Shogmaster
Sloppy seconds
If SteamDeck was PC gaming's main dev platform, it would be true.That kind of talk is nonsense really. It's like saying the Steamdeck holds back the RTX 5090.
If SteamDeck was PC gaming's main dev platform, it would be true.That kind of talk is nonsense really. It's like saying the Steamdeck holds back the RTX 5090.
I am well aware, but it still isn't beating the Steam Machine let alone PS5 in GPU compute. Yes, it will effectively produce better image in many cases due to newer advancements. If they can do slightly higher clocks, it could reach 7 Teraflops.Don't do FLOPs comparison between different generations. Doesn't work.
That's exactly what PS6S is, the Canis chip that goes in the handheld. 16 CU RDNA5 GPU, 24 GB LPDDR5X RAM, 4 Core Zen 6C CPU plus 2 Zen 6LP cores for OS.If Sony is already making a PS6 Handheld, then couldn't a PS6S just be the handheld spec as a console?
What would be bad with that if devs already have to account for it?
That's a 16x upscale which is very hard to do without image artifacts/blur. Upscale cost is also based on output resolution, so if we assume 1080p upscale FSR5/PSSR3 is around 2ms on Canis then 4K upscale would be about 8ms.I've been speculating that a "docked Canis"/"PS6S" could be done by just using the extra power to override the upscale factor of PSSR2/3 by 2x so if a game targets 1080p on the Portable itll automatically support a docked mode in 4K (of course remains to be seen if a 540p -> 4K can even look okay, but that's just an idea), would that not be a reasonable solution to make sure a PS6S does not even need developer interaction?
7 TeraFLOPs mean exactly what? I mean RDNA 2 in consoles is doing single issue FP32 and later are doing dual so now what?I am well aware, but it still isn't beating the Steam Machine let alone PS5 in GPU compute. Yes, it will effectively produce better image in many cases due to newer advancements. If they can do slightly higher clocks, it could reach 7 Teraflops.
Canis = 16 CU RDNA 5
Series S = 20 CU RDNA2
Steam Machine = 28 CU RDNA 3
PS5 = 36 CU RDNA2
Xbox Ally X = 16 CU RDNA 3.5
Why would Sony divert chips for the handheld to try and sell a second home console at the same time they have have a main home console(PS6 Orion) in a mobile first world? Its not like TSMC has enough capacity for all these chips. It will just lead to all sorts of marketing issues, production line and developer issues. I think Sony is focused on one mobile device offering 1080p gaming on the go at 30 fps and one next gen home console running games at 4k 30- 60 fps. That is much much easier especially considering that there will likely be a midgen pro consoleThat's exactly what PS6S is, the Canis chip that goes in the handheld. 16 CU RDNA5 GPU, 24 GB LPDDR5X RAM, 4 Core Zen 6C CPU plus 2 Zen 6LP cores for OS.
People assumed Sony wouldn't mandate handheld support, but they will 100% mandate support for a Home Console. Canis will have mandatory support regardless of what happens.
How? They don't build games to run only on high end cards. Games are still made to be scalable with cards weaker than PS5.
Because Orion is too expensive, and Handheld won't sell as much as a cheap home console could. You take away the screen, battery, controller components of a handheld, simply put the internals inside a small TV box, that makes for a decent entry level console. Good enough for the casual userbases.Why would Sony divert chips for the handheld to try and sell a second home console at the same time they have have a main home console(PS6 Orion) in a mobile first world? Its not like TSMC has enough capacity for all these chips. It will just lead to all sorts of marketing issues, production line and developer issues. I think Sony is focused on one mobile device offering 1080p gaming on the go at 30 fps and one next gen home console running games at 4k 30- 60 fps. That is much much easier especially considering that there will likely be a midgen pro console
They might if they end up with lots of binned Canis that blow past their wattage budget? But yields on 135mm² SoC can't possibly get tgat bad. LOLThis PS6 lite is completely made up by MLID, it wont happen, doesnt make technical or marketing or economic sense.
Huh, I always assumed Sony would mandate Canis support.That's exactly what PS6S is, the Canis chip that goes in the handheld. 16 CU RDNA5 GPU, 24 GB LPDDR5X RAM, 4 Core Zen 6C CPU plus 2 Zen 6LP cores for OS.
People assumed Sony wouldn't mandate handheld support, but they will 100% mandate support for a Home Console. Canis will have mandatory support regardless of what happens.
