Microsoft Is Working on Two New Xbox Consoles, One of Which Is Streaming Only

Megalith

24-bit/48kHz
Staff member
Joined
Aug 20, 2006
Messages
13,000
Thurrott has learned new details regarding Scarlett, Microsoft’s upcoming family of Xbox devices. Redmond is reportedly working on two consoles: one will be a traditional piece of hardware, a la Xbox One, but the other will be a weaker but less expensive cloud console that can only stream games.

The cloud console will have a limited amount of compute locally for specific tasks like controller input, image processing, and importantly, collision detection. The downside of this is that it since more hardware is needed locally, it will raise the price of the streaming box but it will still cost significantly less than what we are accustomed to paying for a new-generation console which should help expand the platform’s reach.
 
I can see the streaming box selling very well if it a good deal cheaper than the normal Xbox.
 
I'd imagine the costs of the streaming service would add up quickly to the costs of a full blown console anyway.

I don't think gamers outside of baby boomers addicted to facebook games are inclined to such a service just yet. It would also be impossible for "competitive" esports style games due to latency.
 
not so sure splitting resources like this is a good idea. MS looked like they were getting the xbox brand back on track to being the company they were in the early days of 360. PS4 didn't have a hard time selling compared to xbox, because sony focused on what a game console is suppose to do..
 
I think it is a good idea provided that they can figure out the latency issues. That has always been the problem with streaming games from the cloud. There is usually a degradation in image quality too, but most people don't care about that provided the game is running smoothly.
 
This streaming shenanigans has been heralded for years as the future and never *actually* takes off. All the big issues remain, they show zero signs of being remedied, probably wont for decades. Streaming as a future for gaming is an idea that ignores past experiences and glaring elephant-in-the-room issues. It's a total and utter waste of time and money for microsoft as big as the xbox one tee vee tee vee tee vee pitch and will go just as wrong again. Simply another blind alley to wander down that distracts from the thing that works: gaming focused machines for gaming.

Xbox one looked meek and inferior when it was revealed and they even went full arrogant and out of touch with always on and no trading used games. They're only a few monkey brained decisions from another console wars drubbing.
 
This streaming shenanigans has been heralded for years as the future and never *actually* takes off. All the big issues remain, they show zero signs of being remedied, probably wont for decades. Streaming as a future for gaming is an idea that ignores past experiences and glaring elephant-in-the-room issues. It's a total and utter waste of time and money for microsoft as big as the xbox one tee vee tee vee tee vee pitch and will go just as wrong again. Simply another blind alley to wander down that distracts from the thing that works: gaming focused machines for gaming

Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).
 
I see issues. From what I've read, the streaming console will process some data loads locally while others are sent to the server due to latency issues. While this makes sense, it also puts a lot of limits on how developers can create their games. Essentially, the weaker streaming console's limitations on processing power will be the standard for the generation, putting the new Xbox at a major processing disadvantage.

Simply put: Stupid idea.
 
I dunno, I'm curious to see how the streaming console handles latency. If this 'splicing' idea works and mostly eliminates latency - I could be on board for one. My PC is my 'main' rig, so if I could purchase a cheap 'stream' console to play the few exclusives for the machine to complement my PC - great.
 
Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).
4K streaming at the bandwidth needed for 1080p at a minimum.
 
Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).

Uhmmm,..nope.

1) Latency is a matter of physics. Unless you want to break a couple of those laws, latency is always going to be a problem.
2) Netflix is NOT a real time service (latency wise), like a game is.
 
Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).

Except Netflix 4K is pretty shit quality compared to UHD-BD and is not so latency dependent as games would. Resolution does not mean quality, others factors are important too. Netflix and streaming popularity gained due to ease of use and convenience and people sacrificed quality. I suppose cost too, UHD BD player and discs are quite expensive in comparison. I personally find Netflix and other streaming quality way too bad so I don't even want to use it. I still bother with discs and rip them for convenience of use at home.
 
Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).

Yeah, the latency streaming from my PC to the living room using Nvidia Shield or Steam Link is too much for anything other than slower RPG's for me and it's not like I'm some FPS guy doing everything fast twitch to begin with. I don't see how putting the rendering hardware 75+ miles minimum away rather than feet is going to be solvable.
 
The only way this makes sense is if the non-streaming Xbox is pirced at $1k+.... which at current GPU pricing trends.. will see what the GTX 1160 launches at.
 
With streaming games you'll probably end up using more bandwidth than if you just downloaded the game instead.
 
Im not a console gamer (PC's only for me !!) but I don't like the idea of a streaming console. Sounds like another income generation scheme where users have to subscribe to a service or pay rental on the streamed games.
 
Or they will resolve latency and provide services/games people want and it will be a big hit. I would bet on it working and being profitable.
Look at Netflix. "No one will give up having actual disks for low quality video streams" (4k streaming now available).

Put it this way, is your average reasonably priced internet connection in the US capable of streaming games? If no do you see that changing in the next 5 years? There's a good chance we got no twice in a row. If streaming is going to take off it'll probably have to do it first in the usa and/or china (the mega markets). Compared to the rest of the developed world american internet is garbage. So there's one giant roadblock right there.

Then there's shifting consumer tastes. Over the last 10 years we've gone from being 'ok' with playing crysis 1 at an average 25ish fps, to having our games at 60fps... at a vastly higher res. The leading edge amongst us aim even higher. So the trend right now is toward high refresh rate gaming, both in pc and slowly in console realms to. How on earth does that trend reconcile with streaming games and all the latency padding that goes with it?

So there you are, going to the buying public thats grown a taste for the finer things in gaming and saying hey everyone, for a small monthly fee you too could play games in a syrupy fashion on your 1st rate internet you pay an absolute fortune for. No CoD though it doesn't work well enough.

Over time technology moves forward. But I'm betting that'll go into making standalone consoles output 60fps at higher and higher resolutions and ISPs in america will go on milking the consumer.
 
Last edited:
Put it this way, is your average reasonably priced internet connection in the US capable of streaming games? If no do you see that changing in the next 5 years? There's a good chance we got no twice in a row. If streaming is going to take off it'll probably have to do it first in the usa and/or china (the mega markets). Compared to the rest of the developed world american internet is garbage. So there's one giant roadblock right there.

Then there's shifting consumer tastes. Over the last 10 years we've gone from being 'ok' with playing crysis 1 at an average 25ish fps, to having our games at 60fps... at a vastly higher res. The leading edge amongst us aim even higher. So the trend right now is toward high refresh rate gaming, both in pc and slowly in console realms to. How on earth does that trend reconcile with streaming games and all the latency padding that goes with it?

So there you are, going to the buying public thats grown a taste for the finer things in gaming and saying hey everyone, for a small monthly fee you too could play games in a syrupy fashion on your 1st rate internet you pay an absolute fortune for. No CoD though it doesn't work well enough.

Over time technology moves forward. But I'm betting that'll go into making standalone consoles output 60fps at higher and higher resolutions and ISPs in america will go on milking the consumer.

Streaming Game service does not have to be great to be profitable. It has to be good enough. Also, the next generation of consoles will likely have a 5-7 year life span (and they have not even been released yet). This would push boxes that primarily stream content into 2025.

I also think most people are thinking about "streaming" games in the same way you might stream video. My bet would be the technology would not be the same. We won't be rendering 100% of the game on servers and streaming a video. I imagine a box that streams large portions of highly rendered pieces of background and interactive objects would be rendered in large part locally. Imagine the perfomrance of a Street Fighter Game if all you had to render was the character models. Backgrounds, particle effects, etc would all be "streamed". It is complicatred, but why can't they resolve it?

Also people, just because it is better doesn't mean it is what will win commercially. Firefly was cancelled before the first season finished airing. The Bachelorette is in Season 14.
 
