By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Microsoft - Cloud Processing won't mean a thing for the Xbone

 

The Cloud will...

Significantly improve games 63 12.57%
 
Improve some games somewhat 65 12.97%
 
Might be used well here and there 101 20.16%
 
Merely PR BS 271 54.09%
 
Total:500
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.



Around the Network
SvennoJ said:
scottie said:
ViktorBKK said:

There is a lot of buzz lately about the how the cloud will change everything, not only in gaming but in general.

What many seem to ignore is that internet infrastructure does not improve at the same rate as silicon. The cost of GPU power for example, has massively improved in the last 5 years. The cost of bandwidth on the other hand, has improved only slightly, and in some areas of the world, not at all.

Ok, from the sounds of that post, you don't really know what cloud processing is. Internet speeds don't need to progress at the same rate as processing speeds, because the amount of processing required in video games increases orders of magnitude faster than the amount of data a streaming service needs to send.

 

The advantage of cloud processing is that you do the work remotely, and then just pass the results to the console. The size of the results needed to be sent will either be constant, or slightly increasing.

 

Take as an example, if we wanted a complete cloud system, much like Ouya, where all the processing is done elsewhere and what is sent back is just a video.

 

If we had done this in 1985 to replace the NES, we would have had to send 256*240 pixels, and must change 30 times per second. This means we need to send 1.8 M pixels per second. The NES uses 2 digit hexidecimal for colours, so the total amount of data sent is 1.8M * 8, as a hexidecimal digit is 4 bits. The NES would require 14.4 Mbit/s

I can't find details for the Famicom modem, but a 3rd party alternative worked at 2.4 kbit/second. Internet speeds would need to be 7000 times faster to display it this way.


Nowadays, for the X1, we can get at best 3840*2160 pixels at 60 fps, which is 500 M pixels/second. Each pixel requires 24 bits, requiring 1.2 GBit/s. This is actually attainable for the X1 in locations with Google fibre. We can also drop the framerate to 30/second and resolution down from 4k to 1080p to massively reduce the load. I am not suggesting that this will be how the X1 streaming works, just demonstrating that the internet is developing faster than the load required to stream.

 

We can think of partial streaming solutions that the amount of data sent has not even increased since computers were invented. Lets say we had a chess game. The graphics and the users turn are handled by the Xbone, and the A.I. is handled in the cloud. This allows for use of super computers, creating an A.I. good enough to beat grand masters. What is the actual data sent? The location of the piece to be moved, and the location is is to move to. All that needs to be sent is A7A6 to move a pawn forward. This is the same amount of data that would have been sent with a chess cloud streamign service in 1950, and it'll be the same in 2050.

Slight miscalculation there 4K video at 60fps 24 bit is 11.1 Gbit/s. But with h.265 it can be compressed down to 30 mbps for reasonable streaming quality. Besides the point anyway. 

Yes chess is a good example, maybe also to speed up an end game of civilization. And to keep track of NPCs and objects over a large world and some dynamic changes in that world. (No too dynamic though since streaming world data cannot be lossy compressed and easily costs a lot more then a simple video stream)

What can it do for real-time gaming? Off loading background AI is pretty much the only thing that leaves some resources to speed other things up a bit. Yet there still needs to be a fall back in place in case the connection is lost. And does it really add to the game if every car and every person in a GTA game is persistant and animated by the server?

I'm enjoying your posts. Gives credibility to the thread - thanks.



Dr.Grass said:
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.


Yes dude they speculate.  The author is neither a developer or engineer.  They speculate because they cannot develop code themselves and they do not have a way to test what is possible or not.  Just like any technology you put it in the hands of the builder and let them find creative ways to use it.  It's interesting people put up articles like they are gospel from someone who has no more technical knowledge on the subject then yourself.  Personally I would listen to actual developers who are using such techniques.  I want to here from developers like avalanche, Bethesda, Rockstar and others who may be able to leverage this tech and here there opinion.  I really care not about an opinion peace that conclude that MS must show me it works because we all came to that conclusion without knowing anything.

Sony talked about how powerful the cell was back in the day but when the first games games came out in relation to the 360 people were hard pressed to see the difference and the 360 games even looked better.  As the years went on developers figured out how to send there code as jobs to the cell processor.  The developers figured how to tap into the power of the cell by changing the way they develop for the system.  This will not be any different with the cloud compute.  Developers will test what works and find solutions that allow them to take advantage of the power.

Poeple who naysay the technology and state it will not do anything to enhance a game are the same people who stated the same thing about the cell but look at some of the games coming out now.  



scottie said:
ViktorBKK said:

There is a lot of buzz lately about the how the cloud will change everything, not only in gaming but in general.

