Support NeoGAF

[Shtof]

To clarify the topic: Would it be possible for Sony to create a handheld, that with minimal/low-effort porting could be adapted to play most PS5 games?

Clearly, this was something Sony already wanted to do with PSP/Vita (align with PS2/3), but it was too difficult because of different hardware architectures.
However, with hardware architectures having converged in recent years, it is now much more common to make handhelds that run software originally intended for stationary hardware.
The Switch, Steam Deck, ROG Ally and Lenovo Go are testament to this.
But making a handheld that runs most PS5 games sounds like a very daunting task, even for Sony.
I asked GPT4o to sketch what kind of hardware would be required to make a handheld PS5 with a release fall 2026 feasible.

• Custom SoC: ARM-based CPU + RDNA 3/4 GPU (~3-4 TFLOPS) with PS5-like features (ray tracing, VRS, geometry engines).
• Unified Memory: 8-12 GB LPDDR5 for shared GPU/CPU access.
• Storage: NVMe SSD with PS5-like decompression for efficient asset streaming.
• API Compatibility: Match PS5's GNM/GNMX APIs for minimal code changes.
• Tempest Audio: Lightweight version for consistent audio.
• Display: 720p native with scalable output to 1080p.
• Input Parity: Built-in DualSense features for seamless control adaptation.
• Devkit Tools: Asset scaling and performance profiling for manual optimization.

 

[TintoConCasera]

Battery life would most likely be a problem unless they pack in one with a LOT of juice.

 

[Kumomeme]

before, it is impossible to do with feasible hardware price.

but with PSSR it might opened the possibilities on smaller and lower handheld screen resolution.

 

[Shtof]

Battery life would most likely be a problem unless they pack in one with a LOT of juice.

I've noticed on my gaming laptop that DLSS seems very hardware efficient.
So PSSR should help a lot with battery life probably.

 

[Fafalada]

Clearly, this was something Sony already wanted to do with PSP/Vita (align with PS2/3), but it was too difficult because of different hardware architectures.

Vita like setup that has similar visual fidelity to current gen consoles already exists on the market (7840u handhelds).
What you get is devices that trade blows with a Series S (usually somewhat lower resolution, but with portable display that's fine, and fairly often better framerate owing to faster CPU/more memory), battery life of 2-6hrs, and large-ish form-factor (smaller than SteamDeck, but larger than Switch).

Other items on your list would be satisfied(or could be customized) - but I noticed you omit price targets (I suspect this would MSRP for as much as PS5 does today, in 2026).
Downsides of requiring ports are the same as they were with Vita(so it's pretty major negative) - AND you're launching to what's likely much more hostile environment to such a handheld (especially since it'll go against Switch 2 as well). So - tbh, I don't know what is the market fit for this device, other than tiny enthusiast crowds. Maybe PSVR2 like sales, on the optimistic end?

 

[Roman Empire]

Between this and a Pro, I would say this. I had all the PSP models and even the Vita at launch. It doesn't matter being weaker than the PS5, as long as it can do the same as portal but better it's already a huge win in my book. And this is a way Sony can somehow compete with Nintendo in the Asian market again, because right now Switch is King and Switch 2 will follow suit.

 

[lh032]

I think this handheld will be pricey , because it will not be considered as a portable console to fight with switch , instead it will be considered as an accessory along side with portal

 

[asdasdasdbb]

Feasible yes. The price though will be an issue. Probally released alongside the PS6 in 2027.

It'd have to have more CUs than the PS5 Pro and downclocked very far. And on a very expensive node.

 

[DaleinCalgary]

I can't see it happening. It would have to be Arm and that isn't feasible.

 

[crackajack]

with PS5-like features (ray tracing, VRS, geometry engines).

RT is something that a handheld definitely would not need. a feature that is hardly efficient on PC yet, but would be an insanely bad idea on a handheld that has to be about efficiency.

 

[coachmcguirk91]

tbh, I don't know what is the market fit for this device, other than tiny enthusiast crowds. Maybe PSVR2 like sales, on the optimistic end?

100+ million gamers having their entire library on Playstation. There is a large portion of that audience that enjoys handhelds. Pretty simple.

