[Rumor] Brad Sams: Xbox Scarlett Cloud box information leaked (Next-gen; Traditional system as well)
- Thread starter Iichter
- Start date
Why would you need fibre to the premises when it shouldn't use more bandwidth than Netflix?
It is when the result is you not having access to software you've paid for.
I think the price will tempt people, but I'm not sure the average persons internet (here in the UK at least) is good enough for it to snowball into negative coverage for the machine.
Maybe they can put "NEEDS 5MBPS MINIMUM" on the box, but:
A. Your average customer doesn't really know what that means.
B. It'll lead to resentment for customers who have to spend £300 more because their internet doesn't meet the requirements.
The last thing they need is further negative connotations with the brand, they've worked hard to overcome it with the X1, I can see the boxing day newspaper reports now if the infrastructure isn't there.
Netflix isn't sending packets back every second it's connected. There's more things to consider with a game.
I'm confused why they would need to split them. Couldn't the Xbox Two also just be a streaming box too? Just a price worry?
I have zero interest in game streaming though, so fuck it if it means a cheaper box, I guess.
I have zero interest in game streaming though, so fuck it if it means a cheaper box, I guess.
It has to start somewhere, this is likely the future whether people like it or not.
You already have several PS4-owners in this thread saying they'd purchase the XBox "Cloud" kit. MS has to think about that too.
Not everyone has great internet, but by the same token there are a lot of people do have pretty decent internet speeds and with 5G coming that also opens up more options.
But when you play games online they are unless you see other players warping around everywhere.
I work for a very big networking company. They are known for its awesome hardware, but most of their profit was made from services. Now their focus is on superscription models. Not everybody wants to buy hardware since a lot of people work from home, mobile, remote offices, cost of renting office space, storage between others.
Now software that was a once a time purchase doesn't make sense. Why would you buy a permanent license of an Adobe suit, if once the next one is out, yours will be lacking. Instead companies are selling the opportunity to always be up-to-date and receive free support as long as their subscription is current.
Streaming technology is already part of the ecosystem and it will just grow from now, I agree it requires investment, but MS is already investing on streaming and so are other companies. Interactive Internet connection services was a meeting I attended a couple of months ago and they talked about how many big players like Google, Apple, MS and Amazon are investing on streaming services and what potential opportunities they have. Gaming is one of them of course.
I don't think this means that the "Nextbox" will lack game streaming. This is meant to be an additional SKU that offers game streaming as the primary method of playing games. It will most likely mean a much cheaper piece of hardware that makes the brand more accessible to those who have the right internet speed.
It really depends on how all of this is expected to work, the Kahawai tech demo essentially has the local client play the game locally in low quality and the cloud service would then provide a high quality pass on top of it, this would make the game very responsive from an input lag perspective at the cost of sacrificing visual quality if the visuals on the screen changes very rapidly.
The thing is that we know very little about what their solution really is and how it works in practice, but Microsoft R&D has spent considerable resources into cloud rendering for things like games/VR/AR and pushed out some really promising demos.
The Kahawai demo:
This is something I hadn't really thought of yet. If I have to increase my internet package to make this work, say $30 more a month. That's $360 a year, or $1,800 dollars over a 5 year console generation. That's enough money to buy a $400 console and 23 $60 games. And that's without even considering the price of the streaming box or gamepass.
???
I know they are, and that's why it required more bandwidth than netflix.
Interesting, curious to see how it all pans out, personally I'm more interested in the traditional console.
And? When the original Xbox came out no one used Ethernet connections, but they still put one in there cause they knew it was the future. They also put an internal HDD in there which is another thing no one did cause it wasnt needed. Just cause the future isnt here now doesnt mean it isnt coming soon.
I really needed to explain this in 2018?
I'm thinking back to my initial impressions, and I'm really wondering what this box will mean for the "One" brand. I don't see the One S/One X sticking around for longer than the usual cross-gen period, but the streaming box opens up new avenues.
Would it need to be at least as powerful as the One S? If so, could they make it a disc-less version of the One S that's also able to stream "Nextbox" games?
Would it need to be at least as powerful as the One S? If so, could they make it a disc-less version of the One S that's also able to stream "Nextbox" games?
Don't see why it would take more bandwidth when they are both doing the exact same thing sending a video and if Microsoft are combing local and cloud rendering as shown in the Kahawai video above it would require less bandwidth than Netflix.
I'll buy whatever the higher end version is, but I'm genuinely interested in seeing this cloud stuff in action. I hope they are hard at work building the backbone if these are their plans. They have the infrastructure, but implementation is the key.
Nintendo and Sony simply can't do anything like this. We just don't know if it's something that is even worth doing right now.
Nintendo and Sony simply can't do anything like this. We just don't know if it's something that is even worth doing right now.
I can't see how this would lead to any negative response if the issues are user side. Obviously there'll be a few idiots out there who will blame MS but enough people are intelligent to understand the basics of the internet that the negative response wouldn't be that big.
Obviously if MS's servers went down on Christmas night that'd be a completely different story.
I agree My internet will be fine for it I expect, but trying to push it before it was ready was what tripped them up last time. Putting boxes in peoples homes that don't work on day one will/could be a disaster for them, although I don't think they'll launch the streaming box in all regions, in fact; I think that's the real reason that there's two of them...
The ping to the closest azure data center is more important though
We're talking about average users though. We'll all be fine, of course we will, but parents buying the new "fortnight box" Christmas 2020 might not. That it's actually their fault isn't going to make a difference.
