Crackdown 3 and the Power of the Cloud

[George P Global Moderator]

Makes sense it's limited to online MP only, it's still early days for this tech and we all don't have the ability to play online all the time without issue, but it's a step in the future IMO.  One thing I'd always liked to see in games, be it SP or even MP to a degree, is for any damage you do to buildings, cars and so on, to not automagically get fixed the next time that part of a level gets loaded up.  I want damage to last a bit and actually take time to repair itself, it'd be cool for open world games to actually show you how they're fixing that house you just trashed or that building is gone and replaced by something else, not right away, maybe the next day depending how quick the time goes by in game.   Just little things like that which could add to the realism. 

[BajiRav]

If they keep it to MP that is fine. SP worlds should be able to run on the grunt power of the console, as to prevent issues where there is no online connection, and preserve the ability to play titles you love 10 years from now. That and I'm really wary of Plus/Gold ever being able to stop non-paying subscribers playing SP content with the excuse SP relies on MP, so paywall activate! Stuff like the Souls series is fine online in SP, because you can play it all offline. Sim City is was not.

On another note this is one game I'll be jealous of if I don't pickup a One before it's out. Crackdown 1 was superb, 2, not so much, but feedback about 2 was given loud and clearly.

I kind of see your point but how many games do you really get that nostalgic about? I would rather have online enhanced SP NOW than worry about it playing 10-15 years down the line. I think there might be more factors to limiting it to MP only. I guess that way you get people to pay for the cloud compute. You also can't have a SP story when people can blow the whole map instead of playing through it.

[vcfan]

Crackdown 3 effectively turns your Xbox One into the most powerful console ever made

 

 

So, Crackdown 3 might be the most impressive demo I've ever seen. We'll have a full preview up later today, but suffice it to say that all that bluster about "leveraging the Cloud" to bring hitherto unseen levels of physics-based destruction is totally accurate.

 

 

If you, say, blow a chunk off of a building, which then flies into an adjacent area and smashes the window of the tower block next door, that neighbouring server then helps the original to resolve this. Destruction is persistent, and every piece of rubble remains interactive, and can continue to be shot, blown up or pushed around. Servers can be piled on servers to keep this working - in our demo, we saw 11 being used at once. Producer, Dave Jones, assured me that that was the tip of the digital iceberg.

 

 

You'd think this would require an immense internet connection to keep it rolling, not least when four players (this is the current maximum size for a multiplayer party, although it could increase) are doing the same thing in four separate corners of the city, but the relative ease of swapping information between Xbox and server means the strain is fairly small. Jones says that his team are optimising the game for a 2-4mbps connection.

http://www.gamesradar.com/crackdown-3-effectively-turns-your-xbox-one-most-powerful-console-ever-made/

 

Crackdown 3 hands-on preview – the Xbox One killer app

GameCentral reports back from Gamescom on what may be the most technically impressive console game ever made.

 

http://metro.co.uk/2015/08/05/crackdown-3-hands-on-preview-the-xbox-one-killer-app-5329378/#ixzz3i31sTbtw

 

[+Audioboxer Subscriber²]

I kind of see your point but how many games do you really get that nostalgic about? I would rather have online enhanced SP NOW than worry about it playing 10-15 years down the line. I think there might be more factors to limiting it to MP only. I guess that way you get people to pay for the cloud compute. You also can't have a SP story when people can blow the whole map instead of playing through it.

Well, I never want any SP game I've bought on disc to become defunct just because it's sitting on a shelf for a few years. No trade off is worth that. That's all I'm saying. I still have copies of PS1/PS2/Dreamcast games I enjoy owning, for what little to no time they get used, at least they'll work if I pop them in

Other devs do SP online enhanced fine, so I'm never going to accept it's ever a reasonable trade off for a SP game not to work unless it can communicate with a server.

[theyarecomingforyou]

I've played RF:Guerilla (good game) and it doesn't compare.  once the smaller buildings collapsed in that game, they would pretty much disappear.

rival gang has building locked down and don't want to come out... Blow up building, flush gang out... sounds like tangible benefit ...

Obviously it's not to the same fidelity but then that game is six years old and was released on the last-gen consoles. My point is that I'm sure a similar system could be implemented locally with minimal difference in fidelity. The limitation is still the GPU, as when the buildings break down the particle effects and increased polygon count still have to be rendered locally.

