oldpuck
Koopa
- Pronouns
- he/they
The same gamers complaining about this we’re bitching about multi core CPUsMicrosoft has been asking for more NPU for AI based tech from AMD, which is bottlenecking gaming based improvements.
The same gamers complaining about this we’re bitching about multi core CPUsMicrosoft has been asking for more NPU for AI based tech from AMD, which is bottlenecking gaming based improvements.
Microsoft has been asking for more NPU for AI based tech from AMD, which is bottlenecking gaming based improvements.
Page 350 - Discussion - Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
Page 350 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.forums.anandtech.com
Microsoft has been asking for more NPU for AI based tech from AMD, which is bottlenecking gaming based improvements.
Page 350 - Discussion - Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
Page 350 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.forums.anandtech.com
Lol at people complaining about this and saying it’s useless, meanwhile ML technologies like upscaling, frame gen, and RT denoising are becoming more and more important. Nvidia has worse raster performance than AMD, but still wrecks them because of their ML hardware/software and RT performance. AMD dragging their feet on ML hardware is a major reason many current gen console games have such poor image quality and people on the PC side have been begging for a real AMD alternative to DLSS since FSR2 is nowhere near as good. AMD needs to dedicate space on their desktop cards to ML hardware and develop an AI version of FSR in addition to including this hardware on mobile stuff.The same gamers complaining about this we’re bitching about multi core CPUs
A significant issue is... you really do not want to run your Steam Deck at max speed. Aiming for 10W and locking the screen to 40Hz is common to get good battery life. Reviewers and internet debaters will run the Steam Deck at max, even above the official cap of 15W, report the high frame rates and settings, and then lower the settings and down clock the screen and SoC so they can actually enjoy the system. Specially in the LCD model.Y’all are making my point You can’t give a simple answer that is also nuanced.
The question was, will handheld Drake be more powerful than SteamDeck and the answer is absolutely not. It’s not a question.
On every single benchmark, AMD outperforms Nvidia on similarly specced hardware.
On single core performance, Steam Deck’s CPU will stomp Drake’s. It benchmarks better and it’s clocked much higher.
On multi-core performance it’s the same. SSD performance? Well ahead. Memory? Lucky if Nintendo is tied.
“How much will it matter” is a much more complicated question.
Y’all are making my point You can’t give a simple answer that is also nuanced.
The question was, will handheld Drake be more powerful than SteamDeck and the answer is absolutely not. It’s not a question.
On every single benchmark, AMD outperforms Nvidia on similarly specced hardware.
On single core performance, Steam Deck’s CPU will stomp Drake’s. It benchmarks better and it’s clocked much higher.
On multi-core performance it’s the same. SSD performance? Well ahead. Memory? Lucky if Nintendo is tied.
“How much will it matter” is a much more complicated question.
largely because we have nothing else to go on but theoreticals. until the system is released and games are compared, it's the best we gotWhy do we even compare “raw power” if “how much will it matter” is such an open question now? When does it actually matter? Does “Drake will be less powerful than Deck” actually have any practical implications?
I see Nvidia and AMD GPU performance comparisons and they’re littered with caveats like “without DLSS AMD comes out ahead” and “with DLSS and RT involved Nvidia jumps ahead by a pretty big margin.” Maybe this is important for the PC space where we do see games with poor or no DLSS and RT. But for a console that guarantees its feature set and DLSS availability, the latter statement feels like the only relevant one.
Also with Switch 2 titles being a hypothetical 1080p to 4K on TV, is such a thing even possible on Deck?
It can do FSR2/3 to resolutions over 1080p as well. But it would be lower quality and a bigger relative drain on resources.Also with Switch 2 titles being a hypothetical 1080p to 4K on TV, is such a thing even possible on Deck?
If we're ignoring IQ, then we could do even worse types of scaling for even cheaper, like FSR1. Better IQ means bigger performance gains if you don't have to do like-for-like scaling factors.DLSS isn't a panacea. it gives better IQ than FSR, but Drake will still have lower input resolutions and people don't care that much about IQ. the performance gains are the same with like for like scaling factors.
Chat we might be so back.
Not a prototype. Just the Wii U Gamepad.that a Wii U prototype not Switch sucessor
There’s no reason for it not to be able to.If drake can't turn on dlss in portable mode does that mean we're stuck with 800-900p resolution?
I meant that in comparison to other systems using FSR while Drake uses DLSS. while you can go lower in the resolution with DLSS for more performance, because we're already seeing PS5/Series games go low, those versions aren't afraid to go low either.If we're ignoring IQ, then we could do even worse types of scaling for even cheaper, like FSR1. Better IQ means bigger performance gains if you don't have to do like-for-like scaling factors.
