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

Crackdown 3 uses cloud-based physics processing to enable environmental destruction

GAF pre gamescom- hahaha, power of the cloud!!! Lolz!!

GAF post gamescom- oh but Sony can do this too if they feel like it!!!

I need the cloud to help me process all the back flipping :)
 
Who cares about this:
oxdruas.gif


When you can have 1080p?

What if Crackdown 3 is 1080p?

Now what?

Well then we'll just belittle the games graphical style or accuse it of being all flash and no substance...

Or we'll just say it isn't fun

We'll do whatever it takes to avoid giving credit when it's due...

That's what fanboys are for...
 
No, azure has always had the ability to spin up on demand... That's its claim to fame...

Cloudgine portion is its work on engines that are designed to use local and cloud processing together deliver a game...

yes I know that, thats why, if you go read the post carefully, i said isn't it the Cloudgine technology here the more important part not Azure, since they can probably tailor the Cloudgine tech to be used on other clouds, be it AWS or whatever new company comes up with one
 
Who cares about this:
oxdruas.gif


When you can have 1080p?

What if Crackdown 3 is 1080p?

Now what?

This looks really fun. It looks like they built the entire engine around complex collisions.

But that's not direct footage is it?

The contrast looks way off. Way too dark and crushed.
 
yes I know that, thats why, if you go read the post carefully, i said isn't it the Cloudgine technology here the more important part not Azure, since they can probably tailor the Cloudgine tech to be used on other clouds, be it AWS or whatever new company comes up with one

Yeah I misread you...

Well, they are both important... The important part of Azure is that it's inhouse, which makes this use case economically sound...

I'm curious to know the extent of MS' and cloudgine's relationship. Surely MS provided a ton of R&D, funding and engineering help to pull this off. Surely they wouldn't let this tech be redistributed so easily.
 
Well, they are both important... The important part of Azure is that it's inhouse, which makes this use case economically sound...

cloud margins are razor thin (the profit thoroughly comes from the scale; it is to compute what Guangzhou is to manufacturing), so there is really not that much of a benefit of running it in house.

And as I pointed out earlier, the economics of it are somewhat strange, and that is just considering electricity cost, no maintenance or IT.
 
I remain rather skeptical (my daily work consists of running computation on often thousands of computers). Still smells strongly of marketing to me (which does not mean it does not run on servers) - it's a cool marketing line, and it somehow doubles down on the dubious marketing of the xbox one regarding cloud tech.

First, it's true the physics are impressive (not blown away, though) but it's not something that games traditionally focus on. . It's better than red faction guerilla, but that's one of the few games that goes for that kind of feature, and that was last gen. I am personally significantly more impressed by something like flex from nvidia, which runs real time on a GPU (yes, one much faster than the one in an xbox one).
https://www.youtube.com/watch?v=0vMllWRX6SI

The money aspect is a bit strange, too.
Azure is not magic. Big data center have driven down the cost of IT and software deployment. They provide tools so that people can use compute as needed, and charge them for it.

It drove down the maintenance/IT cost- but not to zero. It does not drive the electricity cost to 0, in fact, it barely does anything to reduce it.

So yes, let's pretend it takes 14 servers to run the game (is that the number I read somewhere?). Let's count 100W per server (a very generous estimate). Let's say an average gamer is going to play the game for 20 hours. That's 28kW of electricity, something that will cost Microsoft 4 dollars or so. It might be small, but its nothing to scoff at; for a $60 game, for which about 50% go to the publisher/console owner (in this case the same thing, Microsoft), that's a significant amount of money. It also makes it difficult to lower the price.
A gamer that plays the game hundreds of hours ends up costing money to microsoft.
That's bonkers. I guess it's doable with IAP, or continued monetization of some sort (ads? live is not enough).

We could probably work it out if I could find which model PowerEdge they're using...
 
Well I think we would see far less doubters if this was a Sony demo. I don't see the thought process in some of the concerns raised. They did a ton a research into this prior to making a game.