I don't see a modern PC having any trouble rendering the same scene. Here's a video from a few years back with all the physics done in real time:

https://www.youtube.com/watch?v=FIPu9_OGFgc

[Showan]

Obviously it's not to the same fidelity but then that game is six years old and was released on the last-gen consoles. My point is that I'm sure a similar system could be implemented locally with minimal difference in fidelity. The limitation is still the GPU, as when the buildings break down the particle effects and increased polygon count still have to be rendered locally.

I don't see a modern PC having any trouble rendering the same scene. Here's a video from a few years back with all the physics done in real time:

https://www.youtube.com/watch?v=FIPu9_OGFgc

And yet Battlefield 4 on a modern PC still has canned destruction.  I not knocking it because it looks good, but this looks better.

 

[+Asmodai MVC]

I think this is cool.  Kudos for MS making Xbox Live Compute free for Xbox game developers.

Let's be clear here though, cloud compute is always going to be online only because you can't have cloud compute without the cloud.  Also cloud compute doesn't make the Xbox One more powerful, you need cloud compute to do things via network servers (i.e. the cloud) that the console is too WEAK to do locally (this isn't a Xbox vs. PS thing, PS is likely too weak to do it offline as well).  This is the pioneering days here but as this technology matures you might see comparatively weak devices like Apple TV Next, Amazon Fire TV Next, nVidia Shield Android TV Next, etc. leveraging cloud compute to play games that approach the quality of the stand alone consoles when the stand alone consoles aren't cloud connected.  Which brings us to the last point, while MS is certainly a Pioneer in this area there is nothing in the concept of Cloud compute that is Xbox exclusive.  Nothing is stopping a developer from using cloud compute in a Amazon Fire TV game and hosting it in Amazon's EC2 cloud.  Heck a developer could make a game that uses cloud compute to hosted on Microsoft's public Azure cloud and release the game on PlayStation if they want to.  The big deal MS is doing is just making that cloud free for Xbox games and providing some game specific APIs for developers.  Microsoft provided APIs may or may not be desirable to any given developer (many like to roll their own) and free is awesome but one of the big things with the cloud is it's cheap and it scales according to use.

[Showan]

I think this is cool.  Kudos for MS making Xbox Live Compute free for Xbox game developers.

Let's be clear here though, cloud compute is always going to be online only because you can't have cloud compute without the cloud.  Also cloud compute doesn't make the Xbox One more powerful, you need cloud compute to do things via network servers (i.e. the cloud) that the console is too WEAK to do locally (this isn't a Xbox vs. PS thing, PS is likely too weak to do it offline as well).  This is the pioneering days here but as this technology matures you might see comparatively weak devices like Apple TV Next, Amazon Fire TV Next, nVidia Shield Android TV Next, etc. leveraging cloud compute to play games that approach the quality of the stand alone consoles when the stand alone consoles aren't cloud connected.  Which brings us to the last point, while MS is certainly a Pioneer in this area there is nothing in the concept of Cloud compute that is Xbox exclusive.  Nothing is stopping a developer from using cloud compute in a Amazon Fire TV game and hosting it in Amazon's EC2 cloud.  Heck a developer could make a game that uses cloud compute to hosted on Microsoft's public Azure cloud and release the game on PlayStation if they want to.  The big deal MS is doing is just making that cloud free for Xbox games and providing some game specific APIs for developers.  Microsoft provided APIs may or may not be desirable to any given developer (many like to roll their own) and free is awesome but one of the big things with the cloud is it's cheap and it scales according to use.

You are correct that, this isn't exclusive to MS.

But one thing we have all witnessed, this is expensive to do. Seems Very Expensive to do on a world wide level.

If this takes off, it seems that only those with deep pockets will be able to sustain it for long term.  

This being in-house for MS, doesn't incur them any extra charges.

It seems that MS shut their mouths, took the punches (in regards to cloud compute), and rolled with it.

Not sure how real world results with play out, but this tech demo is pretty awesome.

 

[+Eternal Tempest MVC]

For the benefit of others (Guerilla destruction example):

https://www.youtube.com/watch?v=U93lGcMC4mc

Did they have any huge high buildings in Guerilla? I dont recall huge buildings more 2/3 floor buildings in it but it has been a while. Also the scenes were a lot more spartan than the crackdown demo.
If its so easy to do without servers sony will obviously create something similar but only time will tell.

Goes back even further with the geo mod ? engine that was present in the Early Red Faction Games.

