Yeah I think stadia is going to stick with console performance,it will be just as much 60FPS as consoles are,resource heavy games will be at 30FPS.So unless you think the Xbox one X or PS4 are potatoes, Stadia should be fine (especially considering that the tests and impressions so far have been at 30 FPS not the final 60 FPS).
Yeah I think stadia is going to stick with console performance,it will be just as much 60FPS as consoles are,resource heavy games will be at 30FPS.
It just doesn't make sense for them to spend so much hardware on getting 60FPS for games that are planned out for 30FPS.
If it's all virtual they can make each instance as powerful as they want,what I said was that it doesn't make sense for them to use too much hardware on each instance because cloud cpu cores are a sought after commodity that they can use and rent for other things that make them more money.Stadia basically has twice the performance of the Xbox one X or PS4 Pro, so maintaining 60 FPS shouldn't be much of an issue, assuming that most of the games will primarily be console games and similar multiplatform games.
It's also worth noting that Google has said that resource heavy games will be able to use multiply Stadia instances, in other words multi-GPU setups.
If it's all virtual they can make each instance as powerful as they want,what I said was that it doesn't make sense for them to use too much hardware on each instance because cloud cpu cores are a sought after commodity that they can use and rent for other things that make them more money.
It's also worth noting that Google has claimed that anyone using Stadia will never be more than 64km from the datacenter running the game, so if you are getting excessive network latency it's probably your ISP's fault not Stadias.
So if this is true, only a small portion of the US population will be able to use this anyway. It would mean Google would need a data center every 80 miles, which they don't have.
The vast majority of the US population lives in metropolitan areas, so if Google could simply cover those (or at least the majority of them), then they wouldn't need anywhere near a data center every 80 miles.
This would of course leave people living in rural areas high and dry, but those areas tend to have crappy internet connections, so chances are that they wouldn't qualify for Stadia anyway.
Well, for instance, I am in California. I am more than 64km from the closest Google datacenter. But I can have gigabit fiber to my house if I chose. California has 36 million people in it, and while a lot of them are in certain areas that are close, about half the population would not be in range as California is a very large state geographically speaking. The same would go for a state like Texas.
Agreed with the distance irrelevance. Speed of light says 1ms every 186 miles or 300Km.That's cool and all, but it's not really relevant. I'm sure there are plenty of people like you who has adequate connections, but are further away than 64 km from the nearest data center. Btw, how do you know that you are not within 64 km? Google has over 7500 data center nodes, but I don't think there's a publicly available list of all of their locations.
However my point was that the number of people within 64 km (i.e. the ones Google are targeting ) is large enough to support a service like Stadia. Sure down the line they might want to expand their reach, but I don't really think they feel any immediate need to do so.
Also it's not like distance really matters, even if you went from coast to coast in the US you're still only looking at about 20ms of added latency. Latency generally doesnt come from distance but from data processing at the nodes.
Yeah I think stadia is going to stick with console performance,it will be just as much 60FPS as consoles are,resource heavy games will be at 30FPS.
It just doesn't make sense for them to spend so much hardware on getting 60FPS for games that are planned out for 30FPS.
Why not just buy a console then? They aren't that expensive and work OK with our garbage internet connections already.
People make arguments for this streaming all the time but all the half way decent arguments seem to boil down to "like a console, but worse".
That's cool and all, but it's not really relevant. I'm sure there are plenty of people like you who has adequate connections, but are further away than 64 km from the nearest data center. Btw, how do you know that you are not within 64 km? Google has over 7500 data center nodes, but I don't think there's a publicly available list of all of their locations.
However my point was that the number of people within 64 km (i.e. the ones Google are targeting ) is large enough to support a service like Stadia. Sure down the line they might want to expand their reach, but I don't really think they feel any immediate need to do so.
Also it's not like distance really matters, even if you went from coast to coast in the US you're still only looking at about 20ms of added latency. Latency generally doesnt come from distance but from data processing at the nodes.
Google lists all the locations here: https://peering.google.com/#/infrastructure
Also, it would see the 64kM range was for ID's testing, so other games may vary.
EDIT: Also note that 64kM is telefeet, not 'as a bird flies' distance.
That's a really cool site, didn't know about that one, but it clearly doesn't include all 7,500 locations, the map has 1,000 locations at best.
Regardless though, as I already mentioned above the distance doesn't really matter as far as latency is concerned.
The distance itself really doesn't. Speed of light and all that. However, typically the farther you go, the more routing equipment you pass through. This is where delay is induced, and is thereby why distance is typically a factor. Its the reason I ping 100ms to the east coast, but ping 15ms to the bay area. That extra 75ms isn't from the signal taking that long to travel across the country, its the equipment causing delay as the signal makes its way east.
Agreed with the distance irrelevance. Speed of light says 1ms every 186 miles or 300Km.
And as antihelten said distance is relevant is it will affect number of hops that will occur.
That is why I asked about the solar high altitude internet drones. Up and down. and unlike geosynchronous, still close.But you would have to double the latency. You click then it has to go to the data center and then the result comes back.
And as antihelten said distance is relevant is it will affect number of hops that will occur.
looks like someone managed to get some actual latency numbers for DOOM:
Not sure how this is measured or how accurate it is, but one interesting thing to note is that DOOM has roughly 25ms lower latency than AC: Odyssey. Assuming 166 ms total latency for AC:O (as measured by Digital Foundry), that would put DOOM at about 140 ms, which is indeed a tad high for a shooter (and would explain some of the floatiness reported earlier), but if we can assume they will manage to shave of 33ms by moving from 30 FPS to 60 FPS, they would get down to about 105ms, which is very close to what consoles are currently running DOOM at (about 90 ms), and should be perfectly playable.
That's weird that the input lag would be that high from a booth at a convention. Where was the server in relation to gdc 2019?
I realize that pings are not exactly the same thing as streaming in encoded game frame data, decoding the data, and displaying it, but I get 6 ms pings to game servers in the nearest major metropolitan area . . .
Why do you think that an average of about 70 ms of input lack (as seen here for DOOM), is particularly high?
There ya go.
https://www.file-hunter.com/AMIGA/
The page has adds?Didn't even notice, uBlock with chrome, really makes the interwebs tolerable.Kinda cool. I haven't made it through the strange ads yet.
The page has adds?Didn't even notice, uBlock with chrome, really makes the interwebs tolerable.