Also where was this scepticism when Sony showed streaming games which is far more demanding from a bandwidth comparison.
 
We could probably work it out if I could find which model PowerEdge they're using...

Compute/W is not _that_ variable, so what we'd really need to know is the load. I suspect the 20 xbox ones is inflated - I actually don't think it's all that high. But I am surprised that they'd be willing to bear any cost normally born by the customer.

But if it makes the game that much more fun and creates an addiction they can monetize somehow, perhaps.

Well I think we would see far less doubters if this was a Sony demo. I don't see the thought process in some of the concerns raised. They did a ton a research into this prior to making a game.
This is the kind of annoying comment that really brings nothing to the discussion. I'd be equally skeptical if sony showed this.

As for your example, I think it's interesting sony does not offer the possibility of buying the game at all- you can only rent it - at a pretty shitty price on top of that
( though I doubt anyone actually uses playstation now much - it's quite impressive it could be done, but ultimately I assume the latency kills it for most people)
 
Well then we'll just belittle the games graphical style or accuse it of being all flash and no substance...

Or we'll just say it isn't fun

We'll do whatever it takes to avoid giving credit when it's due...

That's what fanboys are for...

Irony so thick you could cut it with a (cloud computing enhanced) knife
 
Averages are very misleading.

I live in a rural location, and even I've got 70mb fibre.

Agreed. I live in a town of about 1300 people, and have a 200/50 connection, with 1000/200 slated for the end of the year. I would guess all the dial-uppers are to blame for the low average.
 
Well I think we would see far less doubters if this was a Sony demo. I don't see the thought process in some of the concerns raised. They did a ton a research into this prior to making a game.

Also where was this scepticism when Sony showed streaming games which is far more demanding from a bandwidth comparison.

oh it was (and still is) everywhere.. where have you been?
 
Well I think we would see far less doubters if this was a Sony demo. I don't see the thought process in some of the concerns raised. They did a ton a research into this prior to making a game.

Also where was this scepticism when Sony showed streaming games which is far more demanding from a bandwidth comparison.

Check the PS Now threads. Plenty of skepticism to be found there regarding bandwidth and latency.
 
I"m not going to say Crackdown is definitely going to be 1080p (because the singleplayer is still going to have to process physics locally), but part of the reason this tech is so impressive is precisely because it can boost certain aspects of a game to an exponential level (in this case, physics for Crackdown 3) while freeing up processing power for the console to use on other things (making it easier potentially to reach a higher resolution, framerate, etc.).

That's the assumption, and they have a year or so to prove that, but so far outside of the physics, the game ins't exactly an absolute stunner, not in terms of it's visuals, lighting, textures, animations or IQ. Then again, it's single player might be much better looking, or not if campaign gameplay doesn't use cloud.

crackdown-3201508-04-15008-1438711956.jpg


If leveraging cloud for physics computation leaves you more resources to use elsewhere, in terms of urban open world cream of the crop graphics on the Xbox One, these are the kinds of visuals Crackdown will compete with.

 
This looks really fun. It looks like they built the entire engine around complex collisions.

But that's not direct footage is it?

The contrast looks way off. Way too dark and crushed.

That's footage from the trailer, but I think it's been "enhanced" as someone has sharpened it to shit
 
Compute/W is not _that_ variable, so what we'd really need to know is the load. I suspect the 20 xbox ones is inflated - I actually don't think it's all that high. But I am surprised that they'd be willing to bear any cost normally born by the customer.

But if it makes the game that much more fun and creates an addiction they can monetize somehow, perhaps.

Compute on the bone is what, 1.2tf? So taking the PR line at 14x we're looking at 16.8tf, or about one AMD Ultra workstation.
 
If leveraging cloud for physics computation leaves you more resources to use elsewhere, in terms of urban open world cream of the crop graphics on the Xbox One, these are the kinds of visuals Crackdown will compete with.