Almost everything was destructible. In multiplayer maps, and enough missiles, you could blast a tunnel to the opposing sides base.  
Latter Red Faction games were actually less destroyable. 

[+Asmodai MVC]

You are correct that, this isn't exclusive to MS.

But one thing we have all witnessed, this is expensive to do. Seems Very Expensive to do on a world wide level.

If this takes off, it seems that only those with deep pockets will be able to sustain it for long term.  

This being in-house for MS, doesn't incur them any extra charges.

It seems that MS shut their mouths, took the punches (in regards to cloud compute), and rolled with it.

Not sure how real world results with play out, but this tech demo is pretty awesome.

 

What's expensive to do?  Cloud computing is cheap, that's one of the major selling points for it.  It's extremely cheap compared to building your own dedicated servers. Heck I have a friend who hosts a minecraft server on Amazon's Cloud and I'm pretty sure he said it costs him nothing to a few bucks a month depending on that month's usage.  Now clearly he has like dozens of users and a AAA title would have a ton but you only pay for what you use and if your users all paid $60 to buy a game on launch day then if millions buy it you have plenty of money to pay Google or Amazon or Microsoft of whatever cloud provider you chose (and there are many) to provide service for you.  There's big competition in the could services business too so I'm sure EA or Ubisoft or other big players could get amazing deals and the situation is only going to get better as cloud computing continues to mature.  In some ways in fact Xbox game developers are limited in that they can ONLY choose Xbox Live Compute whereas those outside of Xbox Live can choose Microsoft Azure (on which Xbox Live Compute is based), Amazon, Google, etc... Again the big advantage of Xbox Live Compute is just that it's totally free so it's a trade off.  You get free cloud compute and but in return you're locked into Xbox only (including Win10).

In theory EA could release a game on say Amazon's cloud that worked with Mac, Windows, Linux, and Playstation versions of the game.  The only thing that's really expensive is setting up your own cloud but there's very little reason for most developers/publishers to do that when cloud computing is already a highly competitive field.  Maybe the super big guys like EA and Ubisoft can afford to set up their own server farms but most will just buy service on the existing cloud service providers.

I totally think there's a huge future in cloud compute in gaming, and I give massive kudos to MS for really being the first to try to push it for gaming but I really think this is going to be a general industry thing even if MS does it first and not just a Xbox/Win10 thing.  There also seems to be a lot of misunderstanding that it's somehow tied to Xbox or will make the hardware more powerful or something. (not from you specifically, just in general)

[Showan]

http://www.ign.com/videos/2015/08/06/17-minutes-of-explosive-crackdown-3-gameplay-gamescom-2015

 

 

[Showan]

What's expensive to do?  Cloud computing is cheap, that's one of the major selling points for it.  It's extremely cheap compared to building your own dedicated servers. Heck I have a friend who hosts a minecraft server on Amazon's Cloud and I'm pretty sure he said it costs him nothing to a few bucks a month depending on that month's usage.  Now clearly he has like dozens of users and a AAA title would have a ton but you only pay for what you use and if your users all paid $60 to buy a game on launch day then if millions buy it you have plenty of money to pay Google or Amazon or Microsoft of whatever cloud provider you chose (and there are many) to provide service for you.  There's big competition in the could services business too so I'm sure EA or Ubisoft or other big players could get amazing deals and the situation is only going to get better as cloud computing continues to mature.  In some ways in fact Xbox game developers are limited in that they can ONLY choose Xbox Live Compute whereas those outside of Xbox Live can choose Microsoft Azure (on which Xbox Live Compute is based), Amazon, Google, etc... Again the big advantage of Xbox Live Compute is just that it's totally free so it's a trade off.  You get free cloud compute and but in return you're locked into Xbox only (including Win10).

In theory EA could release a game on say Amazon's cloud that worked with Mac, Windows, Linux, and Playstation versions of the game.  The only thing that's really expensive is setting up your own cloud but there's very little reason for most developers/publishers to do that when cloud computing is already a highly competitive field.  Maybe the super big guys like EA and Ubisoft can afford to set up their own server farms but most will just buy service on the existing cloud service providers.

I totally think there's a huge future in cloud compute in gaming, and I give massive kudos to MS for really being the first to try to push it for gaming but I really think this is going to be a general industry thing even if MS does it first and not just a Xbox/Win10 thing.  There also seems to be a lot of misunderstanding that it's somehow tied to Xbox or will make the hardware more powerful or something. (not from you specifically, just in general)

 

 

Physics, calculations, to a dedicated server farmer in almost every corner of the Earth isn't cheap. That's why other major players don't bother.

