Microsoft Xbox Reveal Event - May 21, 2013

[Arwin]

I think there are some definite interesting possibilities for cloud computing to enhance gaming, especially multi-player and social gaming, with persistant world integration. GaiKai like features are also possible for certain situations. Of course this is not necessarily something that benefits from a 'platform holder', but is something that is inherently available to all who are willing to invest in writing the necessary software and finding a hosting partner/host yourself, so the question is what Microsoft brings to the table ...

 

[DJ12]

When they introduced the cloud it was as live. I imagine when push comes to shove that's all its for.

Multiplayer experiences like has been mentioned. No one is going to realistically send anything critical to the cloud. There are too many factors that can go wrong to rely on it.

Also surely nothing can be done on a per user basis anyway as the service will just get overloaded. Lets say 3 million people are playing cod at the same time how is that going to work if physics is done in the cloud or anything else?

 

[expletive]

If the cloud is only introduced as an idea, it might never happen. Wouldn't be the first time someone suggested a tech or use of a console that never materialised (Ken Kutaragi's numerous visionary ideas that never amounted to anything...). I certainly wouldn't buy XB1 trusting it'll get an online boost - I'd have to see it with my own eyes.

It's up to MS first parties to lead and define the platform in this way, particularly if it is as intrinsic to the platform as MS states. As ERP has stated (and i'm paraphrasing), multiplatform developers probably see this as very low ROI right now and investing in immature and unproven tech is not appealing. (See also:Rare making first gen Kinect games)

 

[Shifty Geezer]

It's up to MS first parties to lead and define the platform in this way...

That's if the platform ever even gets off the ground. Might be an idea that, on further reflection, MS decides isn't worth the investment and silently back-pedals on.

 

[expletive]

That's if the platform ever even gets off the ground. Might be an idea that, on further reflection, MS decides isn't worth the investment and silently back-pedals on.

3,000 to 300,000 servers sounds like more than an idea to me but i suppose they have plenty of other cloud services that could make use of the farm.

 

[Scott_Arm]

That's if the platform ever even gets off the ground. Might be an idea that, on further reflection, MS decides isn't worth the investment and silently back-pedals on.

I think they said some stuff would be showing at E3 that might demonstrate cloud computing. It might have been in that architecture panel video. It could have been the Turn10 guy. I don't remember. Still, this could end up being vaporware for 3rd parties if the financials don't make sense.

 

[Johnny Awesome]

Many of us here are into our third launch cycle on these boards.

I hope we've learned to take all the marketing talk AND hater talk with a grain of salt.

From where I'm sitting it looks like the next gen consoles are going to be about 10 x more powerful than what we have right now with more interesting add-ons like Kinect integrated and more dedicated servers for online gaming.

Sounds pretty fun to me.

 

[Shifty Geezer]

I haven't heard the original statement so am uncertain as to the current state of play, but AFAIK Azure is MS's cloud platform, and they want to invest in this no matter what. So they basically seem to be saying, "devs can use Azure," without necessarily committing to a minimum service standard for XB1 users. After a couple of developers dabble with Azure and find it's not much good for anything other than cloud storage maybe, MS could drop interest in the cloud and devs fall back the usual local-machine approach.

 

[Jwm]

I see Remedy's game just streaming the live action cut-scenes instead of maybe having them on the disk.

Forza - I see it adding to the crowds, and overall AI car count. How I am thinking about this, you would just send strings of data. Axis of player, position, blah, blah. No AI cars would be calculated on the "One", the cloud would be running the simulation on the track based on my data. While on my end, I would be running the track while getting the string of data back on each AI car. I think it could work rather well, just keeping things synced would maybe take a bit of thought.

An issue I have here is I could actually save local resources and improve other aspects of the game, graphics, sounds, etc. As long as I know for sure that I have a connection to the "Cloud", otherwise I would just default to more cars if the cloud is there vs. less cars if not.

I can see some simple uses that would free up local resources, but making it optional would limit my ability.

