How would you implement multi GPU rendering on a closed box that is already defined as having a single GPU? The difference is the computational power available to developers, to run things as physics simulations or AI. On a PS5 or Xbox Two, they have a single CPU/GPU on each box and they cannot make it work with another PS5/XBox Two CPU/GPU, that is in a separate house. In a server environment they can make the connected hardware cooperate with each other. You can of course create hybrid solutions like the one below, but it would be a less efficient solution than the one that Stadia is implementing.
Fight sim is like the only one so farThe possibilities of cloud gaming has been discussed for like 10+ years but has never achieved what's promised.
Their stance on allowing you to game with out purchasing hardware must be infuriating...
Cool stuff will come from it at some point; just not from Google.I think stadia will offer some reall amazing stuff that we havent seen before. Its a shame lots are so down on it here. I understand peoples fear that they dont "own" the games and googles past history. But some really cool stuff can come from this.
We will see, MS was just as hated when they entered the gaming scene and yet we can thank them for giving Sony to bring up their online infrastructure, unified gamertags, achievements, ect..Cool stuff will come from it at some point; just not from Google.
The issue isn't Google entering the gaming scene, the issue is Google entering the gaming scene with its track record of lack of support, constant fracturing, and killing of of popular services, along with half assing of features that were supposed to be great elsewhere.We will see, MS was just as hated when they entered the gaming scene and yet we can thank them for giving Sony to bring up their online infrastructure, unified gamertags, achievements, ect..
Some times you need an outsider to bring something new to the mix.
Time will tell, either way some interesting times ahead
What you keep discussing is literally what Crackdown 3 does. It has physics and AI simulations happening on cloud servers, main rendering being done on people's local Xbox's.How would you implement multi GPU rendering on a closed box that is already defined as having a single GPU? The difference is the computational power available to developers, to run things as physics simulations or AI. On a PS5 or Xbox Two, they have a single CPU/GPU on each box and they cannot make it work with another PS5/XBox Two CPU/GPU, that is in a separate house. In a server environment they can make the connected hardware cooperate with each other.
Well there response to internet connections were that the internet providers will solve that problem.Still crazy to me that they didn't start developing games to take advantage of this stuff and sell their product years ago. Do they not know how long it takes to make games these days?
Not what I'm referring to. Crackdown 3 was supposed to have impossible-for-local-hardware-to-calculate levels and environmental destruction, showcased in 2015 with the power of the cloud meme and all, and in the end all of that massively scaled down. Digital Foundry did an analysis:Admittedly crackdown 3 was a game struggling to run on console hardware then struggling to sync with a server, which would have to spin up for each player
In this case the game is already running on a much higher powered spun-up server and all those calculations are ready to go
Not what I'm referring to. Crackdown 3 was supposed to have impossible-for-local-hardware-to-calculate levels and environmental destruction, showcased in 2015 with the power of the cloud meme and all, and in the end all of that massively scaled down. Digital Foundry did an analysis:
What's even worse is some people literally won't be able to do anything launch day because the preorder bundles are being shipped over the course of like 2 weeks.
What's even worse is some people literally won't be able to do anything launch day because the preorder bundles are being shipped over the course of like 2 weeks.
The same problems that happened with Crackdown 3 will not be there with Stadia, unless you have anything specific you want to point at, on how the Stadia server blades will not be able to work together.
From the article you posted:
"It's difficult to avoid the conclusion that the original Crackdown 2015 concept we saw was massively over-engineered in some respects - to the point where actually rolling it out to thousands of players could ever be achieved, bearing in mind the vast amounts of compute power it required."
I can see the massive benefits cloud gaming could deliver, but for me the biggest mistake they made is not have a launch game that demonstrates one of those use cases.
Their stance on allowing you to game with out purchasing hardware must be infuriating...
It's funny cause if I did a point by point list like this in any console or PC thread as it has fuck all to do with the topic, pretty sure I would get banned.If i had some time i would add these to the last cell:
- and there's input lag.
- and compression artifacts.
- and buffering whenever the network has a hiccup.
- and performance drops when someone else in your house downloads or streams something.
- and you can't mod games.
- and...
The elephant in the room, who is paying for this server hardware?People still thinks Google will give a free server cluster time without asking bigger cut from devs? Nothing is free, this amazing potential features will cost money.
If i had some time i would add these to the last cell:
- and there's input lag.
- and compression artifacts.
- and buffering whenever the network has a hiccup.
- and performance drops when someone else in your house downloads or streams something.
- and you can't mod games.
- and...
The elephant in the room, who is paying for this server hardware?
People still thinks Google will give a free server cluster time without asking bigger cut from devs? Nothing is free, this amazing potential features will cost money.