If you put yourselves in their shoes, there are many millions of gamers out there with nothing more than an iGPU or GTX 1030 class card(at this point).

There are also many many popular games that are not network latency dependent. Pick almost any of Nintendo's games, they are turn based. The awesome Tale of games...



But logic fails at price... even if it is only $100-$200 dollar device, even at $10/mo subscription, you could have a full size console in a little over a year(more, depending on sales).
 
not so sure splitting resources like this is a good idea. MS looked like they were getting the xbox brand back on track to being the company they were in the early days of 360. PS4 didn't have a hard time selling compared to xbox, because sony focused on what a game console is suppose to do..

Yeah, this is MS's problem. They go half-in on something acting like it's 100% the future, but they also half ass it. When it flops, MS writes it off and falls back on their old businesses while everyone else (the consumers who bought in, the companies/divisions part of it) are left holding a steaming bag of poo.
 
Streaming Game service does not have to be great to be profitable. It has to be good enough. Also, the next generation of consoles will likely have a 5-7 year life span (and they have not even been released yet). This would push boxes that primarily stream content into 2025.

I also think most people are thinking about "streaming" games in the same way you might stream video. My bet would be the technology would not be the same. We won't be rendering 100% of the game on servers and streaming a video. I imagine a box that streams large portions of highly rendered pieces of background and interactive objects would be rendered in large part locally. Imagine the perfomrance of a Street Fighter Game if all you had to render was the character models. Backgrounds, particle effects, etc would all be "streamed". It is complicatred, but why can't they resolve it?

Also people, just because it is better doesn't mean it is what will win commercially. Firefly was cancelled before the first season finished airing. The Bachelorette is in Season 14.

Because your options are to wait for all rendered data to be compiled before pushing it to the screen, or you have a ton of pop-in from either the streamed or locally rendered stuff.

The reality of the matter is developers aren't gonna do all the work required to mitigate pop in and balance the two separate loads properly so they will push it all to the rendering farm and you are gonna end up with an extremely annoying delay between the controller input and the movement on screen.
 
Because your options are to wait for all rendered data to be compiled before pushing it to the screen, or you have a ton of pop-in from either the streamed or locally rendered stuff.

The reality of the matter is developers aren't gonna do all the work required to mitigate pop in and balance the two separate loads properly so they will push it all to the rendering farm and you are gonna end up with an extremely annoying delay between the controller input and the movement on screen.

You can still do something like Occulus's asynchronous timewarp/spacewarp locally to smooth out input without a ton of tweaking from the developers.
 
Yeah, this is MS's problem. They go half-in on something acting like it's 100% the future, but they also half ass it. When it flops, MS writes it off and falls back on their old businesses while everyone else (the consumers who bought in, the companies/divisions part of it) are left holding a steaming bag of poo.
Yeah, with a huge company like MS, very likely, a "new" group is created for an exciting project. They use other projects money, as this is forward looking, and everyone gets free dinners and overtime and new equipment to work on the new project... If they fail, oh it was ahead of its time, it wasnt their money to big with.
 
This will be so confusing for parents that don't know jack to figure out which one to buy for their kids. They will just say F it and buy a PS5.
 
I can see the streaming box selling very well if it a good deal cheaper than the normal Xbox.

Really, beacuse I see a future where the streaming box is advertised with visuals that do not represent what you get out of the streaming box, AND a shitty customer experience in certain areas where the lower price has lots of appeal, combined with shitty return policies and whole host of other things that will make everyone involved regret this.

But maybe I'm beign pessimistic and there will be no latency issues, no bandwidth issues, perfect visual quality, and no geographic bandwidth deserts.

Or maybe they will jsut release samples to reviewers and they will be panned as so much shit they just won't launch it.
 
If they wish to install their game servers in ISP datacenters they may have a chance to conquer the latency issues.

Gotta get that content as close to the customer as possible
 
Because your options are to wait for all rendered data to be compiled before pushing it to the screen, or you have a ton of pop-in from either the streamed or locally rendered stuff.

