But the vast majority aren’t expecting maximum fidelity. There are millions more casual gamers that get a console for fifa cod madden fortnight.
There are millions that aren’t buying consoles for starfield etc
Give the majority a device that can play fifa cod on the go natively to series S standard and they’ll fly off the shelves.
Time frame means a lot here. Are we talking Series S Standard in the next couple of years, or in 2028 when next gen (presumably) begins?
If Microsoft made a Windows version of Steam Deck, they’ll sell about as well as steam Deck, around 5 million units LTD optimistically. Steam Deck, compared to Windows handhelds, have fixed UX, better price points, better battery life, but still lacks in compatibility, support of new games, and out of the box game experience (have to mess with settings)
If you look at the worse selling HW from Nintendo, Microsoft, and Sony, that was the Wii U, at 14 million units. Even the Vita was 16 million.
I think it’s a no brainer why MS should make a dedicated gaming device vs a general purpose gaming device.
Sell the dedicated device at a loss.
Lock down the OS and store so you get the 12-30% of everything sold in your store.
Games just work without much tinkering.
Your BC library just work.
The issue with Vita and Wii U was both lack of a large gaming library (no BC) and incompatibility with new software from first pary and third party developers. An Xbox handheld will not have the same problem, at least not nearly on the same level.
A Series S level handheld has compability with every piece of software in the Xbox library.
Even a Xbox One level handheld has like 99% compability, and 100% compatibility with the major titles, (Fortenite, COD, Fifa). Wii U, Switch, Vita doesn’t come close in terms of software support.
For comparison sake on price and performance, Apple IPhone 15 pro max at around $1200 with the world’s most powerful mobile SoC by a wide margin, just had Resident Evil 4 launch with close partnership with Capcom. The game performed and ran worse than the X1 console from 2013. The Series S portable dream is just that, it won’t happen in the next 7-10 years. With that, and keeping a competitive price to performance handheld the xbox portable would have to be it’s own platform. I would say MS should make a deal with Nvidia, and create a competitor to the Nintendo Switch 2. Platform support is the most important problem to solve.
I feel like there are a lot of base assumptions going on with this topic that make it pretty challenging to have a reasonable discussion on this lol.
Umm. I don’t think you should draw conclusions using Apple devices lol.
Aren’t we all just drawing conclusions and assumptions for this topic, I thought that was mutual. 
Aren’t we using the Apple M2 as an comparison to the Zen 2 as a basis for why ARM is good enough?
Architecturally yes, but not to indicate how expensive an ARM SOC will be.
A SOC using arm will cost less than one using x64.
I’m not sure it will provided the following:
Microsoft needs a GPU on the same feature level as DX12 Ultimate or higher, this is the AMD RDNA2 / Zen. They wont be able to use any existing ARM-based GPU cores like the existing tablets and phones offer.
Intel isnt there yet from a BC perspective on GPU, while they ask for substantially higher costs / margins for their SOCs. Nvidia has the GPU aspects but also ask for higher costs / margins for their SOCs at the same feature level relative to AMD. That means using AMD for the GPU portion.
On the CPU side, AMD has done some interesting optimizations for x64 using the Zen 4C compared to the Zen 4 CPU. What do they have planned for 5 and 5c or even 6 and 6c CPU variants.
Would an AMD ARM CPU core be substantially cheaper than an AMD x64 core 5c or 6c? I have my doubts as to any realized cost savings to Microsoft.
Arm CPU clusters will take up less die space than comparable x64 CPUs on the same node, resulting in cost savings.
Not considering r&d and royalty cost.
Interestingly, DF just uploaded a video comparing UE5 games on Steam Deck vs Xbox Series S. Maybe that Xbox next 2026 rumour could be a handheld and the ng consoles 2028.
Those high end APUs are capable of series S performance but are limited by memory bandwidth which is still DDR5 for GPU.
The topic is to discuss specs of an handheld. But it has turned into a discussion to convince some people that the handheld is possible.
Actually, it’s possible to get Series S bandwidth today using 7500mhz LPDDR5x on 256bit bus. The BW in such a setup is 240GB/s.
The only technical limitations from a Series S portable today is purely battery.
AMD for some reason does not allow 256bit bus on the APUs. Limited to 128bit bus, That’s the main issue.
Battery wise, handheld will never be possible unless it is Nintendo level performance.
All my assumptions are based on 1 HR for heavy games like Hellblade. And 2+ HR for optimised and lite games
What AMD decides for their own APU isn’t a problem for MS.
Nothing preventing MS from ordering a design for an APU that has 256bit bus. Their console APU has a 320 bit bus lol. Obviously buses results in a larger chip means additional cost.
Nintendo Switch battery is 16 watt hour. Handhelds can go up to 60 watt hour lol. The Deck OLED has 50 watt hour.
Xbox portable imo should have 20 watt total system power draw and 50 watt hour battery. That gives up 12 watts for the APU which will run circles around Switch 2.
No problem for MS if they are making xbox handheld. But bandwidth is currently the limitation of existing PC handhelds APUs. That is what I was implying. That is why xbox handheld is important which unfortunately many doesn’t understand.
PC handheld players are not putting in much effort in R&D. Xbox will do that and bring the best handheld.
12 watts for the APU is nice for BC and lite games.
For games like Hellblade, standard 28 watts for the APU will be required.
So I suggest a 40 watts total system power draw. It will give roughly 1 hr of gameplay for heavy games and 2.5 hrs for BC and lite games.
Would people be open to Xbox doing a handheld streaming device akin to the Logitech or a PS Portal (but without requiring connection to your own console)? A lot of games require online connections anyway, and that type of device would be much cheaper to create. It also would be able to keep up with newly released games longer because it wouldn’t have to do the processing locally
What would differentiate such a device from a Logitech G? How well did the Logitech G sell?
The current handhelds on the market that are the closest in experience to a console is Switch and Steam Deck, both of which has sales many times better than the Logitech G.