I assume most of us will be dead before that becomes practical. 10 years is way too optimistic of a timeframe.I apologize for the question if it’s poorly worded, or if it doesn’t really make sense. Assuming that one day the game consoles, including Nintendo’s, will be replaced by cloud gaming services, depending on what hardware the games will be developed?
Would this mean the end of an integrated approach or would it simply mean that the hardware would be dematerialized for us but would continue to be developed somehow as a console? Can we imagine that Nintendo only sells a controller like Google did with the stadia but is still affected by hardware limitations including in the hypothesis of a dematerialized service?
I have a hard time imagining how this can be constantly presented as the future of the industry when I can’t imagine the costs or the revenues of such a model. I mean a switch entirely to cloud gaming is how much costs compared to developing a console, and what revenue perspective does it really offer you to push games that client will not own in fact? I have no idea. Yet I feel like I’ve been hearing from everybody for 10 years that this is the future.
The nature of internet connections makes this unviable for a long, long time.I apologize for the question if it’s poorly worded, or if it doesn’t really make sense. Assuming that one day the game consoles, including Nintendo’s, will be replaced by cloud gaming services, depending on what hardware the games will be developed?
Would this mean the end of an integrated approach or would it simply mean that the hardware would be dematerialized for us but would continue to be developed somehow as a console? Can we imagine that Nintendo only sells a controller like Google did with the stadia but is still affected by hardware limitations including in the hypothesis of a dematerialized service?
I have a hard time imagining how this can be constantly presented as the future of the industry when I can’t imagine the costs or the revenues of such a model. I mean a switch entirely to cloud gaming is how much costs compared to developing a console, and what revenue perspective does it really offer you to push games that client will not own in fact? I have no idea. Yet I feel like I’ve been hearing from everybody for 10 years that this is the future.
I don't see consoles being replaced by Cloud service like the defunct Stadia. In reality, there's much less hype around cloud gaming now than a few years ago.I apologize for the question if it’s poorly worded, or if it doesn’t really make sense. Assuming that one day the game consoles, including Nintendo’s, will be replaced by cloud gaming services, depending on what hardware the games will be developed?
Would this mean the end of an integrated approach or would it simply mean that the hardware would be dematerialized for us but would continue to be developed somehow as a console? Can we imagine that Nintendo only sells a controller like Google did with the stadia but is still affected by hardware limitations including in the hypothesis of a dematerialized service?
I have a hard time imagining how this can be constantly presented as the future of the industry when I can’t imagine the costs or the revenues of such a model. I mean a switch entirely to cloud gaming is how much costs compared to developing a console, and what revenue perspective does it really offer you to push games that client will not own in fact? I have no idea. Yet I feel like I’ve been hearing from everybody for 10 years that this is the future.
I know pretty much all of the consoles state they can handle FP16 operations, but seeing as developers actually used this function on Switch to punch above its weight. I can see Switch 2 games also using mixed precision a lot to get closer to the other systems in image quality and performance.Why do we even compare “raw power” if “how much will it matter” is such an open question now? When does it actually matter? Does “Drake will be less powerful than Deck” actually have any practical implications?
I see Nvidia and AMD GPU performance comparisons and they’re littered with caveats like “without DLSS AMD comes out ahead” and “with DLSS and RT involved Nvidia jumps ahead by a pretty big margin.” Maybe this is important for the PC space where we do see games with poor or no DLSS and RT. But for a console that guarantees its feature set and DLSS availability, the latter statement feels like the only relevant one.
Also with Switch 2 titles being a hypothetical 1080p to 4K on TV, is such a thing even possible on Deck?
Yeah, even graphics card reviewers are asking these questions. The standard way of talking about this stuff doesn't make sense anymore. But no one can quite agree on the correct replacement, and when you're talking about something like a benchmark, the whole point is it's standardized.Why do we even compare “raw power” if “how much will it matter” is such an open question now?
I’ve got an example that might be illuminating.When does it actually matter? Does “Drake will be less powerful than Deck” actually have any practical implications?
I just missed the days of how many gamecubes duct tape together as measurement of of power. But I guess "real world" use is just fine enough.Yeah, even graphics card reviewers are asking these questions. The standard way of talking about this stuff doesn't make sense anymore. But no one can quite agree on the correct replacement, and when you're talking about something like a benchmark, the whole point is it's standardized
I've mentioned it before, but there was legitimate debate about whether or not the N64 was truly next gen because it wasn't a "real" 64 bit console, and it would fail like the Jaguar.I just missed the days of how many gamecubes duct tape together as measurement of of power. But I guess "real world" use is just fine enough.
Really? I had no idea that someone even would have thought that would be the case.I've mentioned it before, but there was legitimate debate about whether or not the N64 was truly next gen because it wasn't a "real" 64 bit console, and it would fail like the Jaguar.
