News & Rumors: Xbox One (codename Durango)

Status
Not open for further replies.
i actually really like most everything about this console (could mostly care less about no used games, online every 24 etc) but...

why the hell did ms not "splurge" on a few more cu's? you know, 14, 16, something to close the alleged gap a bit.

so shortsighted for such a wealthy company.

I am at two minds about why they didn't do this.

1. They don't care that Sony is ahead, they only care about there vision

2. They expect the cloud(tm) to makeup the difference.

I hope its not the second.

EDIT:.

This might explain it.

http://www.viddler.com/v/10738ac4?secret=81714898

~5:40

"We purposefully did not target the highest end graphics. We targeted it more as a broad entertainment play"
 
But it is not sending the computed data back down piecemeal to be used in local software. It is rendering the completed frame and sending it. What will a game do when the required data gets delayed on its trip? What if it is lost completely?

It doesn't matter if it is lost if it isn't something that directly affects gameplay.

They sound new to me, MMOs don't do physics calculations and send results to local computers, they send meta-data like mob position. They don't do it to offload the local computer, they do it because they are serving hundreds of other players which need the same shared data. How can a dev who needs more FLOPS in a single player game use something unreliable with random latency?

Maybe I'm missing the point of MS's cloud features.

Yes. :)

While I and others can see the potential for gameplay affecting features. I think it's more likely that developers might play it safe and use it for game enhancing features.

So think of it another way. Tesselation does absolutely nothing for gameplay. Its primary virtue is that it allows for a better looking game world. The same could be said for HDR. The same could be said for textured polygons. As far as gameplay is concerned gourard shaded polygons would be just fine. You don't NEED all that fancy stuff like shadows or accurate lighting or even character models with more than 100 polygons.

And yet a significant amount of render/game time is spent on all these things that only serve to enhance the visual appeal of a game.

There are lots of things that don't require low latency that can enhance the visual appeal of a game. That can make games more immersive and believable.

I mentioned previous a case where you had more NPCs with better AI respresenting random people you would see while walking around the busy streets of New York City, Paris, London, Rome, Tokyo, etc. There isn't enough computing power in even the most powerful PC to make that a reality. But maybe the cloud can slowly make it less empty and less repetitive than the open world games currently are by hosting that AI. AI That is re-useable across multiple people playing that game potentially allowing for hundreds of NPCs featuring more robust and random AI than currently exists in any game in existence. Hell, the cloud could allow the developer to uniquely texture many of those NPCs as well as change the clothing they use based on the time of year and the season in real life. You just stream the appropriate textures from the cloud based on time of year. So rather than shipping a game with all possible clothing combinations, etc. you have the cloud store most of them and just ship what you usually have in games currently with the game disk. Hell, the developer can also then add new textures/clothing without requiring the user to download anything before playing the game as it's all done while they are playing.

Or a somewhat less demanding example that Bkilian used, where instead of a 2D crowd in the stands of a sports title or racing game, you have a more fleshed out and believable 3D crowd with AI that can actually react properly to what is going on in the stadium or race track. In such a case, it doesn't even matter if there's 2 seconds of lag.

Or going even farther. You could offload global world illumination to the cloud while keeping local illumination local. Global illumination not affecting the character and what is immediately around doesn't require low latency. Throw in offloading things like birds flying around, or cars driving on a distant street and that would fit right along with the offloaded illumination. Something like this probably requires more computing power than is realistically available for games, but 2 years from now? 4 years from now?

You can even extend that out to extensive weather affects with accurate physics, globally (as in affecting the world the game is in) it doesn't require low latency. Anything that actually affects the player would require it being done locally. So, being able to relieve the console from doing ALL of the game worlds physics in such a situation would allow for a much larger physics calculation budget for the character and anything affecting the character.

Hell, you can theoretically render the entire world outside of a few feet around the character and enemies on the cloud without it affecting gameplay in any way. Basically the only things that HAVE to be done locally for the most responsive play possible is anything the character can potentially interact with in the next, say, 2 seconds or less. Otherwise does it matter if NPC #21347 in the scene walking down the street with random AI that has decided they are on a window shopping excursion who stops to talk to NPC #12234 who is heading to the dry cleaners is delayed by 4 or 5 seconds? Since it doesn't rely on player interaction, no, it doesn't.

I could go on an on about ways in which cloud computing could enhance gameplay.

Basically just like a developer spends the majority of their frame rate budget on non-character interactive things such as rendering the scene, so too you could use cloud computing to enable scene enhancing effects or additions that aren't possible without more resources.

Now, as to what's possible? That's far more difficult to say. How much cloud capacity will be allocated for this? How much of a safety margin will be built in to prevent a breakdown of the system due to overwhelming load if everyone started playing a game with cloud compute enhanced features?

