• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

"The Power of the Cloud" - what happened?

So, I guess we need a poll on who thinks the Crackdown Cloud demo is a) complete bullshit vs b) a viable proof of concept.

That MS oversold what cloud computing would do for the average One game is obvious, but I don't think you dedicate an entire GDC demo and hire a dev under the name "Cloudgine" for tech that doesn't f'n exist. Of course, if it's only ever used for a handful of games, you can still chastise MS for putting out weaker hardware and promising the world for a tech platform that only affected 1% of the lineup.
 
Wasn't that Kampfheld guy banned because he couldn't prove where he worked/talked BS about where he worked? I don't remember it anymore...

There are so many PSN problems, that making fun of MS (in this case) is just dumb.
What has PSN to do with this? Other than having more problems with DDoS and having server overloads (which doesn't really happen anymore) the service itself works fine.
 
So, I guess we need a poll on who thinks the Crackdown Cloud demo is a) complete bullshit vs b) a viable proof of concept.

That MS oversold what cloud computing would do for the average One game is obvious, but I don't think you dedicate an entire GDC demo and hire a dev under the name "Cloudgine" for tech that doesn't f'n exist. Of course, if it's only ever used for a handful of games, you can still chastise MS for putting out weaker hardware and promising the world for a tech platform that only affected 1% of the lineup.

It's obviously a viable proof of concept.

Whether or not we actually see that tech end up in Crackdown itself is to be seen.
 
I put many, MANY hours into Quake and UT across multiple builds. I also loved different bot types you could get. I use to play with aggressive bots, defensive ones, etc.. I had a blast against their bots, but they're absolutely NOTHING like the AI in Titanfall.

In the hoard mode they added it can be crazy fighting two Titans on your own, let alone 8 Titans and 30 different kinds of ground troops.

Watching the AI Titans drop smoke, put up shields to catch your projectiles, sending your projectiles back at you while simultaneously hitting you with a missile barrage and then hitting you with heavy machine gun fire after dropping your shields....that's nothing like Q3 or UT...lol.

That's a linear expansion of behaviors; not a marked difference in type. Halo is notable for it's AI behaviors in fairly large groups and how they react to your position and attempt to pin you down and flank you; all run locally. Can it make a difference. Maybe. TLoU had notable problems on the ps3 because a lack of memory and CPU; so their enemies did weird things. A server side bot could have more resources and get stuck less often. But you trade that off w/ latency.
 
Exactly. It's not like Live is bullet proof.



I've had the same attitude towards a cloud powered CD from the beginning. I've expressed a number of times how worried I am about the game using cloud powered tech when they could just go with an offline solution or just not have destruction as it's not something CD needs.

I haven't been as openly critical because I honestly did start to believe that Kampf was legit (claims of verified helped with this) and wanted to see if his claims had any basis before ripping into it completely. I did want to believe, but after he was exposed as a hype man, the realisation that CD was completely fucked quickly sank in and has left me rather bitter...

I won't deny the tech they demoed looked really good and I hope it works out as they hope, but I don't have any faith it will. I've seen nothing to suggest that tech will work on a much larger scale with lots of player influenced events happening and still be able to work as intended.

No faith in Dave Jones?
 
It's obviously a viable proof of concept.

Whether or not we actually see that tech end up in Crackdown itself is to be seen.

This. It will be interesting to see how they would even sell the consumers on it, make Crackdown online only a la Titanfall? Or just tell the consumer that it would benefit them by being online (and have Gold required possibly???) or the game would be nerfed.
 
Not being real-time is a incredible silly argument for trying to disprove the effect of server-side calculations.
I am not trying to disprove the effet of server-side calculations. I'm trying to say that they are very limited in real-time applications and that the "power of the cloud" isn't limited do XBO (hell, most big developers are already using their own cloud for analytics and simulating player behaviour with that). It can be used for innovative feature that have real value, just like the Drivatar system.

The way you're describing what Drivatars do is extremely simple and portrays the technology in a way that makes it look trivial and pointless.