Technology is more interesting than fans want it to be.
Cloud gaming being the future is something mostly pushed by analysts and cloud providers with servers they want to rent out. It's not really an especially workable model.I apologize for the question if it’s poorly worded, or if it doesn’t really make sense. Assuming that one day the game consoles, including Nintendo’s, will be replaced by cloud gaming services, depending on what hardware the games will be developed?
Would this mean the end of an integrated approach or would it simply mean that the hardware would be dematerialized for us but would continue to be developed somehow as a console? Can we imagine that Nintendo only sells a controller like Google did with the stadia but is still affected by hardware limitations including in the hypothesis of a dematerialized service?
I have a hard time imagining how this can be constantly presented as the future of the industry when I can’t imagine the costs or the revenues of such a model. I mean a switch entirely to cloud gaming is how much costs compared to developing a console, and what revenue perspective does it really offer you to push games that client will not own in fact? I have no idea. Yet I feel like I’ve been hearing from everybody for 10 years that this is the future.
I don't see why not. But of course impossible to know if there's any fan lottery or other issues.Alright, time to unignore the thread! Yes, the delay/Q1 2025 news broke me. That took a while to get over lmao
Here's something I believe hasn't been asked much about the Switch 2 during these... 2628 pages, mostly because it's probably hard to actually answer?
Will the successor be as beautifully quiet as the Switch is? I ask this because I've noticed that the thing that bugs me the most with anything gaming hardware related is fan noise. I know for a fact that I prefer playing on Switch rather than PS5* because of the noise differential (even when it's only truly a problem during those rarer ultra quiet moments, such as BGM-less cutscenes).
Any way of narrowing down our possibilities? I see that the Deck for instance is quite loud (way too loud for my liking), and if I understand correctly that can be mitigated by lowering the TPD. Hence power-draw could be an indicator? I dunno I'm all ears.
*I've never checked but I'm pretty sure I got a Nidec fan, so if your PS5 is dead quiet you probably got one of the other 2 fans. And yes, for those who didn't know the PS5 has a fan lottery problem.
Cloud is not viable because it offers literally no benefits to end users and mostly drawbacks.I apologize for the question if it’s poorly worded, or if it doesn’t really make sense. Assuming that one day the game consoles, including Nintendo’s, will be replaced by cloud gaming services, depending on what hardware the games will be developed?
Would this mean the end of an integrated approach or would it simply mean that the hardware would be dematerialized for us but would continue to be developed somehow as a console? Can we imagine that Nintendo only sells a controller like Google did with the stadia but is still affected by hardware limitations including in the hypothesis of a dematerialized service?
I have a hard time imagining how this can be constantly presented as the future of the industry when I can’t imagine the costs or the revenues of such a model. I mean a switch entirely to cloud gaming is how much costs compared to developing a console, and what revenue perspective does it really offer you to push games that client will not own in fact? I have no idea. Yet I feel like I’ve been hearing from everybody for 10 years that this is the future.
And they no longer want to rent those servers for low profit game streamin as now they have the hot AI market.Cloud gaming being the future is something mostly pushed by analysts and cloud providers with servers they want to rent out. It's not really an especially workable model.
I've just finished watching Nintendo Forecasts latest video about the costs the Zelda BoTW and ToTK. He had a Miyamoto quote in it that got me thinking about if Nintendo had a minimum sales target for hardware. The quote was "... or basic premise is to create software that will sell ... 2 million ..."
While there's no equivalent of big hits subsidising sales flops, I wondered what Nintendo's minimum sales target for hardware was, that they used to judge how much they could invest in hardware. I guessed a figure of maybe 50 million units, but that's just a guess based on previous sales. Does anyone else think similar or different?
I don't remember where I got it from but the electric bill for a GPU was running 180 a month while the CPU is around 3.80.And they no longer want to rent those servers for low profit game streamin as now they have the hot AI market.
Just like cars, honestly. People always quote horsepower because it's an easy number to point to but there are so many other things that influence a car's performance (the car's overall weight, where in the rev range the engine makes its power, transmission gear ratios, wheel size/weight, etc) that cars with less horsepower can be just as fast (or faster!) than cars with more.IMO what makes a piece of hardware more "powerful" is not based solely on raw horsepower and should also take hardware features into account. I would consider Drake more powerful than Steam Deck, but it's all just semantics really. It's like SNES vs. Mega Drive again; technically the Mega Drive has an advantage in raw power (though MUCH smaller than people make it out to be), but the SNES has a lot of graphical features that give it an advantage over the Mega Drive.
T239 at 4NM with an original Switch fan would be capable of being just as quiet because at expected performance the TDP is pretty similar.Alright, time to unignore the thread! Yes, the delay/Q1 2025 news broke me. That took a while to get over lmao
Here's something I believe hasn't been asked much about the Switch 2 during these... 2628 pages, mostly because it's probably hard to actually answer?
