• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

PS6S Budget Model Rumored at $350-$550 According to MLID

We're talking about Series S compute and bandwidth here, and at 720p, dropping resolution further has harsh diminishing returns on performance gains. We may be looking at those games being completely unplayable on Canis.

If we're getting games running at 720p on a base PS5/Pro, then the issue is most likely going to stem from raytracing, which means that's even less of an issue for canis

And no, we're not talking about Series S bandwidth, this has been explained to you multiple times.
 
Last edited:
If we're getting games running at 720p on a base PS5/Pro, then the issue is most likely going to stem from raytracing, which means that's even less of an issue for canis

And no, we're not talking about Series S bandwidth, this has been explained to you multiple times.

Supposedly the Switch 2 has superior architecture for RT than Series S and is close in compute, yet it still gets soundly beaten by the junior Xbox console in Star Wars Outlaws and AC Shadows. I expect the base PS5 to best Canis in raytracing performance for Unreal Engine games.
 
Last edited:
We're talking about Series S compute and bandwidth here, and at 720p, dropping resolution further has harsh diminishing returns on performance gains. We may be looking at those games being completely unplayable on Canis.

p28Hm0M.jpeg


F9FvtSo.jpeg
 
Supposedly the Switch 2 has superior architecture for RT than Series S and is close in compute, yet it still gets soundly beaten by the junior Xbox console in Star Wars Outlaws and AC Shadows. I expect the base PS5 to best Canis in raytracing performance for Unreal Engine games.

Yeah, no lol. Switch beats S in raytracing in Outlaws. The raytracing in shadows is pretty much inconsequential for both of them.
 
Last edited:
"Fortunately we have a product for people who don't want to pay the $1000 for a PS6, it's called PS6S"

you know if this thing exists, that's the exact PR line they're gonna use 😂
 
Last edited:
Xbox Series S is by far the most-seeling Xbox Series.
This is 100% not true. At best its split 50/50. Series X always outsold Series S when there was supply. MSFT produced more Series S consoles at first thats why there were more of them initially but Series X has always had more demand which explains the 50/50 split. It really didnt sell well


But you agreed..... you don't agree on the timing? I was thinking stop PS5 Pro production in September 2027, stop PS5 production in December 2027, launch new consoles in November 2027. Then next 6 months the hardcore buy the consoles while they also clear out PS5 inventory. Then release the Canis TV box in summer 2028 along with upgrading PS+ Premium to Orion streaming.

And the Pro consoles are always the first to cease production.
You correctly identified an issue related to the growing prices of consoles but then trying to fix that with a 1080p gaming box that costs $399 isnt going to work. 1080p gaming will only fly on a handheld. There isnt going to be a Canis Home console, especially when the PS6 portable itself could connect to a TV display. Sony doesnt need to create a separate home console using the same APU they designed specifically for mobile gaming. MLID just made a suggestion like he always does without thinking it through. Its really not going to happen

And also there is another reason why the home console Canis box isnt going to happen, there is a scarcity of fab capacity, diverting mobile chips that will sell higher for a second home console releasing at the same time as the main home console which most consumers are going to demand is a recipe for disaster. It will be worse than the Series S situation. most consumers will either buy the portable or the PS6 or stick to their PS5 or PS5 pro than playing 1080p games on a Canis box for $499.
 
Supposedly the Switch 2 has superior architecture for RT than Series S and is close in compute, yet it still gets soundly beaten by the junior Xbox console in Star Wars Outlaws and AC Shadows. I expect the base PS5 to best Canis in raytracing performance for Unreal Engine games.
No. it won't.

Cannis could actually be as much as 2x better than the base PS5 in RT. And that's before even taking into account that Cannis would be running all its games at a lower internal resolution. So when a game runs at 1080p on PS5, the same game would run at like 540p or less on Cannis. You keep ignoring that Cannis has and will use PSSRx...

A lot of this has been explained to you already; all the evidence is laid out for you to see, you are just refusing to see it. It's not even hard to find videos of games running at as low as 360p and upscaled to 1080p using DLSS (an ML-based upscaler like PSSR), and see for yourself how it looks and performs. But rather than just do that, you are here incessantly trying to do all this fear-mongering.
 
If there's a legit $400 entry point for next-gen, that could be huge for casual players. Not everyone is going to drop $700+ on a flagship console.
 