I always thought the original crackdown was a really distinct looking pretty game, it kind of looked like a comic book, with it's flat shading, it's subtle cell shading style outlines around characters and it's bright orange explosions.

This does look like a next gen version of that to me
 
GAF pre gamescom- hahaha, power of the cloud!!! Lolz!!

GAF post gamescom- oh but Sony can do this too if they feel like it!!!

I need the cloud to help me process all the back flipping :)

Clearly.

Well, it's one step closer to wide acceptance and the potential of generating more interesting conversation around its applications, which is better than the tired lolz. Cheers to progress.
 
cloud margins are razor thin (the profit thoroughly comes from the scale; it is to compute what Guangzhou is to manufacturing), so there is really not that much of a benefit of running it in house.

And as I pointed out earlier, the economics of it are somewhat strange, and that is just considering electricity cost, no maintenance or IT.

I feel like you are trivializing what's going on here...

Firstly, most gaming revenue is front loaded... The amount a company will spend on servers will ultimately limited by that. So unless they are generating significant revenue throughout the titles life , there will be a limited spend.

Secondly a typical game will use 1 VM to host multiple gaming sessions. crackdown will use up to 20VMs at once for one session. That will eat up a developers server budget pretty quickly...

Being inhouse is a definite advantage...
 
Compute on the bone is what, 1.2tf? So taking the PR line at 14x we're looking at 16.8tf, or about one AMD Ultra workstation.

It's not the whole amount CPU powe of the Xbox x20, it's actually just the typical level of CPU that a game normally allocates to in game physics x20.
 
GAF pre gamescom- hahaha, power of the cloud!!! Lolz!!

GAF post gamescom- oh but Sony can do this too if they feel like it!!!

I need the cloud to help me process all the back flipping :)

Well, people never had complelling arguments why cloud computing wouldn't work. Especially since we slowly saw it implemented in more meaningful ways(AI in Titanfall, Drivatar in Forza and Crackdown 3 is now the most obvious).
All people said was that Microsoft was overpromising like they did with Kinect but while Kinect had some issues that couldn't be overcome, the cloud never had the issues, at least not when it comes to outsourcing stuff like physics.

Was kinda funny to see people looking at PsNow... "Whole game rendered in the cloud? No problem." ...and then at MS approach... "Outsourcing specific tasks to the cloud? Against all the laws of nature!!"

But just like that people have no complelling arguments why only Microsoft would be able to do it and nobody else.
There is no special hardware in the console needed to make it cloud compatible. The only question is, are others willing to spend the money to make it happen?

I think right now Crackdown 3 is a test that'll show whether or not deeper cloud computing integration is worth it. If its a success I could see Microsoft pushing it in its 1st party titles and when Sony has something similar in place I could see 3rd parties implementing the tech, too.


Personally I think physics alone can open up so many new possiblities when it comes to gameplay, but there is also AI and other thing that can be vastly improved with the cloud.
So I hope it goes mainstream, but for 3rd parties to make use of the cloud for their games it needed to be possible on both consoles and PC.
Because as the CD3 developer mentioned in the IGN interview, a fully destructible world has to be built like a real building, with an interior, something you ususally don't have. Devs won't be going to great lenghts to enhance games via cloud for a single plattform. Its too much work, but if they can do it everywhere they can build their games for it from the ground up.
 
Clearly.

Well, it's one step closer to wide acceptance and the potential of generating more interesting conversation around its applications, which is better than the tired lolz. Cheers to progress.

Yep, it's actually nice read a thread about something Xbox related where the trolling people are actually in the minority.
 
That's the assumption, and they have a year or so to prove that, but so far outside of the physics, the game ins't exactly an absolute stunner, not in terms of it's visuals, lighting, textures, animations or IQ. Then again, it's single player might be much better looking, or not if campaign gameplay doesn't use cloud.