The price is going to be the primary consideration, not; does our internet at home work with it... "Well the X1 works fine, so this will work" is going to be the thought process.
Last edited:
This feels pretty much in line with Microsoft's approach right now. Imagine an Xbox Game Pass box, essentially, and a classic console for the rest of us. Makes perfect sense.
You take the latest (now out of date) frame from a video feed sent by the server, and move bits of that frame around based on your latest up-to-date game world state calculated locally to give a visual approximation of the latest game world state that you can return to the player immediately.
So my last frame from the server has a rendering of my sword at location x. In the next frame it should be at location y, based on new input. You take the pixels for the sword in the last frame, and transform them to where they should be in the next frame based on the new sword location. To do that well, you might have more than just the color data for the sword from the server - it might also send depth, so you can reconstruct the world position of each sword pixel, and move them more correctly to their new location in the new frame.
A very simplistic example. Things can get more complicated.
In the general case, you won't get an exactly correct next-frame input response. But you hopefully get a decent enough approximation that gives you an impression of a plausible, and most importantly immediate, response. At least for input responses made up of rigid transformations of objects already in-frame, in the last frame received from the server.
(This is assuming a system with no local rendering, and 'just' local cpu-side simulation...i.e. not like the Kahawai video linked earlier, where the client renders a low fidelity version of the game. That would work differently, but would have higher local device requirements and/or put greater limitations on the complexity of games. I doubt they're doing it that way.)
Last edited:
What about controller inputs? User data?? Whatever anti-piracy/cheating implementation they're running???
A lot of people are throwing around $99 as a price point but I wouldn't be too sure about that. Read this quote again:
I think of a new streaming box price to be around $99-$150. Apple TV 4k is like $150 and the Amazon Cube is $120. I think we could at least expect that level of tech in this so I'd put baseline lowest price around there. Then a new-gen console price is around $400-$500 so we know it will be cheaper than that. Probably significantly so or else it starts to not look like such a great deal.
So personally I'd expect this to be more around $200-$250 depending on what local hardware they need to put in it (plus don't forget it will likely include a controller etc). Maybe even $300. I could be wrong, but I'd set my expectations more around there.
I think of a new streaming box price to be around $99-$150. Apple TV 4k is like $150 and the Amazon Cube is $120. I think we could at least expect that level of tech in this so I'd put baseline lowest price around there. Then a new-gen console price is around $400-$500 so we know it will be cheaper than that. Probably significantly so or else it starts to not look like such a great deal.
So personally I'd expect this to be more around $200-$250 depending on what local hardware they need to put in it (plus don't forget it will likely include a controller etc). Maybe even $300. I could be wrong, but I'd set my expectations more around there.
That data still represents far less bandwidth than streaming video does, by a considerable amount.
Did you not read they're also releasing a traditional console, they're giving you options, it's not a must, you can still buy the traditional console.
A Kawahai style system might require less bandwidth.
A system where the local box is only doing cpu-side simulation calculations - which is more like what the report describes - would probably require more bandwidth than a regular video stream - you'd probably want depth data and more from the server, on top of the regular video/colour data.
Yes, but they're essentially streaming video on top of that. Ergo; it will probably take more bandwidth than Netflix.
This is extremely exciting, A 99$ next generation console could open up the Xbox platform to a new level of engagement. Xbox is coming for those supposed two billion gamers.
When do we hear that google, MS, Amazon, Netflix, Sony, Valve, are going to somehow keep comcast (and others who have actual large swaths of people in a monopoly for broadband) from using this to make a lot of money by slapping fees / caps on this stuff?
I mean, if my ISP said "OK, new price for services to your house is $500 a month for internet", I'd be shit out of luck..I'd have no internet. Or if they said $200, I'd pay it, and feel so so taken advantage of.
I mean, if my ISP said "OK, new price for services to your house is $500 a month for internet", I'd be shit out of luck..I'd have no internet. Or if they said $200, I'd pay it, and feel so so taken advantage of.
Like everything a current multiplayer game would have to send to a server.....
Yup. Phil's entire push is Xbox Anywhere. I'm sure his grand vision is to take your Xbox library and play it tons of devices ranging from PC to high end console to mobile phone to a handheld to a streaming device. I hope they pull it off.
So they're answer for the cloud/lag issue is a hybrid solution. Clever shit if it works. Nice outside the box thinking by MS
Most video streaming data is delivered in burst packets with buffering at the client layer. You don't continuously download at a constant bitrate, you have bursts of transmissions, which are then sequenced locally. This burst activity would be less helpful in a game streaming context if you are constantly uploading local data that needs to digested server side and impacts what comes next from the server for video. Streaming works well because the video is static. Now Live Streaming using some magic sauce to get by this, but this is still a little different and more complicated than that.
Again, we don't know the exact implementation, but if the Kahawai demo is to be believed it should require less data than traditional video streaming due to it not having to send full frames all the time. Once we know more about the service and can do actual technical measurements on bandwidth usage, it's moot to discuss what uses more since no one can give an accurate answer.
Every current multiplayer game has hard data at each end.
It's going to take more bandwidth than Netflix.
You may be right, and because Microsoft is developing a traditional console this is more of an academic discussion, but VR is consumer ready now. There's still plenty of improvement, sure, but I wonder how comparable cloud assisted consoles are right now.
Or those who already own Xbox ones, would we even need to buy the new box? The cloud is handling all the heavy lifting, so an xb1 should be able to host the cloud app. That's fascinating.