PS6 already has conservative specs (power draw, memory bandwidth, number of CUs etc.).
are you mad?

It has 30GB of Ultra Premium High Speed Memory.

The market will not allow it to exist at affordable mainstream prices regardless of how much Sony compromises on power consumption and such.
 
Last edited:
Where are you getting this from? I've never heard this rumor before
Purely speculative from me.

It's 41-44$ on the BoM. If you looked at the Bill of Materials for a PS5/PS5 Pro, you'd immediately identify the controller as one of the possible items to cut to lower cost.

Though it is possible that Sony believes that selling a second controller to everyone that needs it by requiring DS6 on PS6 might out weigh that.
 
If there's a legit $400 entry point for next-gen, that could be huge for casual players. Not everyone is going to drop $700+ on a flagship console.
Whats the point in investing 400 dollars and only be "next gen" in name ... since cross-gen games are probably the norm, just stay at the ps5.

Like why buy a Ps6 S (for shit) if you are going to play games at 30 fps or 60 fps with heavy limitations just like the ps5 ? ...in fact just like it happened with the series S vs Series X

In the case of series S crossgen was not that present so in part was justified, now with crossgen being so prevalent, it seems just pointless a false "next gen" console.
 
Now would be the time to cancel the PS6 altogether, skip it until 2030 at least.
Make a small form factor handheld PS5 in 2028 (small like a Vita at most), and call it a day.

Big deck-sized handhelds are okay, but what I really miss is the PSP. And the PS5 barely got any game it feels like, it can hang on for a long while yet.
 
Last edited:
Xbox Series S is by far the most-seeling Xbox Series. Just because his price
The Xbox Series X is a failure, it was sell about maybe 8-9 millions...
That's why Microsoft was so agressive with the parity and force developpers to optimise their game for Series S.

Let's be real here, the only damn reason as many people chose the series s as they did was because for the first 2 or so years of this gen you couldnt hardly find a ps5 or series x at retail price.
The choice for a lot of people wasn't between a $300 series s and a $500 ps5/series x - it was a choice between a $300 series s and a $700 - $900 ps5/series x from ebay/facebook marketplace until the supply situation at retail got normalized somewhat.
 
The whole point of something like a Series S is next gen architecture but targeting ~1080p to offer a lower cost option. You don't achieve that with just extending a last gen system. Last gen consoles could not do what Series S did and not just because of the SSD but the CPU in particular.

Honestly I think it's going to be hard for Sony to just release a ~$750 system and then hope for the best. To get a ~100M install base they are going to need a lower cost option.
 
Last edited:
They need the PS6 because they botched the PS5 Pro.
I disagree. The Pro has been fine and really starting to step up its game. I am not sure what people's expectations were. It delivered what I expected, pretty much PS5 quality at performance settings. A Pro model is never going to be a generational leap.
 
Let's be real here, the only damn reason as many people chose the series s as they did was because for the first 2 or so years of this gen you couldnt hardly find a ps5 or series x at retail price.
The choice for a lot of people wasn't between a $300 series s and a $500 ps5/series x - it was a choice between a $300 series s and a $700 - $900 ps5/series x from ebay/facebook marketplace until the supply situation at retail got normalized somewhat.
yes thats why the split last time I checked was 51% Series X and 49% Series S. MSFT produced so many Series S boxes and they didnt sell out.
 
The whole point of something like a Series S is next gen architecture but targeting ~1080p to offer a lower cost option. You don't achieve that with just extending a last gen system. Last gen consoles could not do what Series S did and not just because of the SSD but the CPU in particular.

Honestly I think it's going to be hard for Sony to just release a ~$750 system and then hope for the best. To get a ~100M install base they are going to need a lower cost option.
Series S was marketed as a 1440p 60 fps device which turned out to be false advertising, it wasnt marketed as a 1080p device and even then demand for Series X and PS5 was much higher despite Series S being available during launch. It was always available while the other two sold out. Consumers generally want to buy 4K capable machines for their home TV. Now imagine trying to market a 1080p device for gaming at home on the TV in 2027 and beyond. Consumers would rather buy a PS5 or PS5 pro if they cant buy the PS6. Or PS6 portable to play next gen games on the go but at 1080p
 
So the S console will just be the PS6 handheld but in a TV box format, with presumably slightly higher abilities due to the form factor and thermals compared to the handheld.