If leveraging cloud for physics computation leaves you more resources to use elsewhere, in terms of urban open world cream of the crop graphics on the Xbox One, these are the kinds of visuals Crackdown will compete with.

A game can be impressive without having to be impressive in every single area. Those examples are all going to look hilariously static when compared in motion... it'd be like having a character with long hair that doesn't flow... sure the detail may be there, but as compared to even a simpler model with L'Oreal-tech, you're going to notice the differences immediately.
 
A game can be impressive without having to be impressive in every single area. Those examples are all going to look hilariously static when compared in motion... it'd be like having a character with long hair that doesn't flow... sure the detail may be there, but as compared to even a simpler model with L'Oreal-tech, you're going to notice the differences immediately.

Why would they look hilariously static in motion? If anything they'd look even better, as a lot of them have serious post processing effects that are far better appreciated in motion, e.g. things like smoke, fog, dust, wind, weather, debris, realistic cloth motion etc.
 
In a broad sense, yeah the framerate supposedly won't chug with all that destruction but I'm talking about a 30 to 60 fps kind of improvement.

The 30 to 60fps kind of improvement is a direct reflection of what the game is trying to accomplish. It's not a hard limitation of the hardware (not even for a NES). So as long as the cloud is suitable for processing whatever is causing the game to run smoothly, then yes it can (and in this case is probably more like 10fps to 60fps). It won't however be able to take any game and do the same...

.. well actually, even that's bollocks, because if you cane stream a 60fps image, then you can basically do whatever you want, and handle as little as you want locally.

Why would they look hilariously static in motion? If anything they'd look even better, as a lot of them have serious post processing effects that are far better appreciated in motion, e.g. things like smoke, fog, dust, wind, weather, debris, realistic cloth motion etc.

The backgrounds are completely rigid, regardless of what you do to them. It's not something that generally stands out to you consciously when playing... but is being recognised as you play. You know that rocket shouldn't only leave a black mark on this brick wall, but you accept it as that's how it always is in games. Once that's not how it always is in games though... then these things are going to stand out a lot.
 
Sure, it's real. That's great news.

Doesn't mean they weren't seemingly blowing smoke up my ass for months when they showed literally nothing while talking a big game about the cloud being secret special sauce.

I DON'T GET WHY THIS NEW TECHNOLOGY WASN'T BEING USED RIGHT AWAY WAH WAH WAH I SHOULD HAVE HAD DAY ONE CLOUD GAMES WAH WAH WAH

That's what you sound like.
 
That's the assumption, and they have a year or so to prove that, but so far outside of the physics, the game ins't exactly an absolute stunner, not in terms of it's visuals, lighting, textures, animations or IQ. Then again, it's single player might be much better looking, or not if campaign gameplay doesn't use cloud.


If leveraging cloud for physics computation leaves you more resources to use elsewhere, in terms of urban open world cream of the crop graphics on the Xbox One, these are the kinds of visuals Crackdown will compete with.

Are you seriously showing "concern" about the visuals of a game that's over 1 year from release? Then also comparing said game to finished and released games with completely different artstyles?

Come on man.
 
Sony has nothing to gain for putting on a tech demo for PR, they are not the ones with the weaker console. MS is just leveraging what they can to change the 'underpowered' message. It's not like MS is going to have dozens of different titles wasting Azure time, they just need one example to wave around, red meat for the fans.
Ok. Whatever you say buddy.
 
So I hope it goes mainstream, but for 3rd parties to make use of the cloud for their games it needed to be possible on both consoles and PC.
Because as the CD3 developer mentioned in the IGN interview, a fully destructible build has to be built like a real building, with an interior, something you ususally don't have. Devs won't be going to great lenghts to enhance games via cloud for a single plattform. Its too much work, but if they can do it everywhere they can build their games for it from the ground up.

Shouldn't be too much of a problem with modern buildings - pretty much anything over 3 stories these days is based around lift core, shell and wrap, and the first two of those could be procedurally generated or simply C&Pd as assets.