Box Art:
Better with Kinect
Requires Kinect
Better with the Cloud
Requires the Cloud

Oh my! Haha

I do love the ideas of what you can do with an RPG, along with turn based games.

 

[James Car]

I haven't heard the original statement so am uncertain as to the current state of play, but AFAIK Azure is MS's cloud platform, and they want to invest in this no matter what. So they basically seem to be saying, "devs can use Azure," without necessarily committing to a minimum service standard for XB1 users. After a couple of developers dabble with Azure and find it's not much good for anything other than cloud storage maybe, MS could drop interest in the cloud and devs fall back the usual local-machine approach.

Exactly. It's up to MS to be the one to invest in this approach and prove it's utility and how that then translates to sales. Fact of the matter is, unless they are opening their cloud to Sony and other platforms, a cloud based game on the bone will be an exclusive which at this point is a pretty scary proposition for a 3rd party without enough compensation to make it worth their while.


With this event being held on MS' campus, I see no reason why they could not have demoed a best case scenario of their cloud service assisting games if it was intended to be a near term selling point.

The fact they did not, and this was at their campus, tells me this things is years and years away (if ever, as you said).

 

[mrcorbo]

Claim that The voice command inputs during the conference were prbly fake :
http://me.ign.com/en/news/11383/fyi-the-xbox-one-voice-commands-weren-t-in-real-time


Also, when Yusuf said XbOx Game and switched to Forza, it wasn't a realtime forza game playing, it was just a part of the trailer later shown. It was just a video ! So the smooth switching we saw from tv to game to tv wasn't really real there. That was something I noticed then and there during the conference too. But if that wasn't game, then there is a complete chance that we were just watching a presentation on cue.

Switching between TV channels was the only thing I am sure was not representative of what you would expect to see in the real world. Going from game to TV to video etc., is effectively like alt-tabbing in Windows and you would expect that to be pretty snappy. Based on the system software architecture, 3GB of memory that's dedicated to keeping the system UI and apps resident in memory and the hardware display planes, I don't doubt that it really would be as quick as it was presented to be.

 

[bkilian]

Xbone also doesn't have a/V out either fore those who spent the last gen on sd tvs.

Yep. We tried to get them to remove optical out too (it increases the latency across the whole audio pipeline) but too many people still use it.

 

[Phil]

Forza - I see it adding to the crowds, and overall AI car count.

I really don't want to be mean here, but unless the other AI cars will look like lego blocks, expect the rendering to take place on your console. There's no point in upping the AI cars if the console itself only has so much performance to render them. Also, Forza, I hope at least, will still aim to be a 60fps title - so I don't think AI cars (as explained in my other post) will be particular doable with cloud resources.

The only thing that would realistically work is stuff that is not latency dependant. The higher the framerate you are running in your game, the more crucial it becomes not to be dependant on bandwidth and latency issues.

As I said, there is heaps of potential there, but developers will be keen to avoid things that may not work well over an average internet connection where you might have short overloads or slowdowns - or simply packet loss. People may be tolerant to a certain degree if you are playing a multiplayer online game, but a single player race where you have AI cars jumping around due to a bad connection might be a little unforgiving.

 

[bkilian]

Those people paid to play with other people, not a chatbot.

I don't know, I think I'll take an RPG with dialogue created by a talented writer and performed by a talented voice actor over machine generated quest bots anyday.

For the game important conversations, absolutely. But are you really suggesting that you want to go back to every damn random person telling you about how they used to be an adventurer like you, until they took an arrow to the knee? I don't know about you, but the extremely limited, stilted conversation and random comments from "unimportant" characters drives me nuts. I would much rather prefer AI generated incidental conversations, where they have a framework for how much the AI character knows and can talk about.

Here's a question. Are these simulated worlds going to live forever in the cloud? If I stop playing my Skyrim 2 for Xbox One am I going to find a year later all my deeds have been erased because Azure was tired of simulating a land I didn't seem to want anymore? Can games get retired from Cloud services, like an old EA sports title? Is my game only safe as long as my Xbox Live Gold membership checks clear? How will any of us play anything when Anonymous wages a month long DDOS attack on Xbox servers in anger over the used game situation?