They're making the TV-only Switch that people wanted Nintendo to make.
 
Series S was marketed as a 1440p 60 fps device which turned out to be false advertising, it wasnt marketed as a 1080p device and even then demand for Series X and PS5 was much higher despite Series S being available during launch. It was always available while the other two sold out. Consumers generally want to buy 4K capable machines for their home TV. Now imagine trying to market a 1080p device for gaming at home on the TV in 2027 and beyond. Consumers would rather buy a PS5 or PS5 pro if they cant buy the PS6. Or PS6 portable to play next gen games on the go but at 1080p

"Consumers don't want this - option -" is always an odd argument to me....
 
Last edited:
So the S console will just be the PS6 handheld but in a TV box format, with presumably slightly higher abilities due to the form factor and thermals compared to the handheld.

They're making the TV-only Switch that people wanted Nintendo to make.
And they'd be making a mistake if they do that, imo. The 15 watt Canis would be ok churning out game for a 1080p mobile screen, but it would show its weakness in a 4K TV, let alone that 6-core cpu probably allowing only 30 fps.

A Canis TV only could probably sell for $399. But if Sony can pull an Orion-based entry level console at $499, there will be no need to entertain the idea of Canis-based entry level TV-only console.

If Sony make the Orion-based console with conservative RAM of 20GB and SSD capacity of 512GB, the total cost based on MLID's estimation would be $571.75. Now, remove the dualsense controller and you reduce it further by $40.

At BoM of $531.75, Sony could take a hit selling it at a loss for $499 without disc drive.

An Orion-based entry level is leaps and bounds better than a Canis-based console. With an Orion-based console at $400, a TV-only Canis becomes moot.

PS6 - $499
Orion
20GB RAM
512GB SSD

PlayStation Go - $499
Canis
18GB
1GB SSD
 
MLiD is a total fraud! There, I said it! A shamster and a shill! Always wrong. Spews technobabble but understands nothing about graphics technology or videogame development. I'm sick of listening to it and people quoting it. Just stfu. What a clickbait loser, just like most of the influencers on youtube. Toxic assholes.
 
And they'd be making a mistake if they do that, imo. The 15 watt Canis would be ok churning out game for a 1080p mobile screen, but it would show its weakness in a 4K TV, let alone that 6-core cpu probably allowing only 30 fps.

A Canis TV only could probably sell for $399. But if Sony can pull an Orion-based entry level console at $499, there will be no need to entertain the idea of Canis-based entry level TV-only console.

If Sony make the Orion-based console with conservative RAM of 20GB and SSD capacity of 512GB, the total cost based on MLID's estimation would be $571.75. Now, remove the dualsense controller and you reduce it further by $40.

At BoM of $531.75, Sony could take a hit selling it at a loss for $499 without disc drive.

An Orion-based entry level is leaps and bounds better than a Canis-based console. With an Orion-based console at $400, a TV-only Canis becomes moot.

PS6 - $499
Orion
20GB RAM
512GB SSD

PlayStation Go - $499
Canis
18GB
1GB SSD
Yeah if they really feel the need for a lower end SKU it makes sense to use Orion.

Though I would do a little bit different than what you suggested:

6-core CPU cluster
16 WGP GPU
-10% clock speed on CPU/GPU
128-bit bus with 24GB RAM

This basically guarantees near 100% yield rate for the SoC, reduces the BOM by $60 for the RAM, plus another $20-30 for the reduced board/cooling requirements. Shrinking the SSD to 512GB would be another big BOM reduction but I think they could only get away with that if they forced devs to use NTC.
 
Yeah if they really feel the need for a lower end SKU it makes sense to use Orion.

Though I would do a little bit different than what you suggested:

6-core CPU cluster
16 WGP GPU
-10% clock speed on CPU/GPU
128-bit bus with 24GB RAM

This basically guarantees near 100% yield rate for the SoC, reduces the BOM by $60 for the RAM, plus another $20-30 for the reduced board/cooling requirements. Shrinking the SSD to 512GB would be another big BOM reduction but I think they could only get away with that if they forced devs to use NTC.
They will have 3 skus instead of 2. That devs will have to optimise for.

Series S + Series Y + Series X. Something like that. That would be something to behold.
 
Yeah if they really feel the need for a lower end SKU it makes sense to use Orion.