Will the successor be as beautifully quiet as the Switch is? I ask this because I've noticed that the thing that bugs me the most with anything gaming hardware related is fan noise. I know for a fact that I prefer playing on Switch rather than PS5* because of the noise differential (even when it's only truly a problem during those rarer ultra quiet moments, such as BGM-less cutscenes).
Any way of narrowing down our possibilities? I see that the Deck for instance is quite loud (way too loud for my liking), and if I understand correctly that can be mitigated by lowering the TPD. Hence power-draw could be an indicator? I dunno I'm all ears.
*I've never checked but I'm pretty sure I got a Nidec fan, so if your PS5 is dead quiet you probably got one of the other 2 fans. And yes, for those who didn't know the PS5 has a fan lottery problem.
They most likely do have minimum sales goals though these will be dependent on circumstance & IP. The bigger issues for a Nintendo IP are lack of a company voice, resources, gameplay identity; if you are stagnating or declining then good luck convincing anyone.I think Nintendo probably doesn't have a minimum sales goal, it's just that if there's a series that averages only 300,000 copies sold over a long period of time, it risks being forced to terminate, which used to be the case with Fire Emblem, and Fire Emblem Awakening used to be built to be the last entry in the series
It's a complex issue, I think "chronically sluggish and stagnant" is a better way to describe games that didn't live up to Nintendo's expectations, not forgetting that Zelda was a game that averaged less than 5 million copies sold before BOTW, and according to the usual thinking, the series was stopped a long time ago for such a huge investment with such a small return.It was stopped a long time ago, but in fact it wasn't stopped and has survived to this day.They most likely do have minimum sales goals though these will be dependent on circumstance & IP. The bigger issues for a Nintendo IP are lack of a company voice, resources, gameplay identity; if you are stagnating or declining then good luck convincing anyone.
so Switch sucessor still have dynamic resolution? very few games that will be able to run at 720p?If drake can't turn on dlss in portable mode does that mean we're stuck with 800-900p resolution?
there would have been dynamic res anyway. and there's no reason to think it won't run DLSS in handheld mode right nowso Switch sucessor still have dynamic resolution? very few games that will be able to run at 720p?
Even in such a theoretical case... depends. It's still waaaaay more powerful than Switch, which had plenty of games over that resolution. I sure played at 1080p with the first sub-teraflop PC GPU I bought in 2010.If drake can't turn on dlss in portable mode does that mean we're stuck with 800-900p resolution?
Just FYI but this is about AMD increasing the size of their NPU to match MS requirements due to their AI PC push that will start this year.Lol at people complaining about this and saying it’s useless, meanwhile ML technologies like upscaling, frame gen, and RT denoising are becoming more and more important. Nvidia has worse raster performance than AMD, but still wrecks them because of their ML hardware/software and RT performance. AMD dragging their feet on ML hardware is a major reason many current gen console games have such poor image quality and people on the PC side have been begging for a real AMD alternative to DLSS since FSR2 is nowhere near as good. AMD needs to dedicate space on their desktop cards to ML hardware and develop an AI version of FSR in addition to including this hardware on mobile stuff.
Native 720p resolution is no longer sufficient for today's visuals, my personal expectation for switch2 portable mode is native 800-900p resolution, boosted to 1080p via dlssso Switch sucessor still have dynamic resolution? very few games that will be able to run at 720p?
pretty irrelevant what's acceptable because the wider audience isn't that fussed about it. we're even seeing 720p in PS5 and Series X games and 540p in Series S games. drake in portable will fare far better with these resolutionsNative 720p resolution is no longer sufficient for today's visuals, my personal expectation for switch2 portable mode is native 800-900p resolution, boosted to 1080p via dlss
720p with DLSS to 1080p is... Pretty great, though? Especially on a 7.91" display, games could probably get away with 480~540p before DLSS if they're demanding. Heck, some late generation games might just have to sin, and do a 360p render with 720p output mushed up onto a 1080p display. I'd rather it be blurry than not be there at all.Native 720p resolution is no longer sufficient for today's visuals, my personal expectation for switch2 portable mode is native 800-900p resolution, boosted to 1080p via dlss
Some games will be 1080p native, others will be 900p - 800p DLSS to 1080p, some might be 720p, 540p or even 360p and DLSS to 1080p.Native 720p resolution is no longer sufficient for today's visuals, my personal expectation for switch2 portable mode is native 800-900p resolution, boosted to 1080p via dlss
The expectation should be that the days of games running at below native res are over because rendering all the pixels by itself kinda becomes irrelevant with an ML upsampler.
will Nintendo be able to keep it release cadence on it next console?