Even though I dunked on this messaging it is only because we have been through this song and dance before with Microsoft. The potential of having multiple computers work together is super obvious. Most of us do care but are now old enough to care about merely words.
I'll believe it when I see it. So far everything Stadia is made of vapor and dreams, and not any tangible products. Until then, there is no reason to put my faith in Google.The same problems that happened with Crackdown 3 will not be there with Stadia, unless you have anything specific you want to point at, on how the Stadia server blades will not be able to work together.
From the article you posted:
"It's difficult to avoid the conclusion that the original Crackdown 2015 concept we saw was massively over-engineered in some respects - to the point where actually rolling it out to thousands of players could ever be achieved, bearing in mind the vast amounts of compute power it required."
Unless we have comments from developers or even Google saying something different, this is something we can all expect from Stadia. What you see below is what I have seen from developers.
"I think that the more interesting question is how stuff like Google Stadia will change things. It gives developers something different. In the data center, these machines are connected to each other, and so you could start thinking of doing things like elastic rendering, like make a couple of servers together to do physics simulations that may not be possible on current local hardware. I think you'll see a lot of evolution in this direction."
"When you have an almost uncapped amount of computation sitting in a data centre that you can use to support your game design and ambition – whether it's in vastly superior multiplayer, whether it's in distributed physics, or massive simulation – there are things we can do inside a data center that you could never do inside a discreet, standalone device."
The logic you are trying to use now, you wouldn't accept it in any other situation. Let's say that you are trying to sell a PS4 to someone and the reason someone gives you to not buy it, is that he has seen broken PS4s. Right there you would bring objetive data like the one below to show him how reliable PS4s are. What you are doing is even worse, it's like someone saying to you he won't buy your PS4, because he has seen broken Xbox Ones. So let's try again and bring some objetive information we can all read on how developers that make games for Stadia won't be able to make the server blades/instances work together.
"I think that the more interesting question is how stuff like Google Stadia will change things. It gives developers something different. In the data center, these machines are connected to each other, and so you could start thinking of doing things like elastic rendering, like make a couple of servers together to do physics simulations that may not be possible on current local hardware. I think you'll see a lot of evolution in this direction."
"When you have an almost uncapped amount of computation sitting in a data centre that you can use to support your game design and ambition – whether it's in vastly superior multiplayer, whether it's in distributed physics, or massive simulation – there are things we can do inside a data center that you could never do inside a discreet, standalone device."
Stadia games can only be brought in the Google store correct?
That level of control means there will be less frequent sales and probably also means less steeper sales than other platforms.
On top of that Google is telling premium users pay us 10 usd a month and you don't have to worry about buying new hardware every cycle. Over 6 years based on PS4 MSRP you're paying the same 720 usd as a PS4 owner at launch who also pays for online multiplayer every year.
The weird part is that in 2020 Google will have a free subscription tier. I'm sure ad revenue will be factor but I can't imagine ad conversions being so profitable in aggregate. It will take another year or year and half to figure that out once Google releases their investor reports.
Devs can say a lot of things about the potential of hardware. It's happened before for AMD, Nvidia as well as Microsoft. They still are worried about their bottom line in terms of time management while their leaders are concerned about the money side of things. When a developer makes a game most of them are going to be worried about feature parity across platforms. The rest who don't give a shit will try to make something on Stadia but their numbers are really low. If they don't have proper financial backing they'll settle for VR development instead which has more cachet and nerd cred than game streaming.
The PS4 came out at $400, the PS4 Pro came out 3 years later at $500. 6 Years of PSN is $360. Try doing the math again taking into account that online play is included, that the server hardware will continue to improve over time and that you don't have to pay $10 every month. 6 years of the basic subscription is $0.00 on Stadia, you can take all the money you used to spent on the console, online play and mid gen upgrades and invest all of that on games or take that money and spent it elsewhere if you want. All of this has no value of course if the service doesn't work for you, but if it does, it is a loosing battle on your side to try to deny how much money you can save.
You didn't realize I was responding to a post that pondered how Google could possibly get money off of this. It is sufficient enough to show that the long term costs of buying Stadia are comparable to a current gen console at launch.
If it was a discussion about trying to prove which is more cost effective I would be spending more time on the analysis.
I'll believe it when I see it. So far everything Stadia is made of vapor and dreams, and not any tangible products. Until then, there is no reason to put my faith in Google.
The hybrid solution you linked to doesn't really attempt to solve distributed rendering, it's just a two machine scenario with one doing basically all of the work.You can of course create hybrid solutions like the one below, but it would be a less efficient solution than the one that Stadia is implementing.
Seriously, a special build of EVE Online with hundreds of players and no lagging would be epic for demonstration purposes (if such a thing existed)You'd think they would at least have a (misleading) tech demo or two to back up their claims.