The PS portal has sold 700,000 and until last week it only allows you to stream games from your PS5. Now, you can stream from Xcloud and it is currently outselling the PS5 hardware. And you don't think millions will buy a native Playstation device? Get out of here

 

[RetroGamingUK]

I think this handheld will be pricey , because it will not be considered as a portable console to fight with switch , instead it will be considered as an accessory along side with portal

The point is to expand the PlayStation ecosystem. It’s meant for people who don’t have a PS5 and don’t want to stream games (PS Portal).

 

[BlackTron]

Broadly speaking there are two ways to do it. They can release a shrunken down PS5 that plays everything already available, or a different hardware that isn't as hard to squeeze, but requires a new profile to play every game.

The first solution is really hard to squeeze the hardware together and pay for. The second solution doesn't give you the PS5 library anyway so they may as well make a new platform that plays select PS5 games, ie Vita.

It wasn't a big deal for PS5 Pro to need new profiles to be Pro because it's more powerful and plays everything that came before. You simply can't release something labeled PS5 with lower power than OG hardware that needs a patch for every game to work.

 

[crackajack]

Broadly speaking there are two ways to do it. They can release a shrunken down PS5 that plays everything already available, or a different hardware that isn't as hard to squeeze, but requires a new profile to play every game.

The first solution is really hard to squeeze the hardware together and pay for. The second solution doesn't give you the PS5 library anyway so they may as well make a new platform that plays select PS5 games, ie Vita.

It wasn't a big deal for PS5 Pro to need new profiles to be Pro because it's more powerful and plays everything that came before. You simply can't release something labeled PS5 with lower power than OG hardware that needs a patch for every game to work.

PS5 remote play already needs lower resolution profiles? Isn't that enough to allow a "native" (upscaled) lower resolution device?
the cpu side would need to match ps5 power, the gpu though could scale?

 

[PeteBull]

To clarify the topic: Would it be possible for Sony to create a handheld, that with minimal/low-effort porting could be adapted to play most PS5 games?

Clearly, this was something Sony already wanted to do with PSP/Vita (align with PS2/3), but it was too difficult because of different hardware architectures.
However, with hardware architectures having converged in recent years, it is now much more common to make handhelds that run software originally intended for stationary hardware.
The Switch, Steam Deck, ROG Ally and Lenovo Go are testament to this.
But making a handheld that runs most PS5 games sounds like a very daunting task, even for Sony.
I asked GPT4o to sketch what kind of hardware would be required to make a handheld PS5 with a release fall 2026 feasible.

• Custom SoC: ARM-based CPU + RDNA 3/4 GPU (~3-4 TFLOPS) with PS5-like features (ray tracing, VRS, geometry engines).
• Unified Memory: 8-12 GB LPDDR5 for shared GPU/CPU access.
• Storage: NVMe SSD with PS5-like decompression for efficient asset streaming.
• API Compatibility: Match PS5's GNM/GNMX APIs for minimal code changes.
• Tempest Audio: Lightweight version for consistent audio.
• Display: 720p native with scalable output to 1080p.
• Input Parity: Built-in DualSense features for seamless control adaptation.
• Devkit Tools: Asset scaling and performance profiling for manual optimization.

Just wait till 2025, switch 2 is ps4pr0 docked and ps4 base in handheld mode, problem with handheld archi to be more powerful is powerdraw/batery life, and unfortunately advances in that particular field are extremly slow nowadays =/

 

[Lysandros]

Yeah, platform inflation needs feeding; quick, throw more wood to the fire.

 

[Brucey]

100+ million gamers having their entire library on Playstation. There is a large portion of that audience that enjoys handhelds. Pretty simple.

The PS portal has sold 700,000 and until last week it only allows you to stream games from your PS5. Now, you can stream from Xcloud and it is currently outselling the PS5 hardware. And you don't think millions will buy a native Playstation device? Get out of here

Stream from ps plus premium. XCloud is Microsofts thing and performance is garbage.

 

[BlackTron]

PS5 remote play already needs lower resolution profiles? Isn't that enough to allow a "native" (upscaled) lower resolution device?
the cpu side would need to match ps5 power, the gpu though could scale?

My assumption would have been that the stream is being shrunk to accommodate being beamed over wifi and original hardware power is still necessary to play that game, but I don't know.