It does a hell of a lot more than just record your racing line and attitude.
But that's pretty much how racing game AI works; it's isn't trival or pointless, but apart from the algorithms for creating the drivatars (which is absolute algorithm magic no doubt), as a system it doesn't differ that much from other games. It still uses a base racing line, it still uses a behaviour tree, it still has attributes for certain situtations, it's just that all of that is affected by a set that is calculated within the cloud from attributes and heatmaps created by players.

EDIT: my point being, that while it's a great system and most likely a magnificient algorithm, the baseline for the whole AI isn't any different. It certainly adds value to the system, but the way the system is computed in game isn't different from other games, the set of values are.
 
This. It will be interesting to see how they would even sell the consumers on it, make Crackdown online only a la Titanfall? Or just tell the consumer that it would benefit them by being online (and have Gold required possibly???) or the game would be nerfed.

I'd lean towards wanting it to be online-only, as long as Microsoft can successfully keep up the lofty claims they're making regarding destructible environments and particle effects.
 
That's a linear expansion of behaviors; not a marked difference in type. Halo is notable for it's AI behaviors in fairly large groups and how they react to your position and attempt to pin you down and flank you; all run locally. Can it make a difference. Maybe. TLoU had notable problems on the ps3 because a lack of memory and CPU; so their enemies did weird things. A server side bot could have more resources and get stuck less often. But you trade that off w/ latency.

Unless i'm mistaken wouldn't the point of server based processing be to free up the console's cpu to do other things rather than calculate physics or AI? You could do the AI calculations locally and achieve the same result, but if you can instead just offload it you can use those cycles for something else. In some cases it might not be about it being impossible to do locally, but that you can free the console from having to do it locally. Where as something like the Crackdown demo they say it's something that couldn't be done locally.
 
I am not trying to disprove the effet of server-side calculations. I'm trying to say that they are very limited in real-time applications and that the "power of the cloud" isn't limited do XBO (hell, most big developers are already using their own cloud for analytics and simulating player behaviour with that). It can be used for innovative feature that have real value, just like the Drivatar system.

Who in the conversation claimed that cloud computing was only possible on Microsoft's infrastructure? Or that it was being used to directly improve graphics (as opposed to handling one part of the engine remotely)? It sounds like you're reacting to things I'm not seeing, and I want to be sure I'm not missing something,
 
It's obvious to anyone who has played games with standard server-side bots like Quake 3, and also spent a decent amount of time with Titanfall, to realize what the difference is. Yes, at their core they are still server-side bots. However, the difference in artificial intelligence of the bots is where it matters.

Apparently, it's hard to make dumb AI without the power of the cloud. And you know that AI is a code right? Why would I give credit to the servers that run it and not the programmers that engineered it?
 
IBut that's pretty much how racing game AI works; it's isn't trival or pointless, but apart from the algorithms for creating the drivatars (which is absolute algorithm magic no doubt), as a system it doesn't differ that much from other games. It still uses a base racing line, it still uses a behaviour tree, it still has attributes for certain situtations, it's just that all of that is affected by a set that is calculated within the cloud from attributes and heatmaps created by players.

To be fair, once you get to the point of reducing it to that extent, you can essentially claim nothing is substantially different from prior technologies, whether it be graphics, AI, sound or anything else. Everything is basically just a more complex form of the simulations we were doing before. What sort of difference in AI would you actually accept as substantial if you don't mind me asking?

Apparently, it's hard to make dumb AI without the power of the cloud.

Actually it is, which is why games like Outrun 2 (and its followups) have no traffic when played online.
 
I'd lean towards wanting it to be online-only, as long as Microsoft can successfully keep up the lofty claims they're making regarding destructible environments and particle effects.

And then that brings about another question about Crackdown 3. Will there be no traditional single player component, but rather something that puts you into a world with other players just wrecking a bunch of shit (which could be fun)? You would still go about your own objectives but you would see other players destroying buildings.
 