Hmmm, Its not going to happen because most displays(TVs) now are 4K, for a home console you need it to target 4K unless you think they are going to use Canis to run next gen games at 4K which I highly highly doubt. Sony will just have anyone that wants mobile gaming or cheaper entry into next gen get a 1080p mobile device. A home console using hw meant for 1080p gaming is much much worse than the Series S situation. It will just be a marketing, economic nightmare. And basically it will get zero development support since devs will first build on the base PS6 devkit then get the game running in "low power" mode for the portable targeting 1080p at 30 fps. The primary development hw for the whole gen is the base PS6 console, the idea is since we live in a mobile first world, the portable will sell well so if it doesnt sell well what would incentivize devs to keep on developing for the "PS6 lite"? Unless the PS6 lite outsells the base PS6 Orion which would be another issue if it happened. But looking at how poorly the Series S performed I wouldnt count on consumers buying a PS6 lite that cant even run 4K games over the PS5 proBecause Orion is too expensive, and Handheld won't sell as much as a cheap home console could. You take away the screen, battery, controller components of a handheld, simply put the internals inside a small TV box, that makes for a decent entry level console. Good enough for the casual userbases.
I think this is the gap in some of MLID's analysis, he underestimates the Software Development angle. So I'm not surprised he's pushing the PS6 lite angle but thats just his view not an actual product Sony is going to ship. Because Sony has to look at market segmentation, production lines, software hardware support, consumer trends. Releasing two consumer products at once is already going to be a tougher job. 3 at the same time is asking for it. I dont think developers would be very enthusiastic about 3 new Sony devices unless the requirements are super sharp and barely change. Which never happens on any serious project. Its not as easy as just putting the Canis APU in a home box and then its done.Canis is built with specialized low power libraries, it can't run at high clock speeds no matter how much power you throw at it.
Hmmm, Its not going to happen because most displays now are 4K, [continues without mentioning Switch2]
If the Series S came with ~16 gb of ram it would've been fine.Another Series S situation
Fucking hell man what a terrible idea
Exactly. Canis is more powerful than Switch 2, Series S, more powerful than Xbox Ally X, almost as powerful as the delayed Steam Machine but more advanced upscaling/raytracing. Only thing beating Canis other than the big consoles would be Magnus AT4 based handhelds or entry level consoles with 24 CUs, but they won't reach its price points.Switch 2 exists
That kind of talk is nonsense really. It's like saying the Steamdeck holds back the RTX 5090.
If SteamDeck was PC gaming's main dev platform, it would be true.
Even with 5M base, SteamDeck is more like PC dev afterthought. LOLSteamDeck is fully optional. PS6 HH may or may not be forced on devs by Sony.
350 to 550 is quite the gap, lmao. So basically, this MLID doesn't know shit.
They'll be PS6, why would they move away from that?Besides the fact that this S version is purely MLiD speculation (which he declares in his videos), Sony dont have to name these PS6 anything. They can go with something like 'PS GO' and 'PS GO @Home' or some such.
ANYWAYS
I do wonder if Sony could just zap the shit out of Canis with wattage and clock the GPU like 2.2 GHz or more. 1.6GHz docked seems limited entirely by the chassis DIM.
Even with a modest heatsink and fan, you could cool around 40W easily for a 3nm SoC. Thats more than double of 15W TBP alloted for the gandheld. That would get you much better image quality for 4K TVs it would be stuck to.
Until Sony says so we don't know either way, and I don't think it will have mandatory support I think they'll leverage what they have done with Portal to support all games.People assumed Sony wouldn't mandate handheld support, but they will 100% mandate support for a Home Console. Canis will have mandatory support regardless of what happens.
![]()
According to MLID, Sony might be working on a cheaper PS6S type console alongside the main PS6.
The idea is basically a cut down version of the same APU used in the PS6 handheld, just without the screen and battery. So more like a small traditional living room console focused on efficiency and cost rather than high end performance.
If this turns out to be real, pricing could land somewhere around $350 to $550, making it the entry level option in a possible three tier setup. PS6 as the premium machine, a handheld in the middle, and this budget box at the low end.
Honestly it feels like Sony does not really have much of a choice here. If they want to get back to something like 100 million+ consoles sold again, they will probably need a cheaper entry point like a PS6S. A single high priced flagship console alone just will not be enough to reach those mass market numbers.
It feels a bit like Sony going back to a PS4 style strategy again, pushing aggressive pricing to expand the ecosystem and keep momentum going.
I am well aware, but it still isn't beating the Steam Machine let alone PS5 in GPU compute. Yes, it will effectively produce better image in many cases due to newer advancements. If they can do slightly higher clocks, it could reach 7 Teraflops.