 

[Brucey]

Just wait till 2025, switch 2 is ps4pr0 docked and ps4 base in handheld mode, problem with handheld archi to be more powerful is powerdraw/batery life, and unfortunately advances in that particular field are extremly slow nowadays =/

I would be stunned if the switch 2 approached the ps4 let alone the pro.

 

[coachmcguirk91]

Stream from ps plus premium. XCloud is Microsofts thing and performance is garbage.

Haha, yeah that's right. I forgot people are paying 18$ a month to be able to use the portal

 

[M1chl]

Thing is, I don't think its a problem with shrinking the PC down, because, you can integrate a lot of shit together. But the heat and power usage of current components is the problem. To get comparable power, you would probably had to go with totally different architecture and vendors, so in the end it wouldn't be a PS5

 

[Ceadeus]

Both Vita and the PSP were small and truly portable.

If we're looking for current gen graphics on a handheld device, it would probably needs space for components and be as big as PC gaming handhelds.

I'd prefer the thing to look like the ayaneo pocket S

AYANEO Pocket S - AYANEO

Founded in 2020, AYANEO is a revolutionary brand for portable gaming consoles. Our team consists of passionate gamers and tech-savvy experts who want to provide gamers with the most advanced and up-to-date gaming experiences. We’ve developed a line of powerful and portable gaming handhelds, with...

www.ayaneo.com

 

[Astray]

The most important thing to me is going to be hall effect analog sticks. I do not want to have to deal with stick drift on a portable.

 

[Thick Thighs Save Lives]

Both Vita and the PSP were small and truly portable.

If we're looking for current gen graphics on a handheld device, it would probably needs space for components and be as big as PC gaming handhelds.

I'd prefer the thing to look like the ayaneo pocket S

AYANEO Pocket S - AYANEO

Founded in 2020, AYANEO is a revolutionary brand for portable gaming consoles. Our team consists of passionate gamers and tech-savvy experts who want to provide gamers with the most advanced and up-to-date gaming experiences. We’ve developed a line of powerful and portable gaming handhelds, with...

www.ayaneo.com

Moreover the display itself needs to be at least 7", because the device will be designed to play PS4 and PS5 games which were not made for small screens like the PSP, PS Vita or even the Switch Lite (the UI and text can be very small).

It annoys me a bit when some devs who port their game to the Switch don't account for this and the text and/or UI is very hard to read.

 

[asdasdasdbb]

Thing is, I don't think its a problem with shrinking the PC down, because, you can integrate a lot of shit together. But the heat and power usage of current components is the problem. To get comparable power, you would probably had to go with totally different architecture and vendors, so in the end it wouldn't be a PS5

If they doubled the CUs to 72 and downclocked to 1.5 Ghz, that would actually be above the PS5's compute power and would save a decent chunk of power. Fabbing at TSMC N2P would save a ton of power and using LPDDR instead of GDDR would presumably save power too.

None of this is cheap though. Dunno, it might be more than $599.

 

[M1chl]

If they doubled the CUs to 72 and downclocked to 1.5 Ghz, that would actually be above the PS5's compute power and would save a decent chunk of power. Fabbing at TSMC N2P would save a ton of power and using LPDDR instead of GDDR would presumably save power too.

None of this is cheap though. Dunno, it might be $599 or more.

I also think you could cut a ton of power by using some of those M1-style ARMs, with that comes an issue with GPU, but nVidia is on the report preparing such SoC to the market, something which will be in Switch 2, that way I think they could bring such device to the market. However even then it would probably be expensive as fuck and had like an hour of battery time

 

[existensmaximum]

What is the storage? Wouldn't they need to release a portable version of every game so they're not hundreds of gigabytes and filled with unnecessarily large textures etc?

 

[marquimvfs]

Feasible yes. The price though will be an issue. Probally released alongside the PS6 in 2027.

It'd have to have more CUs than the PS5 Pro and downclocked very far. And on a very expensive node.

Also, with a Relatively expensive screen and not cheap battery.

 

[asdasdasdbb]

What is the storage? Wouldn't they need to release a portable version of every game so they're not hundreds of gigabytes and filled with unnecessarily large textures etc?

Steam Deck has a 1 TB option now. That shouldn't be a problem in 2027.