Getting a server for a few people is fine, but multiply that by 10-20million people and things start to change. Now keep those servers running for say 24months and lets see what it would cost.

Yes EA, Activision, and such can afford to do this, but owning vs. renting are not the same thing.  EA is kind of on the in crowd w/Microsoft where they can get a deal cut very easily, instead of using their own servers

 

It may be cheap to get started, but it sure isn't cheap to keep going.  Time equal $$$... The more time the servers are up, them more $$$ it's gonna cost you.

 

Edit: I'd love to see more games try this.  

 

[+Asmodai MVC]

 

Physics, calculations, to a dedicated server farmer in almost every corner of the Earth isn't cheap. That's why other major players don't bother.

Getting a server for a few people is fine, but multiply that by 10-20million people and things start to change. Now keep those servers running for say 24months and lets see what it would cost.

Yes EA, Activision, and such can afford to do this, but owning vs. renting are not the same thing.  EA is kind of on the in crowd w/Microsoft where they can get a deal cut very easily, instead of using their own servers

 

It may be cheap to get started, but it sure isn't cheap to keep going.  Time equal $$$... The more time the servers are up, them more $$$ it's gonna cost you.

 

Edit: I'd love to see more games try this.  

 

I don't know about "every corner of the Earth" but in major 1st world countries like the U.S. (which is a big chunk of the gaming market by itself) cloud computing is cheap and there is heavy competition right now.  There is an ongoing major shift in government and corporate IT to the cloud.  I deal with it every day in a non-gaming way and that's why I'm so excited about MS expanding it to gaming.  There is HUGE potential there for the entire gaming industry.  Major players don't do it yet because it's still relatively new not because it's expensive.

Again if you have 10-20 million people then you have 10-20 million sales and that should be sufficient to secure cloud service hosting.  You only pay for what you use in many cloud implementations (often with low usage as free so when only a handful of users are using the cloud... such as during development the servers would be free) so if on launch month you get hit by 20 million users then sure you pay a lot (but all those users paid $60 each so you've got plenty of income to cover it). If 24 months later very few to no one is playing the game anymore then the cloud servers are free.  You don't have to pay for the massive hardware it took to cover the 20 million user launch, that's yet another of the major benefits of cloud computing, you only pay for what you use so as your active users shrink, so does your costs, it all scales.  One of the big show stoppers with dedicated servers was if you bought all the hardware to handle launch then most of it would set idle after the initial launch spike, cloud computing solves that.

I don't know if this particular game is going to launch the cloud computing craze, I hope it does.  If it doesn't though I'm confident that something else will and MS seems to be out in front of this so it will likely be an Xbox One game.  I'm confident though that this is the eventual future of online gaming in general though and not something that's Xbox specific.  It just doesn't make sense to do peer hosting or for companies to make dedicated servers when cloud computing is available now.  I really think this is the future.

[vcfan]

http://www.ign.com/videos/2015/08/06/17-minutes-of-explosive-crackdown-3-gameplay-gamescom-2015

that is AMAZING

[Ironman273]

Obviously it's not to the same fidelity but then that game is six years old and was released on the last-gen consoles. My point is that I'm sure a similar system could be implemented locally with minimal difference in fidelity. The limitation is still the GPU, as when the buildings break down the particle effects and increased polygon count still have to be rendered locally.

I don't see a modern PC having any trouble rendering the same scene. Here's a video from a few years back with all the physics done in real time:

https://www.youtube.com/watch?v=FIPu9_OGFgc

The maker of the video states: "Scenes were all setup quickly, simulated in real-time, and rendered overnight."  Doesn't that mean that it wasn't rendered in real time, just the physics were simulated?  That's a big difference...

[Hedon]

Red Faction Guerilla had a similar destruction system six years ago without the need for the cloud.

 

 

[red8Rain]

I don't know about "every corner of the Earth" but in major 1st world countries like the U.S. (which is a big chunk of the gaming market by itself) cloud computing is cheap and there is heavy competition right now.  There is an ongoing major shift in government and corporate IT to the cloud.  I deal with it every day in a non-gaming way and that's why I'm so excited about MS expanding it to gaming.  There is HUGE potential there for the entire gaming industry.  Major players don't do it yet because it's still relatively new not because it's expensive.