Canis = 16 CU RDNA 5
Series S = 20 CU RDNA2
Steam Machine = 28 CU RDNA 3
PS5 = 36 CU RDNA2
Xbox Ally X = 16 CU RDNA 3.5
What about clocks, what about RT HW?, what about ML capabilities?.. did you just focus in one stagnating element of GPUs (raster) to prove a point?
Cmon man you are better than this. Lol
They do matter alot TODAY! Look at Switch 2 (with limited ML) vs Series SRT and ML will matter much more in the future, but rirght now still 90% of games (and over 95% of console games) are raster only.
Not sure what point you are trying to make. I already said it would have superior raytracing and upscaling capabilities than all of those. But as KeplerL2 stated in last page that the device can't clock too high no matter the TDP. I mean, Canis would make a great entry level console/streaming box if price can be around $349. Sony should allow 4k/120 Orion streaming on Canis, that should cover most headaches regarding mandatory support, it would have AV1 hardware decoders.What about clocks, what about RT HW?, what about ML capabilities?.. did you just focus in one stagnating element of GPUs (raster) to prove a point?
Cmon man you are better than this. Lol
They do matter alot TODAY! Look at Switch 2 (with limited ML) vs Series S
Cains will be a monster vs Switch 2 with more advanced features.
No wonder we get the shit we are getting when people like you call DLSS, FSR optimization tools. PatheticGetting used to RT tech and dedicated hardware for it, new engines coming in, a lot of the worst-optimized games being on Unreal Engine 5, plus machine learning and frame generation as optimization tools. There's a lot we can expect in terms of progress based on what we know about the next generation.
I don't believe this rumor one bit, Playstation 6 lineup should be:
PS6P: Handheld
PS6: Stationary console.
Both MUST play ALL PS6 games.
Sony can release Pro model down the line but for launch, this should be it. They shouldn't confuse developers with multiple SKUs configurations (I don't buy the scalability thing) we are talking consoles with lower level access not PCs.
Why is Sony always some kind of a mirror to xbox. Its so weird.
You do realize MLID is the one who has provided us 95% of the specs regarding Orion, Canis, and Magnus. He has seen documents that led to his speculation.How do you watch that, listen to his speculative ideas about a home console with the Canis APU (he's speculated about this many times before) and then post it on here as if it's a rumour?
Nothing is confusing for devs, it would still be two PS6 optimization targets for devs.I don't believe this rumor one bit, Playstation 6 lineup should be:
PS6P: Handheld
PS6: Stationary console.
Both MUST play ALL PS6 games.
Sony can release Pro model down the line but for launch, this should be it. They shouldn't confuse developers with multiple SKUs configurations (I don't buy the scalability thing) we are talking consoles with lower level access not PCs.
Where do you get that it would be a lot more powerful than the PS5? The handheld is roughly 60% of PS5. But the TV box could go up to 70%, and much more power efficient.I don't believe it either.
It'd be a lot more powerful than the PS5, yet cheaper than the PS5?
The only way it'd work would be if they discontinued PS5.
They don't have to sell 100m PS6s .. they would still have huge number of active PS5.It all depends on the prices.
A 599$ handheld and a 799$ console with very little room to cut prices in the future given current market realities is going to shrink their market to definetly less than 100m users.
Mainstream isn't going to buy at those prices regardless of the games.
If they want something to reach these people they need to figure it out at launch.
It makes sense to replace the old PS5 with something targeting a 399-499$ price offering similar power but with modern features.
It's easy to do that using the handheld SOC as a basis.
They don't have to sell 100m PS6s .. they would still have huge number of active PS5.
It would be using all the latest features (like RDNA5) making it a lot more efficient than the PS5 and has a lot more RAM to boot.Where do you get that it would be a lot more powerful than the PS5? The handheld is roughly 60% of PS5. But the TV box could go up to 70%, and much more power efficient.
PS5 will likely be discontinued production and sale.
Correct, but it's still not more powerful. It doesn't need to be. Most indies, AA games would probably run and look better on Canis, which compromise 96% of games anyways. And the AAA games could vary based on certain things.It would be using all the latest features (like RDNA5) making it a lot more efficient than the PS5 and has a lot more RAM to boot.
The main question is how much better would PS6 S be vs PS5? For majority of people keeping PS5 and not buying this PS6 Lite would likely be the better move unless it's a significant difference.But that would just make the more expensive handheld even more expensive. This is obviously still a rumor but I think you are missing the point. If this is true it's to be an entry level console and the cheapest option. The handheld would be priced higher then this (screen, etc).