What many seem to ignore is that internet infrastructure does not improve at the same rate as silicon. The cost of GPU power for example, has massively improved in the last 5 years. The cost of bandwidth on the other hand, has improved only slightly, and in some areas of the world, not at all.

Ok, from the sounds of that post, you don't really know what cloud processing is. Internet speeds don't need to progress at the same rate as processing speeds, because the amount of processing required in video games increases orders of magnitude faster than the amount of data a streaming service needs to send.

 

The advantage of cloud processing is that you do the work remotely, and then just pass the results to the console. The size of the results needed to be sent will either be constant, or slightly increasing.

 

Take as an example, if we wanted a complete cloud system, much like Ouya, where all the processing is done elsewhere and what is sent back is just a video.

 

If we had done this in 1985 to replace the NES, we would have had to send 256*240 pixels, and must change 30 times per second. This means we need to send 1.8 M pixels per second. The NES uses 2 digit hexidecimal for colours, so the total amount of data sent is 1.8M * 8, as a hexidecimal digit is 4 bits. The NES would require 14.4 Mbit/s

I can't find details for the Famicom modem, but a 3rd party alternative worked at 2.4 kbit/second. Internet speeds would need to be 7000 times faster to display it this way.


Nowadays, for the X1, we can get at best 3840*2160 pixels at 60 fps, which is 500 M pixels/second. Each pixel requires 24 bits, requiring 1.2 GBit/s. This is actually attainable for the X1 in locations with Google fibre. We can also drop the framerate to 30/second and resolution down from 4k to 1080p to massively reduce the load. I am not suggesting that this will be how the X1 streaming works, just demonstrating that the internet is developing faster than the load required to stream.

 

We can think of partial streaming solutions that the amount of data sent has not even increased since computers were invented. Lets say we had a chess game. The graphics and the users turn are handled by the Xbone, and the A.I. is handled in the cloud. This allows for use of super computers, creating an A.I. good enough to beat grand masters. What is the actual data sent? The location of the piece to be moved, and the location is is to move to. All that needs to be sent is A7A6 to move a pawn forward. This is the same amount of data that would have been sent with a chess cloud streamign service in 1950, and it'll be the same in 2050.

 

I believe I understand what cloud processing is fairly well. You are correct that compressing a signal can reduce bandwith requirements. That doesn't change the fact that bandwidth usage costs a lot more money than silicon. Even if your data is proccessed in the cloud, it still has to be processed, you the customer still have to pay for renting that proccessing power, but you, as well as your provider(Microsoft) also have to pay for the increased bandwidth usage. It is a lot more cost effective for the consumer, to make a one time purchase and use local hardware.

And btw, last time I checked Google Fibre isn't widely available. Most of the developed world is probably still in the 10-50 Mbps range. Could you stream 1080p with 50 Mbps? Yes you could. But it would cost a hell of a lot less if you just spent 100 dollars more on your graphics card. There is a reason why so many ISPs have caps on bandwidth usage.



Machiavellian said:
Dr.Grass said:
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.


Yes dude they speculate.  The author is neither a developer or engineer.  They speculate because they cannot develop code themselves and they do not have a way to test what is possible or not.  Just like any technology you put it in the hands of the builder and let them find creative ways to use it.  It's interesting people put up articles like they are gospel from someone who has no more technical knowledge on the subject then yourself.  Personally I would listen to actual developers who are using such techniques.  I want to here from developers like avalanche, Bethesda, Rockstar and others who may be able to leverage this tech and here there opinion.  I really care not about an opinion peace that conclude that MS must show me it works because we all came to that conclusion without knowing anything.


I suppose I should stop listening to theoretical physicists since they don't know how to set up experiments themselves.

O_o

You don't need to be a game dev to know EXACTLY how cloud-computing works.

Half my work is writing software. I've played games all my life. I live with 2 game devs. I CAN FIGURE IT OUT. And so can smart people like those at Digital Foundry and EG. It's not rocket science.



Around the Network
Machiavellian said:
Dr.Grass said:
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.


Yes dude they speculate.  The author is neither a developer or engineer.  They speculate because they cannot develop code themselves and they do not have a way to test what is possible or not.  Just like any technology you put it in the hands of the builder and let them find creative ways to use it.  It's interesting people put up articles like they are gospel from someone who has no more technical knowledge on the subject then yourself.  Personally I would listen to actual developers who are using such techniques.  I want to here from developers like avalanche, Bethesda, Rockstar and others who may be able to leverage this tech and here there opinion.  I really care not about an opinion peace that conclude that MS must show me it works because we all came to that conclusion without knowing anything.