Again if you have 10-20 million people then you have 10-20 million sales and that should be sufficient to secure cloud service hosting.  You only pay for what you use in many cloud implementations (often with low usage as free so when only a handful of users are using the cloud... such as during development the servers would be free) so if on launch month you get hit by 20 million users then sure you pay a lot (but all those users paid $60 each so you've got plenty of income to cover it). If 24 months later very few to no one is playing the game anymore then the cloud servers are free.  You don't have to pay for the massive hardware it took to cover the 20 million user launch, that's yet another of the major benefits of cloud computing, you only pay for what you use so as your active users shrink, so does your costs, it all scales.  One of the big show stoppers with dedicated servers was if you bought all the hardware to handle launch then most of it would set idle after the initial launch spike, cloud computing solves that.

I don't know if this particular game is going to launch the cloud computing craze, I hope it does.  If it doesn't though I'm confident that something else will and MS seems to be out in front of this so it will likely be an Xbox One game.  I'm confident though that this is the eventual future of online gaming in general though and not something that's Xbox specific.  It just doesn't make sense to do peer hosting or for companies to make dedicated servers when cloud computing is available now.  I really think this is the future.

I have to chime in.  based on your thread, it seem you obviously have no knowledge of how expensive cloud hosting can be.  Go check out amazon and look at one of the m3.2xlarge.  Try powering that one for a month and see how quickly it would burn a hole thru your wallet.  I have one of those server for my past job and it burn ~5k a month. a m3.2xlarge is 8vCpu, 30gb of memory, and 2x80SSD.  Now imagine paying for 50 or 100k of those on a monthly basis?  You would go broke before you can make a profit. 

O and as to, turning it off while it isn't in used?  BS on that because you have to pay $$ to hold onto your SSD, so you will continue to incur monthly charge.  Wait, you don't need a m3.2xlarge to power anything and want a smaller instance?  I hope you have time to migrate all your data from one server to another because amazon doesn't let you simply move your crap from one server to another.  My devops spent half a day cloning, migrating, etc.. when we had to get off m3.2xlarge.

I like to see ANY independent developer(s) try to level the cloud on the same level as MS for this game.

[Emn1ty]

I have to chime in.  based on your thread, it seem you obviously have no knowledge of how expensive cloud hosting can be.  Go check out amazon and look at one of the m3.2xlarge.  Try powering that one for a month and see how quickly it would burn a hole thru your wallet.  I have one of those server for my past job and it burn ~5k a month. a m3.2xlarge is 8vCpu, 30gb of memory, and 2x80SSD.  Now imagine paying for 50 or 100k of those on a monthly basis?  You would go broke before you can make a profit. 

O and as to, turning it off while it isn't in used?  BS on that because you have to pay $$ to hold onto your SSD, so you will continue to incur monthly charge.  Wait, you don't need a m3.2xlarge to power anything and want a smaller instance?  I hope you have time to migrate all your data from one server to another because amazon doesn't let you simply move your crap from one server to another.  My devops spent half a day cloning, migrating, etc.. when we had to get off m3.2xlarge.

I like to see ANY independent developer(s) try to level the cloud on the same level as MS for this game.

If you have a proper server setup, you don't have to have a completely decentralized data storage system. You can pay for one primary data server, and the clusters will only store the data relevant to them (which is persisted to your central data server once it's no longer needed). So in that case, you can definitely just turn off the servers and not use them.

You think the whole of the cloud based companies wouldn't have solved this problem? It's called disk images, and rapid deployment. Look at things like Docker, Hub and Spoke architecture (http://www.compositesw.com/solutions/hub-and-spoke-architecture/). You can call BS all you want, but these are problems massive enterprise level corporations have already solved. Even startups have solved it, seeing as these are key factors in creating scaleable SaaS companies.

The principle is that for every second you have a server running it is driving revenue. Only spend when you can earn, never leave resources unused (money that's not earning).

[Showan]

If you have a proper server setup, you don't have to have a completely decentralized data storage system. You can pay for one primary data server, and the clusters will only store the data relevant to them (which is persisted to your central data server once it's no longer needed). So in that case, you can definitely just turn off the servers and not use them.