 

[asdasdasdbb]

I also think you could cut a ton of power by using some of those M1-style ARMs, with that comes an issue with GPU, but nVidia is on the report preparing such SoC to the market, something which will be in Switch 2, that way I think they could bring such device to the market. However even then it would probably be expensive as fuck and had like an hour of battery time

The whole point would be so that developers shouldn't have to do much to support the handheld.

 

[Shifty1897]

The battery size needed alone kills this for now, check back in 6 years.

 

[HogIsland]

Steam Deck has made this idea completely pointless.

Plus it's not feasible.

 

[Games of Today]

To clarify the topic: Would it be possible for Sony to create a handheld, that with minimal/low-effort porting could be adapted to play most PS5 games?

Clearly, this was something Sony already wanted to do with PSP/Vita (align with PS2/3), but it was too difficult because of different hardware architectures.
However, with hardware architectures having converged in recent years, it is now much more common to make handhelds that run software originally intended for stationary hardware.
The Switch, Steam Deck, ROG Ally and Lenovo Go are testament to this.
But making a handheld that runs most PS5 games sounds like a very daunting task, even for Sony.
I asked GPT4o to sketch what kind of hardware would be required to make a handheld PS5 with a release fall 2026 feasible.

• Custom SoC: ARM-based CPU + RDNA 3/4 GPU (~3-4 TFLOPS) with PS5-like features (ray tracing, VRS, geometry engines).
• Unified Memory: 8-12 GB LPDDR5 for shared GPU/CPU access.
• Storage: NVMe SSD with PS5-like decompression for efficient asset streaming.
• API Compatibility: Match PS5's GNM/GNMX APIs for minimal code changes.
• Tempest Audio: Lightweight version for consistent audio.
• Display: 720p native with scalable output to 1080p.
• Input Parity: Built-in DualSense features for seamless control adaptation.
• Devkit Tools: Asset scaling and performance profiling for manual optimization.

Yes, get a laptop bag, a 14 inch flat screen 1080p monitor and a PS5 with the external plates removed. There you go. Carry it around to your heart's content.

 

[Fafalada]

100+ million gamers having their entire library on Playstation. There is a large portion of that audience that enjoys handhelds. Pretty simple.

Yes and that audience already has a PS console (or two).
Why would they buy a 500+$ handheld that doesn't even run all of the same library?

The PS portal has sold 700,000 and until last week it only allows you to stream games from your PS5.

Yea - and that's 200$ and still selling less than PSVR did at a higher price point with the same restrictions (needing a console).

Now, you can stream from Xcloud and it is currently outselling the PS5 hardware. And you don't think millions will buy a native Playstation device? Get out of here

I mean - Vita sold 'millions'... and everything we're discussing here would only be a subset of that market (higher price, larger device, possibly worse battery and only relying on BC/ports for the library).

 

[Ceadeus]

Moreover the display itself needs to be at least 7", because the device will be designed to play PS4 and PS5 games which were not made for small screens like the PSP, PS Vita or even the Switch Lite (the UI and text can be very small).

It annoys me a bit when some devs who port their game to the Switch don't account for this and the text and/or UI is very hard to read.

Right now the UI is often my main gripe using steam Deck. For most titles it's just fine but it gets complicated for games like ESO when it's UI heavy with lot of reading.

I could imagine it looking like a chunkier PS Portal.. do you agree ?

 

[balls of snow]

It should be dual function. All your games in the cloud or select games running on hardware.

 

[fersnake]

wait another 10 yrs to get there.

 

[Thick Thighs Save Lives]

Right now the UI is often my main gripe using steam Deck. For most titles it's just fine but it gets complicated for games like ESO when it's UI heavy with lot of reading.

I could imagine it looking like a chunkier PS Portal.. do you agree ?

Yep, I personally agree with all that you've said, and I think they'll go with something the size of the PS Portal if possible (maybe a tad bigger/chunkier), and with a different design to fit all the needed components + adequate cooling.

 

[Fafalada]

If they doubled the CUs to 72 and downclocked to 1.5 Ghz, that would actually be above the PS5's compute power and would save a decent chunk of power.

Problem is it'll still run on LPDDR and thus be very bandwidth limited. If compute alone was enough - current crop of handhelds would outperform Series S already - and at best they trade blows (mostly on account of much faster CPUs).

