Not sure it’s viable to introduce another hardware that only runs last gen games natively and would require specific porting for current gen games, specially after launch of the generation where you could potentially end up with some games natively and other through cloud or not running at all.
I think it would be more sensible to have a very thin client streaming only device now, and in a few years time an apu matching or exceeding SS performance on mobile.
Anything streaming wouldn’t catch on because you have phones for that. The reason why steamdeck is exciting is it’s hardware to play games not stream them. If SD was streaming only most people wouldn’t care
Other than being cool from a tech perspective I don’t really see the appeal for these devices. Is gaming on the go that common? I never see people gaming when I’m out and about and if you’re at home or at friend’s place why wouldn’t you use a home console or PC which is going to provide a much better experience. Is it for laying in bed? Maybe I’m just not in an area where on the go gaming is common, but the cool sounding scenarios show in the early Switch videos I’ve yet to see once in the multiple years of its existence.
Further update. Dream might not be dead. VRR functionality is added by the display’s chip and drivers, not the screen itself. For a handheld MS can order a custom display chip with custom display driver.
The only hurdle remaining is FHD screens in the 7-9inch range. From my research, there are only 1920x1200 displays. Perhaps that’s workable if players don’t mind some black bars on top and bottom.
I watched the video you shared and it’s maybe all about the display connector which is different in gaming monitors and mobile displays. So yeah, VRR is still possible.
But as he mentioned in the video that it would be a tad expensive. But I don’t care. I just want premium experience and may pay for it as well.
And second note… Do you want a 720p display or the exact 1920*1080p display?
I had thought VRR required the screen to be different.
Another small update is removing the digitizer from the display. It’s probably going to take a big effort to make an OS that’s gesture compatible. Since no games on Xbox is also touch compatible, might as well remove touch functionality and save on cost and complexity.
Lack of touch should be marketed as a feature, to cement the fact that this is purely a portable Xbox.
Idk, to be able to play ori, resident evil, perfect dark, starfield on a handheld? That sounds amazing. How about with the one X tech but a series S chip? I’m not a hardcore techie so if someone can break that down for me, thanks.
What’s your thought on Stream only (local and cloud) Xbox handheld?
From our conversation up until now… A good Xbox handheld seems to be 4 years away atleast.
Till then we may have a stream only handheld.
This device could be as low as 100 to 150 and could fit well for portability “engagement”.
This could also lay some ground work for full handheld console and test the waters ( do Xbox user want to play Xbox games on a handheld )
The key technology for this device will be streaming medium.
For xCloud 5Ghz dual band is a norm already.
But for local console streaming it can upgrade to 60mm giga wave. It has much less latency ( practically unnoticeable ) and much better bandwidth ( 4gbps ) for uncompressed 1080p 60 FPS stream
Local like you said will be a great experience if you want it to be.
As for non-local, a stream only handheld needs to be 5g capable.
I think the additional latency inherent to streaming a game is completely depend on the video encoding, decoding, and transport. Ultra low latency in the media streaming world is 500ms! You can ameliorate it by having lots of edge compute like Google does and get it down to 150ms, but that’s still additional 75ms on top of all the existing latencies you’re dealing with.
work very closely alongside Qualcomm, AMD, Intel and Nvidia, and
their last two chips have been SQ1 and the Scarlett/Lockhart APUs –
… here is what I consider as the real design challenge:
Use the Xbox One S specs to build an ARM emulation version of this. (This will also allow MS+Qualcomm to compete with the Apple M1). This will also allow for low wattage.
Use a custom ML section (like Tensor cores in the Google phone) to have deep learning modes. This will also allow for Xbox’s custom DLSS chips. These can be a part of the APU or can be separate in a docking configuration provided that bandwidth considerations between the main unit and the docking unit are accounted for
The haptics+gyro can be directed to a custom chip separate from the main one.
Target resolution could still be 720p upto 900p on the handheld mode, and upto 4k with DLSS.
Targeting 60fps is going to be the real USP with an x86 emulation architecture working on ARM.
Some amount of surplus compute-power is required to run fpsboost or BC titles with their own custom emulator/translators.
Basing it off the AMD’s chip is not going to help. The power draw is still going to be high.
The advantages being that they can design for an Xbox One, and have a single dev platform to code for.
Architecturally, they should be able to work with these 3 SKUs in the future too. The Xbox One based handheld migrates to the Xbox Series based handheld in 5-7 years when the S2 and X2 come out.