You think the whole of the cloud based companies wouldn't have solved this problem? It's called disk images, and rapid deployment. Look at things like Docker, Hub and Spoke architecture (http://www.compositesw.com/solutions/hub-and-spoke-architecture/). You can call BS all you want, but these are problems massive enterprise level corporations have already solved. Even startups have solved it, seeing as these are key factors in creating scaleable SaaS companies.

The principle is that for every second you have a server running it is driving revenue. Only spend when you can earn, never leave resources unused (money that's not earning).

Although virtual, aren't u still charged according to want/need?  Need 4 quad core processors 200GB memory, intense graphics/workstation card to handle calculations, etc, to cater to 1-million gamers, doesn't sound cheap.

I have no doubts that anyone can do this, but I do question how long smaller companies can keep it going.  If initial game is a hit then you may survive. But if your a smaller company and launch day goes wrong, It's game over before the games even started.

[Emn1ty]

Although virtual, aren't u still charged according to want/need?  Need 4 quad core processors 200GB memory, intense graphics/workstation card to handle calculations, etc, to cater to 1-million gamers, doesn't sound cheap.

I have no doubts that anyone can do this, but I do question how long smaller companies can keep it going.  If initial game is a hit then you may survive. But if your a smaller company and launch day goes wrong, It's game over before the games even started.

The point is that it scales with how much business you're doing, and when you're doing it. A flexible system will be running the most servers at peak hours and the least at dips in traffic. At any point during your day, week, month year, you're spending money relative to how much business you're getting. There can be limits imposed (budgets, so to speak) to keep traffic booms (or DDOS attacks) from making your dynamic system go wild but the whole point is for every additional server you're running you should be earning enough money to make up for it and then some. I mean, more traffic = more users = more revenue (be it via advertising, subscriptions or sales in regards to e-commerce).

To put it into perspective for a game studio, you'd budget out how much money you'd need and therefore your pricing model would be based on that best guess. The more players on your servers -> the more games you sold (or have subscriptions to the game) -> more money for running more servers (with some breathing room to make profit off of it as well).

Companies, and even developers, take these kinds of risks all the time. A successful business always has to go through these pains, that brink where they attempt to aggressively expand and see if they survive the night. Very few do proportionally to those that try and it has nothing to do with whether or not these technologies are feasible or doable and everything to do with, unfortunately, luck and/or persistence. They could be a hit, they could be a flop. Look at Lift, it was around before Uber and yet Uber is growing much faster and becoming far more popular. Things just don't work out sometimes.

Edited August 8, 2015 by Emn1ty

[President Devil]

http://www.ign.com/videos/2015/08/06/17-minutes-of-explosive-crackdown-3-gameplay-gamescom-2015

 

 

It's the only "next gen" thing I've seen about this generation of console gaming so far. Well done MS!

It's hilarious watching the Sony fanboys downplay this. Pretty sad.

[+Asmodai MVC]

I have to chime in.  based on your thread, it seem you obviously have no knowledge of how expensive cloud hosting can be.  Go check out amazon and look at one of the m3.2xlarge.  Try powering that one for a month and see how quickly it would burn a hole thru your wallet.  I have one of those server for my past job and it burn ~5k a month. a m3.2xlarge is 8vCpu, 30gb of memory, and 2x80SSD.  Now imagine paying for 50 or 100k of those on a monthly basis?  You would go broke before you can make a profit. 

O and as to, turning it off while it isn't in used?  BS on that because you have to pay $$ to hold onto your SSD, so you will continue to incur monthly charge.  Wait, you don't need a m3.2xlarge to power anything and want a smaller instance?  I hope you have time to migrate all your data from one server to another because amazon doesn't let you simply move your crap from one server to another.  My devops spent half a day cloning, migrating, etc.. when we had to get off m3.2xlarge.

I like to see ANY independent developer(s) try to level the cloud on the same level as MS for this game.

Sure cloud hosting CAN be expensive but that doesn't mean it always (or even usually) is.  You cherry pick an single example of how it's expensive... I'm sorry it was so expensive for you at your last job but the cloud is cheap for many despite your particular experience.  A game using cloud compute wouldn't need super fast or large quantities of storage (SSD's would likely be overkill).  The entire game isn't being streamed like PS Now or Win10/Xbox One game streaming and it's not an MMO so most of the data could still be local to the device.  The cloud compute would be doing things like running AI, physics, etc.  I'm not a game developer (just a corporate developer) so I don't know what exactly is non-latency sensitive and could be farmed out to the cloud and really I'm not sure many of them are.  This is a brand new area that MS is currently trailblazing and so I expect it will evolve quickly as people figure out what does and doesn't work.