Regards,
SB
 
i actually really like most everything about this console (could mostly care less about no used games, online every 24 etc) but...

why the hell did ms not "splurge" on a few more cu's? you know, 14, 16, something to close the alleged gap a bit. what would it have cost them incrementally? 2 bucks? 5?

i realize they couldn't have short term reacted like this, but it should have been in their long term planning to have a bit more beef, ya know?

so shortsighted for such a wealthy company.

i have a feeling they'll "win" anyway, based on tv and all the other crap, and have their shitty power doesn't matter thinking validated.

the answer apprears to be they view the device more as an entertainment hub that also plays videogames rather then a videogame system that acts as an entertainment hub. In being so they probably set a goal of that games look like enough of a step up from the x360 and that was considered enough on the specs end.
 
To be honest all of the cloud computing examples do not make the Xbox One a better alternative as all of these are easily doable on any console or PC.

Even if it can do extraordinary things with cloud computing it most probably won't be a Xbox One specific feature.

Is there anything special on the Xbox One or on their servers that allows it to do cloud computing that the other consoles can't? I really don't see it.
 
To be honest all of the cloud computing examples do not make the Xbox One a better alternative as all of these are easily doable on any console or PC. Even if it can do extraordinary things with cloud computing it most probably won't be a Xbox One specific feature.

It depends on whether Sony are willing to invest enough money in the cloud server infrastructure to match what Microsoft will allocate for Xbox use.

Something like that isn't something individual publishers can do without taking massive losses due to the variability of load they would have to handle. Heck, it isn't even the easiest thing to do for Microsoft.

But in this case Microsoft has invested heavily in this area as has their competition (Google).

Regards,
SB
 
I get that today's presentation was about the hardware and the ecosystem but for me the games were disappointing. Not only from the standpoint of visuals but also in terms of the core game play. The games they showed off looked very derivative of what we have come to expect. For that matter I was similarly not impressed by much of what Sony showed off a few months ago as well.

IMO the publishers are going to have to step up the innovation otherwise why would anyone want to invest in a new hardware platform? Slightly better visuals isn't enough of a reason especially if more rigorous DRM and used game policies are the areas which show the biggest leaps over the previous generation....
 
To be honest all of the cloud computing examples do not make the Xbox One a better alternative as all of these are easily doable on any console or PC.

Even if it can do extraordinary things with cloud computing it most probably won't be a Xbox One specific feature.

Is there anything special on the Xbox One or on their servers that allows it to do cloud computing that the other consoles can't? I really don't see it.

True, and it still reduces any performance benefit PS4 has over XB1.

I see Microsoft's strategy with this (and the 360) almost like an endurance race. Physiologically a small increase in performance comes at an exponential increase in cost to the body. A cost that can bite the front runner in the second half of the race. Advanced technology has similar characteristics. Slightly less performance, significantly less cost.

I think Nintendo took this a bit far though, obviously :)
 
It depends on whether Sony are willing to invest enough money in the cloud server infrastructure to match what Microsoft will allocate for Xbox use.

Something like that isn't something individual publishers can do without taking massive losses due to the variability of load they would have to handle. Heck, it isn't even the easiest thing to do for Microsoft.

But in this case Microsoft has invested heavily in this area as has their competition (Google).

Regards,
SB

Yep. The difference between Microsoft doing this and EA doing it alone is the investment in the infrastructure and the expertise in doing so. I have no idea what benefits cloud computing will bring. They really haven't said much at all about it yet. The idea is interesting. Maybe there will be some talk at E3 that is a little more concrete. Even on a minute scale, cloud computing on a world-wide network is a massive undertaking.
 
http://www.engadget.com/2013/05/21/...blog/?utm_medium=feed&utm_source=Feed_Classic

"Tim Stevens 12:25 PM
Nick is saying that "to get all of this processing out of the box" is a challenge too.The new CPU core can do six CPU operations per core per cycle, on an eight-core CPU."

anyone know if 6 ops per cycle per core is the same as stock jaguar?


apparently not standard jag

Stock Jaguar => 2 instructions, 4 ops.
Xbox One CPU => 3 instructions, 6 ops?


http://semiaccurate.com/forums/showpost.php?p=183829&postcount=993
 
Last edited by a moderator:
True, and it still reduces any performance benefit PS4 has over XB1.

I see Microsoft's strategy with this (and the 360) almost like an endurance race. Physiologically a small increase in performance comes at an exponential increase in cost to the body. A cost that can bite the front runner in the second half of the race. Advanced technology has similar characteristics. Slightly less performance, significantly less cost.

I think Nintendo took this a bit far though, obviously :)

This probably sounds a bit cynical but how much of this so called cloud computing is really just marketing to justify the cost of a Live account and downplay inferior specs?

If I'm paying 5 bucks a month for MP I might think hey thats free on WiiU or PS4 whereas this way I feel like I am getting some additional value even though it is likely to be next to impossible to quantify.
 
I think the hundreds of millions Sony spent on Gaikai should maybe indicate Sony has some investment in Cloud delivery. Just a hunch ;)
 
I'd be interested to see Azure (Amazon EC2 etc) load levels on a 7 day cycle. Is is likely that peak gaming loads in a local area are not going to conflict with peak load times for business?