Sony talked about how powerful the cell was back in the day but when the first games games came out in relation to the 360 people were hard pressed to see the difference and the 360 games even looked better.  As the years went on developers figured out how to send there code as jobs to the cell processor.  The developers figured how to tap into the power of the cell by changing the way they develop for the system.  This will not be any different with the cloud compute.  Developers will test what works and find solutions that allow them to take advantage of the power.

Poeple who naysay the technology and state it will not do anything to enhance a game are the same people who stated the same thing about the cell but look at some of the games coming out now.  


I'm sorry. But you are neither a developer nor an engineer. You cannot develop code yourself and you do not have a way to test if it's possible or not. So your words are nothing but speculations.


Am I doing it right? =P



Dr.Grass said:
Machiavellian said:
Dr.Grass said:
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.


Yes dude they speculate.  The author is neither a developer or engineer.  They speculate because they cannot develop code themselves and they do not have a way to test what is possible or not.  Just like any technology you put it in the hands of the builder and let them find creative ways to use it.  It's interesting people put up articles like they are gospel from someone who has no more technical knowledge on the subject then yourself.  Personally I would listen to actual developers who are using such techniques.  I want to here from developers like avalanche, Bethesda, Rockstar and others who may be able to leverage this tech and here there opinion.  I really care not about an opinion peace that conclude that MS must show me it works because we all came to that conclusion without knowing anything.


I suppose I should stop listening to theoretical physicists since they don't know how to set up experiments themselves.

O_o

You don't need to be a game dev to know EXACTLY how cloud-computing works.

Half my work is writing software. I've played games all my life. I live with 2 game devs. I CAN FIGURE IT OUT. And so can smart people like those at Digital Foundry and EG. It's not rocket science.

Yeah maybe you should stop reading and actually DO it.  Reading about something does not give you practical experience.  I work as a developer on business critical code that are hosted on cloud platfoms.  We run entire business within the cloud.  I know how the technology works very well, what I do not know is how well it will work within games.  There are many different scenerios I can think of that may benefit a game but since I do not develop in those areas they are just ideals.  Most developers know the scene and what the player will be doing in any given moment of their game.  A lot of pre calucations can be sent, processed and stored in memory, thus releaving the CPU/GPU from those task and allow them to concentrate on the immediate input from the gamer.  

You can tell me you have friends who are developers but that means absolutly nothing.  There are many different developers that work on different parts of a game.  Even then, if they have not explored cloud compute, they would be no different from a graphics programmer trying to tell a network programmer how to code against the net.  That whole sentence is a very arrogant statement as if being a programmer immediatly gives you insight into everything without the practical experience.

I have to call this statement out because its one of the big problems with people on the net who THINK they know something without any practical knowledge.

You don't need to be a game dev to know EXACTLY how cloud-computing works.

Have you ever programmed to a platform like Azure or Amazon cloud infrastructure.  Do you even know what it takes to host an entire business on the cloud.  Have you ever coded a game or even an MMO.  Do not tell me about what you think unless you actually have practical knowledge on how its done.  I am not going to sit here and tell you how it will be done just because I develop code.  Without the practical experience, I can only make quesses and even then I would not know if I was on target or not without actually doing it.

There were many developers asked about the whole cloud compute thing and their answers was that they needed to test it.



Hynad said:
Machiavellian said:
Dr.Grass said:
Machiavellian said:

AThe eurogamer has no more clue about what can be archived on cloud compute then the average joe on the street.  

 


(gonna have to to this again)

Dude. Really.


Yes dude they speculate.  The author is neither a developer or engineer.  They speculate because they cannot develop code themselves and they do not have a way to test what is possible or not.  Just like any technology you put it in the hands of the builder and let them find creative ways to use it.  It's interesting people put up articles like they are gospel from someone who has no more technical knowledge on the subject then yourself.  Personally I would listen to actual developers who are using such techniques.  I want to here from developers like avalanche, Bethesda, Rockstar and others who may be able to leverage this tech and here there opinion.  I really care not about an opinion peace that conclude that MS must show me it works because we all came to that conclusion without knowing anything.

Sony talked about how powerful the cell was back in the day but when the first games games came out in relation to the 360 people were hard pressed to see the difference and the 360 games even looked better.  As the years went on developers figured out how to send there code as jobs to the cell processor.  The developers figured how to tap into the power of the cell by changing the way they develop for the system.  This will not be any different with the cloud compute.  Developers will test what works and find solutions that allow them to take advantage of the power.

Poeple who naysay the technology and state it will not do anything to enhance a game are the same people who stated the same thing about the cell but look at some of the games coming out now.  


I'm sorry. But you are neither a developer nor an engineer. You cannot develop code yourself and you do not have a way to test if it's possible or not. So your words are nothing but speculations.