That's a linear expansion of behaviors; not a marked difference in type. Halo is notable for it's AI behaviors in fairly large groups and how they react to your position and attempt to pin you down and flank you; all run locally. Can it make a difference. Maybe. TLoU had notable problems on the ps3 because a lack of memory and CPU; so their enemies did weird things. A server side bot could have more resources and get stuck less often. But you trade that off w/ latency.

I don't necessarily disagree with that.

The possibility of latency is always there when you offload some things. The hoard mode AI is certainly more capable than the initial modes we got too from what I've played. Heck, in the player vs player modes some of the AI stands around like it's a rock. In the hoard mode you have to be careful on the ground and on structures. I've seen some AI scaling and leaping to the high ground right away...sort of human like...in that mode. You still have your cannon fodder, but it seems more menacing and actually capable there.

Needless to say, I actually go back and play that mode..I've not touched the other modes in months, so they may have patched that up some.
 
Who in the conversation claimed that cloud computing was only possible on Microsoft's infrastructure? Or that it was being used to directly improve graphics (as opposed to handling one part of the engine remotely)? It sounds like you're reacting to things I'm not seeing, and I want to be sure I'm not missing something,
I'm just reacting to what Microsoft themselves have said and what the previous threads before the release have been about. Just google "xbo one cloud graphics" or physics and look for their statements. This thread is about Microsoft and XBO and how they sold the "The Power of the Cloud" as their features, which is why I'm talking about how XBO should have made it easier to implement these features (which in some ways they have, but mostly for their own developers) and how those features would be XBO exclusive (according to MS).

I'll drop by tomorrow, but I really gotta get some sleep now, since we are supposed to push a build for a competition this thursday.
 
Unless i'm mistaken wouldn't the point of server based processing be to free up the console's cpu to do other things rather than calculate physics or AI? You could do the AI calculations locally and achieve the same result, but if you can instead just offload it you can use those cycles for something else.

For multiplayer; synchronization is the driving issues over complexity. When you take it off local you have a lot of things to factor; more things that can break or come in late. I'm not sure how often the CPU is the limiting factor for most console games. The PS3/360 had anemic CPU's and they did alright. The hard part of AI anyways is making them convincing AND fun; and the devs seem to have stuff that would work on a super under powered PS3 CPU. Doing it for performance looks like a huge gamble. So far it hasn't been the case.
 
Wasn't that Kampfheld guy banned because he couldn't prove where he worked/talked BS about where he worked? I don't remember it anymore...
Someone at Cloudgine posted saying Kampfheld doesn't work there, and isn't working on the next crackdown. Kampfheld never claimed to work at cloudgine, but instead that he was apart of the other team not yet announced developing crackdown (cloudgine is only doing the engine, not actually developing the game). But since the cloudgine guy said he wasn't working on the next crackdown at all, I guess that's what got him his ban.

I'm still a believer of kampfheld personally... Is he perm banned or is it temp?
 
I'm confused; it is for their plans for F5.
I'm pretty sure that article is speaking of the original Forza Motorsport which would make it about ten years old.

F5's Drivatars aren't processed on the console. They are processed on servers that take account both individual and community-wide telemetry.
 
I'd lean towards wanting it to be online-only, as long as Microsoft can successfully keep up the lofty claims they're making regarding destructible environments and particle effects.

Hell yeah they should make it online only. They don't need to pussy foot around with a halfassed middle of the road solution trying to appease folks. I'm sure developing for two scenarios would be a pain in the ass and cost a lot more. I would rather they put that money into the game instead of a scenario I will not encounter.
 
I'm not the one claiming it isn't "special" or unique, I'm saying that if it wasn't unique or special where are the other examples of it?

I don't know, maybe people (i.e 3rd party devs) don't care that much about something that isn't all that special and a tacked on gimmick or marketing point?
Like gyros in the sixaxis.
You'd think MS themselves would use and abuse the thing though, since it's supposedly a massive advantage they have over the competition.

Drivatars in the original Forza are not anywhere near what we have now.

Well I'd hope AI has made some advancements in 15 years. The whole concept is exactly the same though.
 