Gaming may be heavily subsidised by business cloud services.
 
MS is playing up the cloud crap a bit more than expected in this Q&A

Q: How do consumers benefit by being connected to the cloud?

A: The cloud makes every experience better and more accessible. Because Xbox One is powered by the cloud:

• Your games have more power available to create new gameplay, persistent worlds, and deeper experiences.


Q: Will Xbox One be backward compatible with my existing games?

A: Xbox One hardware is not compatible with Xbox 360 games. We designed Xbox One to play an entirely new generation of games—games that are architected to take full advantage of state-of-the-art processors and the infinite power of the cloud. We care very much about the investment you have made in Xbox 360 and will continue to support it with a pipeline of new games and new apps well into the future.
 
I think the hundreds of millions Sony spent on Gaikai should maybe indicate Sony has some investment in Cloud delivery. Just a hunch ;)

What Gaikai is doing is significantly different than what Microsoft and Google are doing in the cloud. And significantly more expensive.

It depends on whether Sony wants to keep going with what Gaikai is designed to do, stream games or morph it into something more like what Microsoft and Google are doing (cloud computing).

It may seem superficially similar but it's different in many ways. One of the most notable differences when it comes to cost of implementation is that Gaikai's service requires video encoding hardware as well as video rendering hardware. You have to first render the game, then encode the subsequent video into something that can be streamed to a user. Each part of that requires specialized hardware. And in some cases that might be a 1:1 relationship. To stream PS3 games for example, you basically need 1 PS3 in the server farm for each person playing a streamed PS3 game. Note: that doesn't mean a retail box PS3. :)

Microsoft and Google on the other hand have focused almost entirely on general compute capabilities along with extensive storage capabilities. Due to the business critical nature of their market focus it also means they have invested heavily in redundant backup and security features. Things that aren't necessary for a game streaming service.

And it's one thing to design a service that is very focused on streaming games, and one that that is all about distributed cloud computing that must be useable by a large variety of applications.

Regards,
SB
 
This probably sounds a bit cynical but how much of this so called cloud computing is really just marketing to justify the cost of a Live account and downplay inferior specs?

If I'm paying 5 bucks a month for MP I might think hey thats free on WiiU or PS4 whereas this way I feel like I am getting some additional value even though it is likely to be next to impossible to quantify.

Services generally are hard to quantify other than the bill. Am I really getting my $350/hr value out of lawyers? I often don't think so, but I still use them... Service providers always have these allegations thrown against them.

However, you may be cynical or you be right on the money. There is no doubt that cloud services seem to be the direction the world is heading as when it is fully realised, there are many benefits. With Sony's interest in Gaikai, I suspect they feel it's worth quite a bit too.

My point is that if all console manufacturers are going down this path, in a few years, no one is going to be too fussed about 12 or 14 or 18 CUs.

Fibre is rolling out around the world, much as ADSL was starting to roll out at the turn of the century. Once fibre is ubiquitous, the practicality of cloud compute services improves dramatically. This may not happen this gen (or it might). When it does happen however, you probably won't want to be waiting for the next generation of consoles to be catching up.
 
That's a good way to look at it. Sony is a more media focused company too.

But anyway, let's just hope this unlimited cloud power from Xbox One doesn't have the same story as SimCity did. Cloud compute I imagine will be very application specific, at least for a long time. I'm going to guess that Respawn's new IP might be one of fewer games to try and get the ball rolling first for those kind of games.
 
That's a good way to look at it. Sony is a more media focused company too.

But anyway, let's just hope this unlimited cloud power from Xbox One doesn't have the same story as SimCity did. Cloud compute I imagine will be very application specific, at least for a long time. I'm going to guess that Respawn's new IP might be one of fewer games to try and get the ball rolling first for those kind of games.

Sim City is really about the most horrible example of cloud compute that there is. :) Other developers and publishers have dealt with online (server based game resolution) for far longer, with much more done on the servers, without the trainwreck in service that EA has gone through with Sim City.

But Sim City is also a perfect example of where a publisher isn't necessarily up to the task of managing online server based game computing.

If Microsoft are really serious about this, they are going to have to host the cloud services (which it appears they will be doing) as well as investing heavily into it on the assumption that the variable load can be mitigated with enough software releases that they aren't in a situation where you may go for months at a time with very low server load then have a month of two of extremely high server load, rinse and repeat.

Publishers get caught out on that because in order to not go bankrupt they provision for slightly above the average of those two extremes. Which means at launch they suffer from insufficient server capacity and then months later they are paying for server capacity that isn't being used.

So that's where things could potentially work for Microsoft where it's more difficult for individual publishers not named Blizzard, NCSoft, etc. They can manage game releases from multiple publishers to attempt to have high utilization year around.

Regards,
SB
 
Status
Not open for further replies.
Back
Top