It sounds to me like you had a bad experience with Amazon cloud and again I'm sorry for that but that's where the a lot of government and industry is moving and it's a fact that LOWER cost is one of the reasons despite there being some cases where you can pay more.  Maybe Amazon isn't the best provider RIGHT NOW.  RIGHT NOW no one has released a game that uses cloud computing that's done anything really amazing that's really got the ball rolling on the whole concept.  I'm hoping this game is that game but we'll have to see how it goes when it's actually released.  If it does then others will likely copy it and cloud providers will adapt their offerings to cater the that new market.  My post here was in part to express my excitement on MS leading that charge but if you disagree that's fine.  I used Amazon and Google and even Azure just because those are the big names in cloud computing right now (along with many other smaller ones) not because I was saying m2.2xlarge is what indy game developers should be using.

As for Amazon specifically they've been getting into gaming and buying up developers.  Since they are one of the bigger cloud providers I wouldn't be shocked if the next Amazon Fire TV or the one after that tried leveraging their cloud compute capabilities for games.  Again not one of the specific offerings they have RIGHT NOW but it would seem like a strong move for them to do to open up gaming options on their platform.

If Apple ever got serious about gaming like in the next Apple TV or the one after that then they likewise could do the same thing.  They may not have a public cloud compute capability like Amazon, Microsoft, and Google but they have loads of money and massive server farms so it wouldn't be a radical step for them to make an Apple cloud compute for gaming.  They have expressed a little more interest in gaming with their new graphics API but I wouldn't really hold my breath on them becoming gaming focused.

Google could make cloud compute work for Android based micro-consoles like the next version or one after of NVidia's Shield Android TV.  Again note I'm not saying the current version of these devices, because again we're still in the VERY EARLY days of this and I'm giving MS mad props for leading the charge here but I just don't see cloud computing as something that's going to stay MS specific, even in gaming.  Again, you are free to disagree.

Heck Steam may even provide cloud compute for Steam games as part of SteamWorks within a few years.  I really don't think these things are off the wall ideas on where the gaming industry may go.  None of that even includes all the other current corporate cloud providers that may branch out into gaming if it takes off and new companies that may form.  It's a really exciting prospect to at least to see where this all could go and again my intent was to give MS props for being on the bleeding edge of this.

 

[BajiRav]

I think this is cool.  Kudos for MS making Xbox Live Compute free for Xbox game developers.

Let's be clear here though, cloud compute is always going to be online only because you can't have cloud compute without the cloud.  Also cloud compute doesn't make the Xbox One more powerful, you need cloud compute to do things via network servers (i.e. the cloud) that the console is too WEAK to do locally (this isn't a Xbox vs. PS thing, PS is likely too weak to do it offline as well).  This is the pioneering days here but as this technology matures you might see comparatively weak devices like Apple TV Next, Amazon Fire TV Next, nVidia Shield Android TV Next, etc. leveraging cloud compute to play games that approach the quality of the stand alone consoles when the stand alone consoles aren't cloud connected.  Which brings us to the last point, while MS is certainly a Pioneer in this area there is nothing in the concept of Cloud compute that is Xbox exclusive.  Nothing is stopping a developer from using cloud compute in a Amazon Fire TV game and hosting it in Amazon's EC2 cloud.  Heck a developer could make a game that uses cloud compute to hosted on Microsoft's public Azure cloud and release the game on PlayStation if they want to.  The big deal MS is doing is just making that cloud free for Xbox games and providing some game specific APIs for developers.  Microsoft provided APIs may or may not be desirable to any given developer (many like to roll their own) and free is awesome but one of the big things with the cloud is it's cheap and it scales according to use.

I don't want to sound rude but your post is meaningless until the competition(i.e. Sony and Nintendo) comes up with something similar and until then Xbox One is absolutely more powerful in physics calculations than those consoles. I get it that anyone can do it but the point is Microsoft is the only company doing it. That is the difference.

Anyone could build an online gaming network but Microsoft was the first one to have it on every single box and easy to setup/use with their first Xbox. It took Sony almost two generations to catch up (and they still are behind in certain aspects) whereas Nintendo is still a joke in online services (don't own Wii or Wii U but based on what I read and saw at friends' places). Xbox Live was a pioneer and they are unarguably still a golden standard.