Though I would do a little bit different than what you suggested:

6-core CPU cluster
16 WGP GPU
-10% clock speed on CPU/GPU
128-bit bus with 24GB RAM

This basically guarantees near 100% yield rate for the SoC, reduces the BOM by $60 for the RAM, plus another $20-30 for the reduced board/cooling requirements. Shrinking the SSD to 512GB would be another big BOM reduction but I think they could only get away with that if they forced devs to use NTC.
Wouldn't make sense having two products so close in power and price on the market at the same time, unless they're planning on discontinuing the PS5 straight away.
 
Yeah if they really feel the need for a lower end SKU it makes sense to use Orion.

Though I would do a little bit different than what you suggested:

6-core CPU cluster
16 WGP GPU
-10% clock speed on CPU/GPU
128-bit bus with 24GB RAM

This basically guarantees near 100% yield rate for the SoC, reduces the BOM by $60 for the RAM, plus another $20-30 for the reduced board/cooling requirements. Shrinking the SSD to 512GB would be another big BOM reduction but I think they could only get away with that if they forced devs to use NTC.
How about my configuration? Would the 20GB RAM bottleneck the Orion as it is which necessitates a middling APU to make it balance?

Or is 20GB really that low for next-gen that makes a 128-bit bus 24GB a better option?

I was thinking this 20GB PS6 is basically the PS6 console. No more 30GB, until perhaps the PS6 Pro three years later.
 
Last edited:
Honestly this is a great idea. I am still using 1080p tvs. My ps5 still looks amazing. As someone who doesn't need the highest resolution, why not? Pcs are scalable so should this be. Just make sure it has enough ram. As otherwise I will be skipping ps6
No way am I paying 900+ for a console.

I get you are all rich or unmarried and don't care about price. Just moar power, but some of us aren't in the same position. I put a cap on what I spend.
 
Last edited:
How about my configuration? Would the 20GB RAM bottleneck the Orion as it is which necessitates a middling APU to make it balance?

Or is 20GB really that low for next-gen that makes a 128-bit bus 24GB a better option?

I was thinking this 20GB PS6 is basically the PS6 console. No more 30GB, until perhaps the PS6 Pro three years later.
It doesn't make sense to have 2 SKUs with the same performance but different RAM amounts. If they need to cut RAM on one SKU to reduce costs they should also reduce performance.

And having only 20GB is not enough imo, remember PS6 will be sold until 2035 at least, and have games made for it until 2040. By then even low end PCs will have 64GB RAM + 32 GB VRAM.
 
Having 20GB RAM is more than enough. Again KeplerL2 is bullshitting.

PS5 | PS5 Pro is going to be supported for first party sony games for practically the entire PS6 generation. Lets be honest here.

>20GB only makes sense within the context of running AI models while gaming(which many games will do and Sony will have to answer DLSS5). A use case beyond that doesn't really exist. Maybe developers unoptimizing but even then NTC exists.
 
Last edited:
They will have 3 skus instead of 2. That devs will have to optimise for.

Series S + Series Y + Series X. Something like that. That would be something to behold.
3 SKUs would piss off devs so much. MS can get away with 3 SKUs due to the scalable Xbox PC/Helix SKUs. Sony will just cause so much developer anger if they tried that.

Besides, we already know the chips are finalized, and there's only Canis and Orion. So Sony has no choice but to use Canis in multiple form factors including TV box.
 
By then even low end PCs will have 64GB RAM + 32 GB VRAM.
Nonesense. Complete nonsense.

Memory moore's law is even more dead than logic. Have you not gotten the memo yet? Look at Nvidia VRAMs.

PS6 adds more VRAM because it's not priced like a PS6, but like a PS6 Pro. And they crippled cooling, power delivery and will likely need to subsidize it with more expensive online.

Memory was also severely underpriced in 2025, 2026 is more or less an overcorrection. But at a baseline level, memory will never be as cheap as it was in early 2025.

It's not only nonsense, there is quite literally no reason a low end consumer device would need anything approaching that amount.
 
Last edited:
3 SKUs would piss off devs so much. MS can get away with 3 SKUs due to the scalable Xbox PC/Helix SKUs. Sony will just cause so much developer anger if they tried that.

Besides, we already know the chips are finalized, and there's only Canis and Orion. So Sony has no choice but to use Canis in multiple form factors including TV box.
Helix devices are PC pretty much. Thats why they need much stronger processor. Everything will scale, will not be as optimised I imagine.