 

[Radical_3d]

I asked GPT4o to sketch what kind of hardware would be required to make a handheld PS5 with a release fall 2026 feasible.

• Custom SoC: ARM-based CPU + RDNA 3/4 GPU (~3-4 TFLOPS) with PS5-like features (ray tracing, VRS, geometry engines).
• Unified Memory: 8-12 GB LPDDR5 for shared GPU/CPU access.
• Storage: NVMe SSD with PS5-like decompression for efficient asset streaming.
• API Compatibility: Match PS5's GNM/GNMX APIs for minimal code changes.
• Tempest Audio: Lightweight version for consistent audio.
• Display: 720p native with scalable output to 1080p.
• Input Parity: Built-in DualSense features for seamless control adaptation.
• Devkit Tools: Asset scaling and performance profiling for manual optimization.

Why are people asking things to ChatGPT and expecting nothing but gibmess? Is a young people thing?

Anyways those specs don’t even start to make any PS5 game port “easy”. And even easy is not market feasible: it must be automatic. And you need more than that. I don’t think we’d see something capable of handling PS5 games in the handheld space in the next 4 years outside overpriced high end phones.

 

[Trilobit]

I would be stunned if the switch 2 approached the ps4 let alone the pro.

I hope it does. I played the intro part of Assassin's Creed Origins yesterday on my PS5 and was confused about whether I was playing an upgraded PS5 version. It looked very good and the framerate felt stable. I still don't understand how it could look and run that well even if it was on newer hardware.

 

[Saber]

Why would you want a PS5 portable when you can have a PS4 portable? Alot of games would be included in the list, the most notable ones are games that later released a PS4 version as well. You could let the games run at 30fps, reduce some resolutions here and there and no one would complain.
I'm betting the Swtich 2 is going to take that route as well, so every game on PS4 catalog can be just ported there.

 

[asdasdasdbb]

Problem is it'll still run on LPDDR and thus be very bandwidth limited. If compute alone was enough - current crop of handhelds would outperform Series S already - and at best they trade blows (mostly on account of much faster CPUs).

They only have 128-bit memory. Would need 256-bit plus faster memory speed to get to where the PS5 is.

 

[T4keD0wN]

It uses ~200W, so most likely 15 years or more since its release till itll be possible to achieve the same raw performance with ~15-25W which is appropriate for handhelds unless there are some massive IPC gains and breakthroughs in cooling/batteries and it actually seems that the advances on the hw side are slowing down.

This obviously doesnt account for things like upscaling and frame gen, which are a straight up cheat answer, just like lower display res and framerate targets would be since similar results could and should be achieved with less hw power, but thats not the same thing.

 

[Loomy]

It is both possible and viable. PS Portal is proof that Sony can do it, and Steam Deck and all the other handhelds are showing that there's a market.

Whether or not Sony succeeds is a whole other story. I don't think this situation is comparable with Vita/PSP.

 

[Brucey]

Haha, yeah that's right. I forgot people are paying 18$ a month to be able to use the portal

Well there's no charge if you have a PS5. But if you don't want a ps5 then plus premium is the "catch"

 

[ZehDon]

PlayStation would likely just create an x86 handheld, similar to the Rog Ally or Steam Deck, running a custom front-end. Porting would just be running their existing PC ports, so they could all be used with minimal if any changes. Because it's Sony, add USD$150 to whatever is a profitable price to bump up the margin and boom: PlayStation Portable.

 

[Valonquar]

If I am playing PS5, I am in pajamas on my couch.

 

[jataple]

the mistake: creating a PS Portable where developers have to create games for the portable console and then games for the home console, big mistake

the good way to go: creating a PS Portable capable of playing existing PS4, PS5, and big maybe PS6 games of your library, devs just create games for the playstation ecosystem that can be played in both systems

and the price will be interesting for sure

 

[Self High Five]

the mistake: creating a PS Portable where developers have to create games for the portable console and then games for the home console, big mistake

the good way to go: creating a PS Portable capable of playing existing PS4, PS5, and big maybe PS6 games of your library, devs just create games for the playstation ecosystem that can be played in both systems

and the price will be interesting for sure

Likely a 1080p PS5 portable with the option to play PS6 through the cloud