I will be very curious to see how Nintendo and Sony leverage cloud computing in their consoles (besides standard MP and media features) to improve traditional console games. I am not holding my breath though.

[+Audioboxer Subscriber²]

I don't want to sound rude but your post is meaningless until the competition(i.e. Sony and Nintendo) comes up with something similar and until then Xbox One is absolutely more powerful in physics calculations than those consoles. I get it that anyone can do it but the point is Microsoft is the only company doing it. That is the difference.

Anyone could build an online gaming network but Microsoft was the first one to have it on every single box and easy to setup/use with their first Xbox. It took Sony almost two generations to catch up (and they still are behind in certain aspects) whereas Nintendo is still a joke in online services (don't own Wii or Wii U but based on what I read and saw at friends' places). Xbox Live was a pioneer and they are unarguably still a golden standard.

I will be very curious to see how Nintendo and Sony leverage cloud computing in their consoles (besides standard MP and media features) to improve traditional console games. I am not holding my breath though.

Quite honestly I think everyone is waiting to see the results from MS first, as there is a truckload of scepticism, and not much more than proof of concepts right now. Otherwise, each of the other companies has their own invested interests that are taking center stage (Sony - Morpheus/PS Now, Nintendo - Trying to become relevant to the PS4/XB1 hardcore). 

[+Asmodai MVC]

I don't want to sound rude but your post is meaningless until the competition(i.e. Sony and Nintendo) comes up with something similar and until then Xbox One is absolutely more powerful in physics calculations than those consoles. I get it that anyone can do it but the point is Microsoft is the only company doing it. That is the difference.

Anyone could build an online gaming network but Microsoft was the first one to have it on every single box and easy to setup/use with their first Xbox. It took Sony almost two generations to catch up (and they still are behind in certain aspects) whereas Nintendo is still a joke in online services (don't own Wii or Wii U but based on what I read and saw at friends' places). Xbox Live was a pioneer and they are unarguably still a golden standard.

I will be very curious to see how Nintendo and Sony leverage cloud computing in their consoles (besides standard MP and media features) to improve traditional console games. I am not holding my breath though.

Well you completely failed at not sounding rude, congrats!  My post boiled down to a congrats to MS for PIONEERING cloud computing in gaming which I personally believe will be big (an OPINION you are free to disagree with of course by no more meaningless than anyone elses opinion... such as for example your thoughts on my post).

I don't know what part of PIONEERING you didn't get but that's constitutes an acknowledgment on my part that Microsoft is doing it first.  I'm not sure why you'd go to such great length to seemingly point out something I stated myself and then act as if you were somehow disproving my point.

As for the Sony and Nintendo thing I never even mentioned Nintendo and while I did mention PlayStation the first time was to specifically make clear I was NOT making a Xbox vs. PlayStation comparison, a point you seem to have entirely missed.  I totally agree Sony is unlikely to do anything cloud related this generation, they're already crushing this generation in sales without doing it so there really isn't much point for them to invest in it right away.  They can afford to ride out this generation and IF cloud computing takes off for gaming (which again I believe it will but that's opinion and it remains to be seen) then they can implement it for the PS5 if they so desire.  Unless of course you think cloud computing is going to make such a huge impact this generation that the Xbox One is going to catch up to and outsell the PS4 (overall, not just on select months or in select regions)... I don't think it will be THAT big of an impact (again, just my opinion though and you are free to disagree).  Furthermore PSN isn't controlled like Xbox Live and if cloud gaming does take off PlayStation developers can access the cloud on any internet accessible provider unlike Xbox Live where what servers games can access it regulated by MS.

Again the most immediate potential competition I see to Microsoft's cloud compute in gaming is NOT from Sony or Nintendo but from the other competitors that are starting to branch into gaming, have their own ecosystem, and/or are already in the non-gaming cloud computing market.  Competitors such as Amazon, Google, Apple, and Valve.  Even then the examples I posted were for future devices not existing ones, once again acknowledging MS was PIONEERING this market:  Apple TV NEXT, Amazon Fire TV NEXT, nVidia Shield Android TV NEXT, etc.

Microsoft is ABSOLUTELY out in front on this, a point I fully acknowledged in my post.  I'm not sure why you feel the need to turn this into a MS vs. Sony vs. Nintendo thing but that's all you and has nothing to do with my post that you for some reason quoted despite seemingly entirely missing the point.