A few comments.
I don’t see the value of gyro when console side doesn’t have it. It’s nice for people are like gyro aiming. It does add cost to the bom.
For targeting resolution lower than Xbox One, that’s going to break BC in a lot of games.
I think Steam Deck’s perf / TDP will tell us if Xbox One and higher level of perf can be achieved on x86 and 5nm.
You can target a lower spec without breaking the game. DLSS works upwards, and subsampling techniques have already in play for many years. The only reason why you would do that is for a smaller power consumption profile.
But yeah - you can opt for an equal target resolution there too.
My point being - don’t do this on an x86 base architecture. It is not architected for a lower power consumption at all. ARM big-little works well for these cases, and install a good x86 emulation circuitry in the big-little architecture. Appe’s M1 does smart branch prediction etc. which can be adopted hopefully with some partnership with Qualcomm for instance - who have the OG Apple chip architects, now.
I feel that Steam can most definitely achieve the perf/tdp that will indicate XBO and higher performance can be achieved for that 5nm process.
I can understand why Steam chose those decisions, but they wanted to be a PC (even though this is now a proper console), Xbox need not go down that route.
Watched this video of Onexplayer and thought it’s a good benchmark to revisit the argument of Xbox Portable
FH5 is running on it pretty nicely at 1920*1080 high 30 FPS. FH5 is excellent example of motion blur implementation at 30 fps, so it makes a really good benchmark of having full HD smooth gameplay on an handheld for a next gen graphics game
Also, lowering the resolution to 1600* XXXX and 1200* XXXX ( ball park of HD 720p) can take you upto 60 fps easily in the above shown bechmark
FH5 running this good gives a ton of hope for the xbox portable
Hope Xbox makes it possible 
The tech simply isn’t there yet, or more accurately gpu manufacturer’s don’t have much interest in the mobile form factor 
COD, WOW, Diablo 4, Halo and hundreds of Gamepass games + your Xbox digital library running natively on a device 2x the power of steam deck doesn’t sound so bad in 2023. Or a series S level handheld in 2024+.
Its a matter of when not why.
I’m down for a XBoy.
With Microsoft in the process of owning Activision Blizzard, MS will soon have enough studios to make a portable system and bringing out a XBoy will also open the door to the Japanese market too.
Maybe then MS might go after Sega or Level 5 and Mistwalker, because they have experience in the mobile market, plus handheld market.
I’d want to make sure the XBoy has R2, L2, R3, L3, a built in hard drive, touch screen and gyro controllers built in too. Plus to be able to stream your other Xbox games to this system too.
Also seeing as it would be a new system having exclusives would also be nice.
If you can do all that, then countries like Japan will be easier to get, to support this system.
Such a idea would heavily rely on countries getting their act together and providing fast internet everywhere. Norway is on a good path.
FH5 isn’t pushing IO and CPU. Would FM or FH6 work, if those games push both IO and CPU? I have my doubts.
Offcourse IO and CPU needs to be taken care of… Most important part is GPU. Even Onexplayer has more potential here if it could get gddr5 memory
It’s a limited comparison for now.
MS could also work on optimizing cloud and compression to make the reqs smaller and smaller while also waiting for the internet situation to get better around the world. It won’t be a one thing fixes all, they will need to make it easier to stream but also need the internet to improve around the world.
I think having a portable device that only streams games is a much smarter idea than one that plays them natively because you would be able to have such a low end CPU/GPU which would consume such little power compared to something that needs to be native and look good.
Having a much lower power draw will mean less heat too so the device won’t be huge and clunky like the stream deck seems to be. The device also won’t be holding anything back and devs won’t have to dev games to fit on it and take time out of making Xbox/PC games they’ll just work without any changes.
Price too, it would be much cheaper to make because of the less cooling components and the weaker CPU/GPU and even ram and storage, you won’t even need storage.
40% of sampled Era is interested in a Xbox native portable. That translates to 80% in the real world.
This thing will do switch numbers!
This video here tells why series S GPU is clocked at 1.565 GHz
Video talks about Steam Deck and it’s GPU clock speed is kept at 1.6 GHz max which is surprise surprise closer to what Series S is
1.6 is basically a point of diminishing return for the RDNA2
Xbox has chosen the best point to stop and make things power efficient… This arises a question, why even think about power for a device plugged to the wall? Answer is simply the title of this thread
Also, 1 RDNA2 CU takes around 1.225 watts @ 1.612 GHz… I would assume for series S it would be 1.2 watt per CU
I,e… 1.2*20= 24 watts for series S GPU
I bet these numbers are in Xbox’s hardware team documents as well.
Picture is getting clear for the portable here 
I still think a xCloud only portable would be a better idea in terms of battery life and size of the device. Maybe they could have a two sku approach a pro Series S model and xCloud only.
I cant see them ever doing a handheld, at least one that’s capable of playing games natively
I can maybe see them creating some sort of dedicated cloud device but sadly, I think the steam deck is the closest we will get
if xbox does make a portable they should keep the screen at 720 so it will be easy for games to scale with resolutions being so low, Switch has proven that 720 in portable form is good. Keep the cpu high enough to keep 60 fps. give it 250 to 500 gigs or storage(expandable) and make it digital only. basically a portable series S. Call it the series G or xbox GO
There is already a Portable Xbox. It’s literally your phone. Simples.
That would be good as well… But point remains the same, cloud can’t beat the local in many ways
you make an handheld you have to call it the Pip-Boy right???
xcloud only portable seems like a waist since we all have phones. A lot of people don’t like streaming or live in areas where it just mite not work. I hope we get a dedicated hardware device
It would be targeting people who either dont have a controller or simply dont like touch controls
There’d be people that would just want to be able to pick it up and play like a tablet with controls like a switch
How big is that audience for a cloud only device? Probably not a massive one, at least right now, a native one would be much more popular but also a lot more expensive to make