Yep, things go away. I know we as humans have a ton of loss aversion, but it happens anyway. With elastic compute, they can scale their stuff so it's only using compute when people need it, so it is probably less likely to go away. But it still might. If this bothers you, I advise against buying either next gen console. And tell those kids to get off your lawn.

 

[wco81]

Greenberg specifically said emerging markets were a big target on major nelson.

That's basically why these slides existed, why they expect growth.

Hmm, wonder if they accounted for mobile devices competing for some of the spending as consoles.

Maybe they don't think games alone will drive consoles sales growth this generation, hence the emphasis on TV?

 

[Scott_Arm]

For the game important conversations, absolutely. But are you really suggesting that you want to go back to every damn random person telling you about how they used to be an adventurer like you, until they took an arrow to the knee? I don't know about you, but the extremely limited, stilted conversation and random comments from "unimportant" characters drives me nuts. I would much rather prefer AI generated incidental conversations, where they have a framework for how much the AI character knows and can talk about.

That's actually a very good idea. I suppose there's all kinds of NPC type behaviour that could be pushed from the cloud without having to roll out game updates.

Yep, things go away. I know we as humans have a ton of loss aversion, but it happens anyway. With elastic compute, they can scale their stuff so it's only using compute when people need it, so it is probably less likely to go away. But it still might. If this bothers you, I advise against buying either next gen console. And tell those kids to get off your lawn.

I think it's just becoming the reality that games are not going to be designed to last forever. That'll happen for PCs, consoles and handheld devices.

 

[MarkoIt]

Considering the box design, the product presentation, the market-focus, the integration of Kinect, I can tell you one thing: it won't be cheap. I expect the two "console" (seems kind of retro to use this term now) to be launched at the same price now.

 

[Nesh]

 

[scooby_dooby]

The most impressive uses for the cloud, IMO, is doing all the things that currently aren't done well in games and thus make them seem to lifeless and empty. Noone wants to spend expensive CPU cycles on AI for non-important NPCs. But making them less robotic and less repetitious (reusing the same AI or a very limited set of AI) would be just as large an advancement in gaming as something like HDR, physics, high res textures, tessellation, etc. And that's just a very simple example of something that could be done with little that has to be rethought.

Regards,
SB

No offense, but you gotta lay off the kool-aid.

Are you seriously suggesting a developer is going to tie a remote-server into their core game code, producing two totally different experiences for people with internet or without? Testing alone is a logistical nightmare there, not to mention you create a maintenance point for your project for eternity. Not to mention it's just a plan horrible idea from a technical standpoint with very little payoff or merit.

The only practical application I see for this stuff is some sort of server-side MMO style global computations, but that is ancient stuff we've been doing forever.

More advanced applications are syncing game saves across different devices, or even streaming from console to a remote device. What would be really cool, is if they used those 300,000 server to provide an OnLive type experience for any device (iOs, Android, Win8 etc), for all the games you own. Now THAT would be an exciting use of Cloud computing, but entirely too sensible for MS to pull off.

 

[scooby_dooby]

. I would much rather prefer AI generated incidental conversations, where they have a framework for how much the AI character knows and can talk about.

So shouldn't the dev just spend his time producing more content, and issue that as a straight forward patch?

* Less time spent fiddling with architecture == more content
* No need to design elaborate fallback mechanisms
* Less bugs / edge cases. Testing time vastly reduced.
* Better user experience (no drops / intermittent behavior)
* Cheaper (no server maintenance costs for lifetime of project)
* Portable. Not tied to any specific 3rd party services
* Trust. Your players know it will always work.

I fail to see any especially attractive benefits from a developers or gamer perspective. Only from MS's, who wants to sell Azure services to developers.

And If I were to do these things, I would design it to be cross-platform anyways, not Xbone dependent. Otherwise you end up with a game that can't easily be ported.