Am I doing it right? =P

Well, I am a developer but not for games.  Yes, I can only speculate based on my experience working on cloud platforms but not hosted game code which is why I stated I would rather hear from people that DO game code for a living and who would have actual practical experience.  I know, its easy to just grab a part of someone post instead of reading the whole thing which is what people are good at on the net but please read the whole thing.

I make no mention that I know what can and cannot be done.  Instead, I am stating we need to wait for people with that knowledge, who have actually tried to use or are using cloud compute.  I hear the naysayers but I am wondering what they are basing their experience on.  Without experience, its just ramdon noise



Machiavellian said:
Hynad said:


I'm sorry. But you are neither a developer nor an engineer. You cannot develop code yourself and you do not have a way to test if it's possible or not. So your words are nothing but speculations.


Am I doing it right? =P

Well, I am a developer but not for games.  Yes, I can only speculate based on my experience working on cloud platforms but not hosted game code which is why I stated I would rather hear from people that DO game code for a living and who would have actual practical experience.  I know, its easy to just grab a part of someone post instead of reading the whole thing which is what people are good at on the net but please read the whole thing.

I make no mention that I know what can and cannot be done.  Instead, I am stating we need to wait for people with that knowledge, who have actually tried to use or are using cloud compute.  I hear the naysayers but I am wondering what they are basing their experience on.  Without experience, its just ramdon noise

Don't bother with this guy...

He reported me for saying the same exact thing (I'm a dev. and I'm using Azure for cloud processing). I said that it is not because he is reading something on the internet that this is true and he reporting me for that (and I got a warning for "implying he was stupid" just because I said he was just an internet reader and not a dev.)... And then he played the "prove it" card.

Just look at a lot of replies here. People are arguing that we need a lot of bandwith etc. to use the cloud processing... The only thing we really need is an internet connection (obvious); then processing stuff in the cloud does not mean rendering it in the cloud and streaming it; like many asynchroneous tasks in a single player game, we can process it somewhere and get the result later (so the lag is not really a problem, no need to have very high speed internet for that).

Everytime you are giving your opinion on the possibilities and potential and why it is doable in today's internet environment, they keep replying with "did you read the EuroGamer article..." like it was the absolute truth :) And when you try to explain them that they are also speculating; they are asking you to "prove" everything... But they believe everything in the article... go figure...

~

Edit: Quote tree shortened by TruckOSaurus - Please refer to this post for Quote Tree Guidelines http://gamrconnect.vgchartz.com/post.php?id=5379187



Imaginedvl said:
Machiavellian said:

Hynad said:

I'm sorry. But you are neither a developer nor an engineer. You cannot develop code yourself and you do not have a way to test if it's possible or not. So your words are nothing but speculations.


Am I doing it right? =P

Well, I am a developer but not for games.  Yes, I can only speculate based on my experience working on cloud platforms but not hosted game code which is why I stated I would rather hear from people that DO game code for a living and who would have actual practical experience.  I know, its easy to just grab a part of someone post instead of reading the whole thing which is what people are good at on the net but please read the whole thing.

I make no mention that I know what can and cannot be done.  Instead, I am stating we need to wait for people with that knowledge, who have actually tried to use or are using cloud compute.  I hear the naysayers but I am wondering what they are basing their experience on.  Without experience, its just ramdon noise

Don't bother with this guy...

He reported me for saying the same exact thing (I'm a dev. and I'm using Azure for cloud processing). I said that it is not because he is reading something on the internet that this is true and he reporting me for that (and I got a warning for "implying he was stupid" just because I said he was just an internet reader and not a dev.)... And then he played the "prove it" card.

Just look at a lot of replies here. People are arguing that we need a lot of bandwith etc. to use the cloud processing... The only thing we really need is an internet connection (obvious); then processing stuff in the cloud does not mean rendering it in the cloud and streaming it; like many asynchroneous tasks in a single player game, we can process it somewhere and get the result later (so the lag is not really a problem, no need to have very high speed internet for that).

Everytime you are giving your opinion on the possibilities and potential and why it is doable in today's internet environment, they keep replying with "did you read the EuroGamer article..." like it was the absolute truth :) And when you try to explain them that they are also speculating; they are asking you to "prove" everything... But they believe everything in the article... go figure...

LOL! Keep em coming!

You clearly haven't read the article. Because the article doesn't suggest there are no ways to use cloud processing. They're taking the wait and see approach, while elaborating on the reason why they think it's not going to have as big an impact as MS claims.  It's really funny how you argue this touted feature of the XBO.  People say that it won't be a "game changer" feature. Most agree that the use won't have as big an impact as MS claimed it would (4 times the power of the XBO has without it)... And so far, you've saying the exact same thing everyone is saying... And there you here, still arguing for nothing. 

But go on. Keep on arguing with yourself. xD