I don't necessarily disagree with that.

The possibility of latency is always there when you offload some things. The hoard mode AI is certainly more capable than the initial modes we got too from what I've played. Heck, in the player vs player modes some of the AI stands around like it's a rock. In the hoard mode you have to be careful on the ground and on structures. I've seen some AI scaling and leaping to the high ground right away...sort of human like...in that mode. You still have your cannon fodder, but it seems more menacing and actually capable there.

Needless to say, I actually go back and play that mode..I've not touched the other modes in months, so they may have patched that up some.

Really good ai comes more form understanding the options the player has and mirroring that rather than hardware I think. Like SC2 has some good built in AI. It can challenge people at multiple levels without too much cheating. A simplistic AI would just have a resource bonus and be more aggressive but SC2 will do timed attacks, expand, and attack your resources more realistically at higher difficulties.
 
Actually it is, which is why games like Outrun 2 (and its followups) have no traffic when played online.

That's a hardware limitation.

The point of the whole conversation, which people here is missing, about how true Microsoft claims about how the "cloud" could do a revolutionary servers side computation and other loads of BS about how will make the Xbone a lot more powerful because of it.

And we've seen nothing other than them being just servers.
 
Someone at Cloudgine posted saying Kampfheld doesn't work there, and isn't working on the next crackdown. Kampfheld never claimed to work at cloudgine, but instead that he was apart of the other team not yet announced developing crackdown (cloudgine is only doing the engine, not actually developing the game). But since the cloudgine guy said he wasn't working on the next crackdown at all, I guess that's what got him his ban.

I'm still a believer of kampfheld personally... Is he perm banned or is it temp?

It also seemed like he asked for the ban. But only the mods really know what happened, I guess.

That being said, I don't see how we can throw out everything he said. The golf ball thing he showed, do we just throw that out now that he is banned? Do we conclude it was just an elaborate hoax? I still think a lot of the stiff he talked about is likely to happen.
 
I'm pretty sure that article is speaking of the original Forza Motorsport which would make it about ten years old.

F5's Drivatars aren't processed on the console. They are processed on servers that take account both individual and community-wide telemetry.

The AI is run locally. The stats compiled locally and sent up, partially digested then distributed. I've been trying to find the interview with a dev where they spoke on 'a few cycles' int he cloud before distribution. However the links I can find are to the defunct OXM.
 
And then that brings about another question about Crackdown 3. Will there be no traditional single player component, but rather something that puts you into a world with other players just wrecking a bunch of shit (which could be fun)? You would still go about your own objectives but you would see other players destroying buildings.

That would be hot.
 
"The power of the cloud" was PR they used because Xbox One is less powerful than the Playstation 4. Forza 5 drivatars and Titanfall bots was nothing special. Maybe Crackdown 3 will bring something to the table, but I won`t buy that anyway, because the game would be online required.

Mark Cerny also said that using the "cloud" to boost up the quality of the graphics doesn`t work. Anyway, they stopped using it after Phil Spencer became boss.

It's not just about boosting graphics, as seen in that GDC demo, engine performance in regards to consistent framerate is a benefit.

http://www.redgamingtech.com/phil-spencer-discusses-xbox-one-directx-12-cloud-gaming/
 
The AI is run locally. The stats compiled locally and sent up, partially digested then distributed.
Yeah, I didn't mean to suggest that Drivatars are run from the cloud in real time. Player telemetry is capture during play on the local console and periodically uploaded to the cloud servers for batch processing. That batch processing also takes into account community-wide behaviors and results in Drivatar profiles which are then downloaded by the game client (at startup?) and run as local AI opponents.

That's how I understand it anyway, from whatever I could find on the topic around F5's release.

It works, and in some ways works very well. Drivers have personalities and are more dynamic than anything I've seen in any other racing games.
 
EA said SimCity required a online connection because the power of the cloud allows them to simulate up to 100k sims.

Turns out that was BS and Cities: Skylines has a more advanced agent AI running all locally.