The reality of the matter is developers aren't gonna do all the work required to mitigate pop in and balance the two separate loads properly so they will push it all to the rendering farm and you are gonna end up with an extremely annoying delay between the controller input and the movement on screen.

You are miss-estimating the balance of complexity vs the technology in 10 years. In 10 years such rendering will be as simple to code as xfire or sli
 
Really, beacuse I see a future where the streaming box is advertised with visuals that do not represent what you get out of the streaming box, AND a shitty customer experience in certain areas where the lower price has lots of appeal, combined with shitty return policies and whole host of other things that will make everyone involved regret this.

But maybe I'm beign pessimistic and there will be no latency issues, no bandwidth issues, perfect visual quality, and no geographic bandwidth deserts.

Or maybe they will jsut release samples to reviewers and they will be panned as so much shit they just won't launch it.

Or maybe it will be ok.
 
I’d just like to point out that people play flash games and think they are great.

Saying that the future is going to be the 6k 200fps heaven.... being right is not the same as being correct.
 
It seems MS is bringing the Phantom console back to life, in a sense.

OTOH, if their "streaming only" console is like... $79 with a wireless controller? It will sell like hotcakes whether it works well or not.
 
Great, I can't fucking wait for a generation of games I can never play again because their streaming servers have to make room for the new stuff.
 
Yeah, the latency streaming from my PC to the living room using Nvidia Shield or Steam Link is too much for anything other than slower RPG's for me and it's not like I'm some FPS guy doing everything fast twitch to begin with. I don't see how putting the rendering hardware 75+ miles minimum away rather than feet is going to be solvable.

I'd love to say gamers care, but I just don't really think that's the case. Modern first person shooters seem to have tons of latency already baked into the game, and no one cares. And it seems to be getting worse every year. If people already don't care about 100 ms delay in their game, they won't notice streaming enough to really whine about it.

pHCiT3m.png
 
If you put yourselves in their shoes, there are many millions of gamers out there with nothing more than an iGPU or GTX 1030 class card(at this point).

There are also many many popular games that are not network latency dependent. Pick almost any of Nintendo's games, they are turn based. The awesome Tale of games...



But logic fails at price... even if it is only $100-$200 dollar device, even at $10/mo subscription, you could have a full size console in a little over a year(more, depending on sales).

I think the price thing is ok - I mean if its $150, and then you pay $10 per month for 2 year you are at about the cost of most consoles when the launch. But you have gotten 2 years worth of gaming by then too. Its not like you just buy the hardware and you are good to go on a normal console launch. Especially at launch the games all stay at that $60 range for a bit with a bad or at least limited selection to start (at least for playstation with typically no backward compatibility).

So at that level it would be acceptable for me.

I agree with everyone's concerns on the latency thing, I don't see how its possible to totally overcome it. But I also think it only has to get "close enough" for MOST people. But not for MOST people on THIS SITE.

Also, I want to play devil's advocate on one aspect - Multiplayer games. Is it not at least conceivable that those could play better - or at least more "fairly".

Think about this:

Normal MP game is:
Controller ->console rendering and displaying locally - > sending the inputs to a server - > results of inputs from all consoles processed on the servers -> sent back to local console. This is where you see all the desync and lag and "Oh Bullsh!t" I hit that guy first.
Couldn't it be just as effective if it is:

Controller > Console > direct send to server > server render (including all inputs) > send results back with display?

Its basically the same thing. What you see might be more... mushy, but at least it would be accurate to what everyone else is seeing too.

I am sure that is not how it would really work, but I kind of see it as not being significantly different for MP. Mostly with MP, the action might be fast, but you don't always actually see what the server sees. So if you are satisfied with MP and the way your actions translate into kills on screen, not sure it would necessarily be much different with a streaming gaming service.

Local single player twitch would probably suck though as everyone points out.
 
Back
Top