Building something like a Creed open world would be a different matter, but you're probably not going to need to simulate building demolition in a historical game...
 
I feel like you are trivializing what's going on here...

Firstly, most gaming revenue is front loaded... The amount a company will spend on servers will ultimately limited by that. So unless they are generating significant revenue throughout the titles life , there will be a limited spend.

Secondly a typical game will use 1 VM to host multiple gaming sessions. crackdown will use up to 20VMs at once for one session. That will eat up a developers server budget pretty quickly...

Being inhouse is a definite advantage...

I am sorry, but this does not make sense. Of course, sony does not need to / probably shouldn't buy their own servers. Even a company the scale of netflix uses (used?) AWS, for instance.

The thing is, running a VM has a cost. hardware cost (amortized, weak), electricity, IT and maintenance. Whether amazon operates those server, or microsoft does, those costs don't vanish.

If microsoft uses its own resources and services, it only has to pay those exact costs.
If sony buys from aws, it has to pay those exact costs, plus whatever the margins amazon makes on those costs.
My point is: that margin is very thin, so the two costs are very comparable. Neither is a great value proposition, especially if it's given away against a one time revenue.

Well, people never had complelling arguments why cloud computing wouldn't work.
There are plenty of arguments why cloud computing is not a great idea for gaming. You just choose to ignore them.

If you think about it, it's unusual that we are told about 'how' those achievements are given. It's the equivalent of saying 'don't judge my product based on its merits, but rather on how hard i worked to offer them to you', which should not enter a consumer's appreciation of said product

If no one had known about azure running titanfall AI or drivatar, I doubt people would have paid nearly as much attention to the features (this is not saying they are bad, but that we wouldn't have cared as much about how they were delivered).

The reason it's marketed that way is not because the titanfall AI is so impressive, or even crackdown physics being that much of a gamechanger - it's because it also constitutes a promise of something even more exciting down the line. A promise I personally doubt they will deliver too much on.
 
Are you seriously showing "concern" about the visuals of a game that's over 1 year from release? Then also comparing said game to finished and released games with completely different artstyles?

Come on man.

It's not concern, it's called an opinion lol. And people tend to offer them for things that publishers show them in trailers, demo's and all the rest. The opinions range from positive, e.g., "man that destruction looks awesome", or "holy shit, the lighting looks amazing", or "damn these effects look amazing", but equally, sometimes the opinions are negative, e.g. "man the textures look really flat", or, "why are there so many jaggies" etc. These are all done with the knowledge that the games shown are still works in progress, and still have time to improve. Being shown before release however, does not nullify the opinions, or mean they cannot be offered.

Unless of course you're doing one of those, you're only allowed to offer an opinion if it's positive type things. Because you seem to have no issue with praising elements of this yet unreleased game, nor with me doing the same….
 
I am wondering what OT of this game going to be when it releases next year.

Crackdown 3 |OT| Clouds Cant Stop Me From Shinin

Crackdown 3 |OT| It's Raining Orbs, Hallelujah!

Crackdown 3 |OT| Like Agility Orbs Lost in the Rain
 
That's the assumption, and they have a year or so to prove that, but so far outside of the physics, the game ins't exactly an absolute stunner, not in terms of it's visuals, lighting, textures, animations or IQ. Then again, it's single player might be much better looking, or not if campaign gameplay doesn't use cloud.

Well, it's pre-alpha gameplay, and for all we know, they could be focusing on gameplay oriented things like framerate (and even resolution). It's way too early to tell, and this is also the first game really trying this. I imagine if this works that it'll only get better and better. Also, as was already said, this just looks like the next step in Crackdown's specific art style.

The singleplayer won't be using the cloud, so in Crackdown's case it may not matter, but games like Destiny could definitely benefit from this due to the online-only nature (not that Destiny will specifically).
 
Top Bottom