And yet people still wonder why many are skeptical of any claims regarding cloud processing.

At the end of the day MS is going to have to prove that the power of the cloud is something that is actually useful for regular games. Server side bots and drivatars are not enough.
 
It also seemed like he asked for the ban. But only the mods really know what happened, I guess.

That being said, I don't see how we can throw it everything he said. The golf ball thing he showed, do we just throw that out now that he is banned? Do we conclude it was just an elaborate hoax? I still think a lot of the stiff he talked about is likely to happen.

and also when he posted it, bish verified him, so I'm really confused about that. what info did he send bish that let him stay until that happened? He never said he worked for cloudgine, iirc. just heavily implied he was working on the next crackdown in some way.
 
It also seemed like he asked for the ban. But only the mods really know what happened, I guess.

That being said, I don't see how we can throw it everything he said. The golf ball thing he showed, do we just throw that out now that he is banned? Do we conclude it was just an elaborate hoax? I still think a lot of the stiff he talked about is likely to happen.
Yeah I think he said something along the lines of "If the mods want to ban me because of that then fine, just remember to look back at my posts after E3"

If the power of the cloud is real then we'll find out at E3. Crackdown will be shown, and it's up to MS/the developers to make us believe the power of the cloud can help games in ways that they are trying to make us believe. If they can't do that, then that's on them and I think everyone will then agree that the power of the cloud is bullshit, but calling it bullshit right now when there will be a game revolving around it due to be revealed in 10 weeks time is a bit silly. I'm skeptical too, and being skeptical is fine, but it's too early to flat out call it bullshit at this point in time in my opinion.

No matter what they say or try to show people will still argue against it, just like the early crackdown demo a few months back, if the power of the cloud is real then they should just sit back and let the games do the talking, which it seems like they are doing now. If it's real then we won't need MS to try and explain it to us, we'll see it for ourselves. If it's not, then CD3 will fall flat like Sims city.
 
From what I can tell it was ninety percent bullshit from the start. A bunch of knowledgeable people here and across the internet gave what i thought were good explanations for why most of what Microsoft was promising was bs, but people kept believing the MS hype machine.
 
So, I guess we need a poll on who thinks the Crackdown Cloud demo is a) complete bullshit vs b) a viable proof of concept.

The problem of showing what is/isn't possible when demoing a new tech is that you're not showing every possibility. You're showing the old way and a new way that uses your tech that you're promoting... and that's when you stop looking. You're promoting your new way and it's absolutely valid that maybe some old tech doesn't scale remotely as well as the new tech. That doesn't mean there isn't a third (or fourth, fifth or even sixth) other way to accomplish the task that don't take advantage of the magic new tech you're trying to promote.

Are they wrong? Not technically. Are they misleading? Absolutely (it's marketing, and making your product/tech look good is priority one).

The claims of DirectX 12 remind me a lot of the claims of The Cloud improving gaming. Yes, using a particular set of tools you're going to see improvements that are absolutely measurable... but hey... you know what? Not everyone is using DirectX 11 to begin with, so the improvements you'll see going from 11 to 12 aren't really pertinent to someone using alternate tech anyway.

The day I see one of these companies come out and say their tech will revolutionize gaming... and THEN IT DOES... then I'll take notice. Otherwise, I'm going to assume it's yet another in a long line of contrived demos designed to make a bullet point and sell a product.

The Cloud will not revolutionize destructible environments... or environmental lighting... and Microsoft's cloud isn't going to corner the market on multiplayer server resources... or AI calculations...

People are tooting their own horns. Don't fall for the hype... let them do the work to actually deliver a PRODUCT that'll prove the point(s) they're trying to convince you they're making now.
 
Yeah, I didn't mean to suggest that Drivatars are run from the cloud in real time. Player telemetry is capture during play on the local console and periodically uploaded to the cloud servers for batch processing. That batch processing also takes into account community-wide behaviors and results in Drivatar profiles which are then downloaded by the game client (at startup?) and run as local AI opponents.

That's how I understand it anyway, from whatever I could find on the topic around F5's release.