I suspect we will see another SKU in Helix line, that will match PS5 Pro and RTX 5060Ti. And will undercut PS6 by some $100.
 
Last edited:
It doesn't make sense to have 2 SKUs with the same performance but different RAM amounts.
I get that. That's why I corrected myself suggesting that if they choose 20GB Orion to get to that sweet price of $499, then they better scrap the 30GB version until perhaps 3 years later with the PS6 Pro.
And having only 20GB is not enough imo, remember PS6 will be sold until 2035 at least, and have games made for it until 2040. By then even low end PCs will have 64GB RAM + 32 GB VRAM.
It doesn't matter when those cards are priced $2000. If expensive RAM prices are here to stay for long, then budget gaming will have to be content on budget RAM. No other way around.
 
Last edited:
No it's not, the cartel has just been profitmaxxing for the last 8 years by not adopting EUV. We are not anywhere near any kind of limit for memory scaling.
By mentioning Moore's Law I am referring to the cost component. Loading up on expensive EUV machines doesn't really help in dropping their Gb/$ costs. So yes we are.

DRAM density scaling has slowed to 2x last decade. It's not going to be even 1.5x for client this decade. It's not about to accelerate to 4x in density and cost smh.

And this is populist nonsense, how is the cartel profit maxing when Samsung was barely making meaningful operating profits until this memory cycle? They were literally bleeding because China's memory champion was memory dumping below cost and it killed the DDR4 market outright.

Memory used to be a commodity. Post AI Boom Memory is a premium product that commands premium margins. Samsung memory alone will be outdoing all of Apple in pure profits this year. This market repricing alone will wipe out the improvement in density for the next decade.

There is no improvement outside maybe speed. The party is over. 6060 will have 8 or 9GB. 12GB would push pricing considerably above where 5060 was. Apply this logic to consoles as well.

The only reason you keep going on this memory will be free bullshit is hyping up the inevitability of memory requirements increasing to hype up PS6. That will simply not happen anywhere close to the magnitude you're describing for the simple reason that AAA games are getting cancelled left and right and no one can afford to be picky about their customers in any meaningful capacity. Games will be more scaleable than ever and console generations will last longer than ever. It's simply the economic reality of this economy and semiconductor manufacturing hitting the limits of cost scaling.
 
Last edited:
If there's a legit $400 entry point for next-gen, that could be huge for casual players. Not everyone is going to drop $700+ on a flagship console.
Just under $400 at ~$375 puts a console exactly where the PS2 was in 2002 ( at $199) after adjusting for inflation.
 
By mentioning Moore's Law I am referring to the cost component. Loading up on expensive EUV machines doesn't really help in dropping their Gb/$ costs. So yes we are.

DRAM density scaling has slowed to 2x last decade. It's not going to be even 1.5x for client this decade. It's not about to accelerate to 4x in density and cost smh.

And this is populist nonsense, how is the cartel profit maxing when Samsung was barely making meaningful operating profits until this memory cycle? They were literally bleeding because China's memory champion was memory dumping below cost and it killed the DDR4 market outright.

Memory used to be a commodity. Post AI Boom Memory is a premium product that commands premium margins. Samsung memory alone will be outdoing all of Apple in pure profits this year. This market repricing alone will wipe out the improvement in density for the next decade.

There is no improvement outside maybe speed. The party is over. 6060 will have 8 or 9GB. 12GB would push pricing considerably above where 5060 was. Apply this logic to consoles as well.

The only reason you keep going on this memory will be free bullshit is hyping up the inevitability of memory requirements increasing to hype up PS6. That will simply not happen anywhere close to the magnitude you're describing for the simple reason that AAA games are getting cancelled left and right and no one can afford to be picky about their customers in any meaningful capacity. Games will be more scaleable than ever and console generations will last longer than ever. It's simply the economic reality of this economy and semiconductor manufacturing hitting the limits of cost scaling.
Very meaningful insight. So budget gaming will have to be constraint with budget RAM amounts moving forward.

I remember though that you think a Canis-based console should be in order to hit a critical price point ($399-$499) for entry level PS6.

I suggest an Orion-based console with only 20GB RAM for $499. What's your take on it? No more 30GB PS6. Just two machines: PS6 and the handheld.
 
Last edited:
Top Bottom