer’s case, they need at least 3 sources to run an article on a scoop. Them not showing said evidence is, again, likely to protect their sources.
Click to expand...

ItWasMeantToBe19 said:
So the Switch is just not declining in Japan (while it's declining everywhere else) just for reasons and those reasons do not include the yen being historically weak?
Click to expand...
the Switch is Strong in Japan and Dominating the Software and Hardware More than any other Market and thats since the 2nd year of the console life , so you Can't Mirror the Japanese to other Markets , the Weak Yen could have an impact but not that much , and you can understand that from the great software sales .
 
0
lemonfresh said:
my percent chances of confidence of each month for release

January - 0%
February - 5%
March- 30%
April - 40%
May - 45%
June - 50%
July - 50%
August - 65%
September - 75%
October - 80%
November - 90%
December - 40% ( Dont see much December launches)
Click to expand...
I don't wanna be that guy, but a probability mass function should sum to 100%!

Personally, mine looks like:

Mar. 2024 or earlier: 1%
Apr: 2%
May: 4%
June: 7%
July: 9%
Aug: 11%
Sept: 16%
Oct: 11%
Nov: 33%
Dec: 5%
2025 or later: 1%
 
ItWasMeantToBe19 said:
Literally the only thing GPT3.5+ is amazing at is never using slurs or offensive language, what is this argument.

Putting an LLM in the game directly would eat up so many resources to make no sense, but every text heavy game within ten years will be using generative text to create more text so NPCs say more stuff. It’s just that GPT3.5+ is a step in the wrong direction compared to GPT2.

I would expect generative machine learning to be creating most art assets and animations (based on core animations and assets created by the artists) in games within 10-15 years as well.
Click to expand...
I imagine the most practical use it'd have is just reporting what language looks like swears and slurs and reporting it back for Nintendo's assessment to update their censor dictionary. Assuming it doesn't slow the game or system down that much, I mean.
 
0
When I read spring, we don’t mean “march and march only” right

Cause if you ask me spring, I’m not thinking early spring myself…. More like May.


Which, if as speculated, is enough time still from now and then. 5-6 months is good enough. Consoles don’t need 2 years or a whole year unless it is coming from someone or something that isn’t really as popular as contemporaries in the same space.

usually. If it’s from popular brand, it advertises itself to a degree and a loyal following does a lot of the rest even for the unsuspecting oblivious denizens of the world.
 
March for a launch won't happen. I think it is best if we move past that.

Would it be possible ? Sure but Nintendo will largely beat its forecasts with their current FY as is and don't need a Switch 2 boost.

They want a smooth transition, in terms of financials too, so next FY it is and I am hoping for May. March could be the date of its large showcase, just like the January one was 2 months before launch.
 
P4bl0 said:
pizza-big-bang-theory.gif


Thank you! Good to be back

Lot of stuff happened I see.. Alright let me add my bit to this
Click to expand...
Don't think the reveal could have happened without you being present. Too bad you were kicked out, we would have had the console 2 weeks ago already, with release in December. Now we gotta wait till March ;)
 
Amazing findings from @LiC and @LuigiBlood you guys are really doing great work,

I've done some investigation as well during my time off based of your findings:

I've found a lot of.. interesting things matching the timing we're currently expecting

To view the content, you need to Sign In or Register.


Realistically I hope and think we're getting the reveal tomorrow or next week. We will have all Switch consoles discounted with bundles by the end of November

The gate is wide open now.

#TeamH1 Lives
 
Last edited:
From a financial point of view, it's far better for Nintendo to release at the start of next fiscal year.
The current one is "good enough", like lelouch said it's smarter to smoothen the transition.
 
Last edited:
Reading this thread, don't comment a whole lot because I don't have much to offer that hasn't already been detailed by people more capable then myself. However I did want to say that it feels like we are getting closer to something. I have no idea what will happen but it kind of feels like the Switch successor would likely come in the first half of 2024 and not the second half but it's really just a hunch based on a lot of what I read here and Nintendo winding down Switch game development (lots of ports, remasters, remakes, smaller scale projects).

Nintendo went out of their way to show us "lots of games are still coming in 2024 for Switch" which seemed a bit odd for Nintendo to show us so many games prepared for Switch that isn't coming out until well into 2024. Usually they might show some games within the fiscal time but it's believed that at least a couple of these games might not come out until spring/summer (Paper Mario, Lugi's Mansion 2) though officially no dates were given. I kind of internally wondered if Nintendo did this in order to have less people speculating that a Switch successor was right around the corner.

Most people will see that the Switch is still getting great games with several great games planned well into 2024 while a small few like myself may speculate that Nintendo plans on releasing new hardware in the first half of 2024 but doesn't want anyone really talking about it or speculating about it until Nintendo themselves reveals it.

Anyways, I'm tired, I'm ranting, it's exciting, I think the new hardware grows closer and closer and we may finally see it sometime soon. Hopefully lol.
 
Reader mode
You must log in or register to reply here.
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