It works, and in some ways works very well. Drivers have personalities and are more dynamic than anything I've seen in any other racing games.

The flavor is really good. Having personality-less AI like in GT5; all driving perfect lines at variable speeds is less fun for sure.

I'm just saying the most interesting thing about that is what the local part was; not the distribution method. They could have used dedicated servers or just peer to peer and had most of interesting parts in tact. Having a more centralized places for the data means they can tweak and the system did need that tweaking.

There are benifits to the cloud; just not int he way it was initially sold. If their policy for servers are maintained it means on XB1 older low pop games can still get a multiplayer server spun up while on the PS4 the servers will go dark.

But claims that it'll help graphics seem a bit of a stretch.
 
There are benifits to the cloud; just not int he way it was initially sold.
On this we agree 100%

I'd also love to see MS show that small population games can have server-support for the long term through spinning up just as many instances are needed. This really needs to happen.
 
If the power of the cloud is real then we'll find out at E3. Crackdown will be shown, and it's up to MS/the developers to make us believe the power of the cloud can help games in ways that they are trying to make us believe.

Even that wouldn't be enough. The game needs to be launched and tested in real life situations, where people have less than perfect bandwidth and latency with the computing servers.
 
That's a hardware limitation.

The point of the whole conversation, which people here is missing, about how true Microsoft claims about how the "cloud" could do a revolutionary servers side computation and other loads of BS about how will make the Xbone a lot more powerful because of it.

And we've seen nothing other than them being just servers.

Isn't that the whole point of our discussion though? Hardware limitations. At the end of the day, the Azure servers are also just hardware... it's not some mystical magic. It's also not only the local hardware that's the only limiting factor, a single person's connection would also be unsuited to handle all of the required data sync'ing between players.

Also, it absolutely can do revolutionary server side computation... if anyone actually employs it for that. This isn't something completely theoretical, it's what servers do. This in turn does actually make the Xbox One (or any other console using such an implementation) more capable, as it is freed from using whatever resources would have been required to do these computations locally. Some of these computations might otherwise absorbed the majority of the console's processing abilities. Saying something is "more powerful" doesn't always have to equate to "better graphics", even though in this case, it very well may. You'd never know though, as you'd never see what the game would have looked like if tasked with computing everything locally.
 
Isn't that the whole point of our discussion though? Hardware limitations. At the end of the day, the Azure servers are also just hardware... it's not some mystical magic. It's also not only the local hardware that's the only limiting factor, a single person's connection would also be unsuited to handle all of the required data sync'ing between players.

Also, it absolutely can do revolutionary server side computation... if anyone actually employs it for that. This isn't something completely theoretical, it's what servers do. This in turn does actually make the Xbox One (or any other console using such an implementation) more capable, as it is freed from using whatever resources would have been required to do these computations locally. Some of these computations might otherwise absorbed the majority of the console's processing abilities. Saying something is "more powerful" doesn't always have to equate to "better graphics", even though in this case, it very well may. You'd never know though, as you'd never see what the game would have looked like if tasked with computing everything locally.

I think the key disagreement is we nay sayers believe the latency will keep the feature very minor. While the believers think it will be expansive.

There are significant engineering problems in the mix and the return on investment is minor.

Even with a 'all in the cloud' feature like PS-now. Available computing power is a huge issue which drives up the price; and latency makes it impractical for a lot of games. Turn based RPG - yes. COD:MW2 - No.
 
I would not doubt this i going to be huge by the end of this gen....its just taking a while, thee things are very much in development. Crack down will blow minds I have no doubt,
 
Isn't that the whole point of our discussion though? Hardware limitations. At the end of the day, the Azure servers are also just hardware... it's not some mystical magic. It's also not only the local hardware that's the only limiting factor, a single person's connection would also be unsuited to handle all of the required data sync'ing between players.

That's not revolutionary, that's been around since the mid 90's.

