Specs for a hypothetical Xbox portable

So I did a bunch of research and landed on the only sane setup for a Xbox handheld that can be made by 2022/2023. Series S is will take another 3-4 years of mobile tech evolution to be fit into a sub 30W handheld.

The key motivation behind the device is hardware BC with Xbox One and dual use with a Surface device. Ironically MS’s misstep with Xbox Ones memory setup allows them to make a Xbox One portable with tech available today. Whereas there is no financially sensible way currently for Sony to get 176GB/s in a portable needed for PS4 BC.

With that being said here is the specs for a Xbox portable with Xbox One S BC.

Octocore Zen at 1.8+ghz. SMT disabled for XB1 BC.

6 RDNA2+ WGP(7 total) for 12 CUs at 1.3+ghz. ~2 teraflops. 16-32 ROPs.

32MB SLC (System level cache). Emulates Esram in BC mode. Only added if necessary for BC.

10GB of LPDDR5 6400mhz, 104GB/s. (2GB for OS).

512GB eMMC / SSD solution with lowest TDP. 1-2GB/s read.

1080p OLED. 8-9" screen. VRR.

4g/5g slot to enable Xcloud on the go.

$499

The performance overhead should allow BC games with fluctuating framerates and resolution to hit their ceiling. The HW feature parity with Series consoles allows for easy porting from S/X builds. The performance target for these ports will be 30fps, perhaps unstable 60fps, instead of 60fps, and further reduction in resolution and settings from the S preset. Built in VRR screen will greatly help the performance targets.

The main sell of the device is a built for XCloud device that can play 99% of your existing digital library natively. I think by 2022, and even 2023 that should hold true.

Further clarification for dual use.

The same APU will be used in a Surface device with SMT on, bumped specs, etc.

I.e. 8 core Zen at 3ghz+ with SMT

14CUs at 2ghz. 3.5+ teraflops

16GB 6400 LPDDR5, 104.2 GB/s.

1 terabyte NVME SSD, 2+ GB/s.

1440p display,

30-50 W TDP.

$999+

Of course the Surface device will be sold as a gaming focused surface. It will be comparable to a Series S in perf.

11 Likes

Price?

S349 for the portable. $999 for the Surface using the same APU.

1 Like

I don’t really see a point in spending that much money on hardware that will never be used in new games. Wouldn’t it be better to keep the device streaming focus and put the money into places that support that? (high quality screen, audio, wifi, etc.)

We should talk about what’s stopping from all this happening

Battery Life.

There are already handheld devices out there which when run on there full intended TDP (30 watts) they produce results equal to Xbox one s or even better.

The only problem is the battery only lasts 1-2 hours and I am being generous here.

At 3 nm even the series S chip could come down to 30 to 50 watts. It’s just that there won’t be enough battery juice to run that APU.

The other most important thing is VRR. Give me 100% sRGB screen with VRR and I am sold.

VRR makes every console comparison useless right now. This feature is that good.

Edit:

Also, a device with series S chip and VRR… I could pay upto $500 for that machine

1 Like

This is why a dedicated xCloud portable in the future will be the best imo, the battery life can be amazing because it doesn’t need a lot of grunt. The hardware won’t get outdated either and devs wouldn’t need to do anything to get games on it. But this won’t work for a while as they need to get xCloud experience more polished.

2 Likes

The cost of such device isn’t going to be far off from the one I proposed. You still need apu ram etc.

The same way that the series s is 1/3-1/4 of the X power wise but more than half the cost.

Of course, I’ve taken battery life into consideration. Compared to the 20 watt steam deck, my APU is wider but slower. At minimum the clock speeds would be a match for base Xbox One S. I also have half the ram and the ram is slower.

Steam Deck also has advanced haptics and gyro that will use up some power.

The peak wattage of the Xbox portable proposed would be 15 watts. Playing games in BC mode should last at least 3 hours.

1 Like

Steam Deck with Windows installed is quite literally an Xbox Portable. No need to think of hypotheticals when Steam Deck is coming…

2 Likes

This solution would require a signifigantly more complex APU (pretty sure the XSS APU is around 3x the size of the typical mobile APU) and significantly more (and faster) RAM than a streaming focused device. It would also require less storage.

1 Like

Of course. But there is a baseline cost that’ll be ballpark similar due to casing, screen, batteries, packaging, assembly etc. At minimum the unit will cost $149.

Maybe I’m in the minority here but a native Xbox handheld will always be more attractive to me than a streaming device. I mainly play Warzone and COD. Those are not available for streaming, and I wouldn’t want to stream them ever. Happy the ability to also play 99% of the Gamepass library anywhere I want is cherry on top.

Yes, but Xbox native handheld plays 99+% of your existing digital library and 99%+ of the Gamepass library.

Steam Deck and Xcloud only device can’t do that. I think it’s a very powerful pitch.

I definitely see value in a native Xbox handheld, but I can’t imagine releasing something that is supposed to allow native gaming and it likely never seeing a new game past 2022.

1 Like

It’ll definitely see new games. It has feature parity with Series consoles. Just like the S will be a 1080p-1440p device, the handheld will be a 720p-1080p device.

For the rare games that demand too much, they can always resort to Xcloud to stream it.

Think of the handheld as a portable Xbox One that has an extra two / there years added to it’s lifecycle playing native games, and indefinitely playing streaming games.

I wasn’t talking power for the 15 watt version which you described actually.

I was talking about the shrinked version of series S chip. For that power wattage will remain anything from 30-50 in the best case scenario. For that battery revolution is required

We’re a generation away from a Series S portable. That needs at least the 2nd / 3rd generation of the 3nm process. That’s a topic that’ll be discussed in 3-4 years time.

2 Likes

A 15 watt handheld in 2022-23 as your suggested specs would only be a good console for current games and casuals.

6-8 CU is really going very low. The 14-18 CU version is critical with no throttling for CPU and that would require atleast 30 watt.

I know current series S could be manageable from game development perspective. But this 15 watt console could be a nightmare in my opinion.

Let’s see if anything happens at first place. Then we will discuss again.

The switch does great because a) it’s cheap b) it’s brilliant for kids c) it plays Nintendo games and is the only option for playing them d) it’s multi purpose.

I don’t see a market for an Xbox handheld outside of real enthusiasts. If you want a cheap box you can buy the Series S as it is. If you want to game on the go you can xcloud to your phone already. People simply won’t pay for a mobile console to play these games - I just don’t see it.

1 Like

14-18CU will need to be clocked at around 1-1.3ghz. Preferably on 5/6nm to fit 15 watts and under.

As for your second point, devs are not obligated to support the handheld with native ports the way they must for Series S.

MS would be very clear that this portable is mainly for playing BC games, indies, Xcloud and a fraction of next gen only games.

The fact that the steam deck will only help the Xbox handheld get native ports of next gen only games.

I guess what your definition of success is. If a Xbox handheld can sell a few million units and gain / retain a few million Gamepass subscribers as a result, the R&D would be worth it in my option. Of course caveat is that I’m not a Xbox suit.

Speaking of Xcloud, what’s the percentage of the entire Xbox digital library and Gamepass library is available for streaming? If the user wants to play all of their digital library / gamepass library on the go, there is no option. If a user wants to play Fortnite / Warzone at 900-1080p60 on the go with minimal latency there is currently no option.

Don’t forget that the silicon will also be repurposed for a lightweight gaming Surface.

There is definitive qualitative and quantitative value to a Xbox handheld with Xbox One S BC.

It’s just a matter of whether the financials make sense.

2 Likes