Also, it absolutely can do revolutionary server side computation... if anyone actually employs it for that. This isn't something completely theoretical, it's what servers do. This in turn does actually make the Xbox One (or any other console using such an implementation) more capable, as it is freed from using whatever resources would have been required to do these computations locally. Some of these computations might otherwise absorbed the majority of the console's processing abilities. Saying something is "more powerful" doesn't always have to equate to "better graphics", even though in this case, it very well may. You'd never know though, as you'd never see what the game would have looked like if tasked with computing everything locally.

As I said in my previous post, all of that is BS until we see the applications run in real life life scenarios. Otherwise, it's hogwash.
 
So, I guess we need a poll on who thinks the Crackdown Cloud demo is a) complete bullshit vs b) a viable proof of concept.

That MS oversold what cloud computing would do for the average One game is obvious, but I don't think you dedicate an entire GDC demo and hire a dev under the name "Cloudgine" for tech that doesn't f'n exist. Of course, if it's only ever used for a handful of games, you can still chastise MS for putting out weaker hardware and promising the world for a tech platform that only affected 1% of the lineup.

Ofc what they have shown in that demo is possible.
The problem with cloud gaming isn't how it works under perfect conditions e.g. the Crackdown demo, but how it works under realistic conditions.
We have no data about that Crackdown demo.
How far away was the cloud, how much bandwidth did the client have?
The ms cloud is like having multiple computer's in a network sharing the work between them - of course this works.
The problem with the cloud is that these other computer's are very far away and all data has to go through the internet massively increasing latency.
The nearer you are to a cloud center the better your gaming experience potentially will be.
 
The cloud is is very real and its main advantage is scalability of resources on demand. It is already replacing local infrastructure in a lot of mid-sized to big companies (provided internet traffic costs and the pricing model don't offset the savings). It really is the future for a lot of business applications, but not so much for games.
It's only viable for games in proceses not heavily reliant on latency and bandwidth. It sure has its uses (AI, session management, data crunching) but nowhere near what Microsoft tried to sell at launch.
If you think developers will start using azure resources (and development time) in gimmicks like grass anytime soon, then you're delusional. Not because it's not feasible but because it doesn't make any sense businesswise.
 
Even that wouldn't be enough. The game needs to be launched and tested in real life situations, where people have less than perfect bandwidth and latency with the computing servers.
Well, that's a different situation.. There's people who don't think the cloud can help gaming at all, then there's those that think it can help games, but latency will be a problem.

I can see them announcing some sort of beta/demo to test it out in your own home though, delorean is supposed to help solve the latency issue but again people are skeptical. The only way to make people believe is to let them see for themselves and I'd hope MS knows that now, so going on about delorean as a solution to latency won't do much if people can't see it for themselves.
 
I think the key disagreement is we nay sayers believe the latency will keep the feature very minor. While the believers think it will be expansive.

There are significant engineering problems in the mix and the return on investment is minor.

Even with a 'all in the cloud' feature like PS-now. Available computing power is a huge issue which drives up the price; and latency makes it impractical for a lot of games. Turn based RPG - yes. COD:MW2 - No.

Latency is definitely a limiting factor yes. But not to the extent that I think a lot of people believe it is. It would basically never be a real issue for AI, as if a game can be played online between two players, then the server-side AI is essentially granted an assload of time to make its decisions, as it will immediately perceive player A's actions to moment they hit the server, unlike how player B would have to wait for the new state to be transferred down to them locally before they begin to make any decisions based on the new information being displayed.

You even get more time for collision based applications than you may think, as time as perceived by you, isn't the same as perceived by the machine. You've probably played numerous games where it'll slow down, or pan to a fancy camera angle or whatever, in response to a dramatic shot or close call etc. In real life these sorts of things would tend to require replay footage, however the game can do them on the fly because it already knew the outcome of what's gonna happen 5 seconds from now, the moment you took the shot. This allows plenty of time to crunch some absolutely stupid numbers in preparation for an event it knows will occur soon. By the time it actually happens, you already have the required information sent to your local console.

Also, comparisons with general cloud gaming are pretty pointless. Processing resources only need be spent for a cloud enhanced game when something on the local box triggers a requirement. if nothing big is occurring in your game, then the server is doing nothing (or just handing out results to other players). When you sit idle when playing Killzone 2 on PSNow however, it's always running a full game of Killzone 2, with all the bandwidth and processing requirements of transferring your 720p footage of nothing happening. The latency issue is far more compounded as well, as every input you press has to go from you, to the server and back before your motion is visible to you. This isn't the case when the core game is running locally, which is why playing a racing game over PSNow is more problematic than playing the same game online.

That's not revolutionary, that's been around since the mid 90's.

Depends what you consider revolutionary really. As mentioned before, it's pretty easy to argue that pretty much no technological progression is truly evolution, as it's always really just more complex implementations of what we did before. I imagine you must absolutely stuff like the iPhone reveal.

I personally consider the fact that I can create a custom game in FH2, invite my brother, and then drive 15 miles down the road to meet him, with all the traffic I pass being synced for both of us, to be pretty revolutionary. Last gen we couldn't even a few basic fixed path cars in Outrun to survive the online transition. It's something that otherwise has only been the domain of persistent MMO type games like Test Drive Unlimited.. except now it's an instance of our own, created on demand when we want it, and we didn't even have to pay to rent a server like back in my old Quake 3 days.

As I said in my previous post, all of that is BS until we see the applications run in real life life scenarios. Otherwise, it's hogwash.

Hey, you were saying it was BS that they "could" do these things. Not that they are doing them right now. They absolutely CAN.. whether they will is a different matter.
 
Latency is definitely a limiting factor yes. But not to the extent that I think a lot of people believe it is. It would basically never be a real issue for AI, as if a game can be played online between two players, then the server-side AI is essentially granted an assload of time to make its decisions, as it will immediately perceive player A's actions to moment they hit the server, unlike how player B would have to wait for the new state to be transferred down to them locally before they begin to make any decisions based on the new information being displayed.

In multi-player it exists because the time for server to player is better than player to player and for syncronization. It would really depend on what AI because a optimistic 30ms ping means you are 1-2 frames behind. At 100ms you are 3-7 frames behind and that can be significant. Doing it for a single player game seems less sensible. You're accepting more lag for 'better' AI. It just doesn't seem like a better solution.

You even get more time for collision based applications than you may think, as time as perceived by you, isn't the same as perceived by the machine. You've probably played numerous games where it'll slow down, or pan to a fancy camera angle or whatever, in response to a dramatic shot or close call etc. In real life these sorts of things would tend to require replay footage, however the game can do them on the fly because it already knew the outcome of what's gonna happen 5 seconds from now, the moment you took the shot. This allows plenty of time to crunch some absolutely stupid numbers in preparation for an event it knows will occur soon. By the time it actually happens, you already have the required information sent to your local console.

Any sort of cloud calculated 'effect' would need something like that to hide the transaction. Like a huge dust cloud then playing the calculated animation. But couldn't you just can 7 animations and run one at random? You could even can it based on a few variables. You wouldn't have to worry about the latency of the transaction exceeding the 'hiding' effects runtime.

Also, comparisons with general cloud gaming are pretty pointless. Processing resources only need be spent for a cloud enhanced game when something on the local box triggers a requirement. if nothing big is occurring in your game, then the server is doing nothing (or just handing out results to other players). When you sit idle when playing Killzone 2 on PSNow however, it's always running a full game of Killzone 2, with all the bandwidth and processing requirements of transferring your 720p footage of nothing happening. The latency issue is far more compounded as well, as every input you press has to go from you, to the server and back before your motion is visible to you. This isn't the case when the core game is running locally, which is why playing a racing game over PSNow is more problematic than playing the same game online.

The key though, is one is just piping up controller input piping down video directly. A easier conceptual problem than finding a computing problem that can be broken up in a way which is agreeable to small outgoing pipes, latency, network traffic variability, and peak load. The set of problems that are agreeable are small and not super impressive. So the results will be small and not super impressive.
 
Top Bottom