Will Google's stadia kill the game console?

[mr friendly guy]

[Mr Bean]

No for three reasons
1. Input delay, they talk about how much better it is but any kind of game with split second timing suffers greatly trying to playing a streamed game. And forget about twitch based multi-player. So Civilization game? Sure, Rainbow Six Seige? Heck no your taking about your mouse click going to a server where it's rendered then back to you then on to the game server then back to you the back again. Even if they get it you to google to server to google to you that middleman is going to add at least 100ms

2. The greed and laziness of monopoly holding American ISPs, I see this doing much better in say South Korea but large amounts of American ISPs are still utter crap and sure Google might have a local rendering farm across the city from me but that does not matter to Comcast which is going to route my traffic through five of their servers six states away then back to the render farm.

3. It's Google, notorious for giving up on big budget projects because they only make some money instead of all of the money.

[TheFeniX]

"Will <X> kill <Y>" should near ALWAYS be met with a hearty boisterous amount of laughter.

Also, what Bean said. This isn't something like a land line losing out to cell phones (though as part of what Bean was talking about, land lines are still a big deal because U.S. infrastructure is shit). This is "Hey, that device that does all kinds of cool shit is going to die because Google has 'The Cloud.'"

The funny part about the Internet these days as well is my ping times to Dallas datacenters used to be in the 20-40ms range. They are now in the 100ms. Modern Internet/Gaming has tons of latency already just sending back "did this guy shoot me in the face?"

EDIT: Want to add "Will X kill the Y?" videos are pretty much the lowest-tier clickbbait. Even lower IMO than a video whose preview is a hot girl in scant clothing. Because at least SOMETIMES that gets delivered on.

[Esquire]

This whole concept seems like a solution in search of a problem. Does anybody actually want this?

[Mr Bean]

This whole concept seems like a solution in search of a problem. Does anybody actually want this?

People with servers they need to sell. They have business needs that have been met and are looking to see if they can resell the same thing to consumers. Google end-game after all is that you do everything on the servers but since they are Alphabet rather than Google these days it means when they try and compete with Facebook (Google+) rather than doubling down investing the money and understanding the issues they just give up. Now they have lots of servers for cloud storage, cloud this and cloud that hey why not have those servers doing something useful with all the extra runtime how about game streaming oh it only made 150% of our investment? Cancel it because that's not 500% of our investment.

[Ace Pace]

This whole concept seems like a solution in search of a problem. Does anybody actually want this?

Yes. In fact, this comes up often. To many people, PC hardware feels like a rat race, while many people would rather not spend a few hundred dollars on console hardware.

Nevermind that if Stadia and similar offerings (such as from Microsoft) work, then say good bye to long patches and install times, the game will just instantly load.

[Elheru Aran]

Nevermind that if Stadia and similar offerings (such as from Microsoft) work, then say good bye to long patches and install times, the game will just instantly load.

While I’m not a tech guy, this part of it makes a lot of sense to me. If you buy the game and it’s “preloaded” (probably not how it actually works but whatever) then you get an overall better experience. For example one of the annoying things about STO was always that periodically it had HUGE patches that took forever to load. If you can eliminate that via Stadia? Heck yeah. I can see a game experience being tweaked to accommodate the limitations of the streaming format as well.

[Solauren]

I fully suspect in a few years, there will be privacy concerns, and Google will get sued over the service.

[bilateralrope]

Previous game streaming services all died due to latency problems for customers in the US. What is Google doing differently that would help ?
Last I checked, the speed of light was still the limiting factor.

Living down in New Zealand, game streaming really worries me. We have internet that is fiber to the home, which should have enough bandwidth. But latency is going to be a problem even with servers in Australia. Even some turn based games like XCOM require mouse precision that is going to be aggravating if I have to deal with input delays.

[TheFeniX]

Yes. In fact, this comes up often. To many people, PC hardware feels like a rat race, while many people would rather not spend a few hundred dollars on console hardware.

Technical implementations are fucking terrible. It used to be you blew a mint on hardware to get the most out of a game graphically while sacrificing only what you needed to maintain X resolution and Y FPS because the games were pretty damn optimized (they had to be). If you were willing to sacrifice EVERYTHING, games like UT99, Quake, and Counter-Strike would run on a toaster.

But people want to spend $200 at Walmart for a Dell and run 4K max everything because that's what the bullshots show. You look at something like the PC port of Batman: Arkham WhattheFuck ever when it was cratering TITANs on release due to shitty (read: non-existent) QA. I mean, fuck, there weren't even SHADOWS at launch. Meanwhile, I STILL have no idea what voodoo was performed at Ubishit to get Farcry 5 maintaining a mostly stable ~40FPS at 4K on an aging 980ti.

Yes, the rat race is there, but only RECENTLY have developers not needed 4X the hardware (Hi, Halo 1 for PC) they should need because they suck. At some point, consumer hardware is going to reach the point where graphics has hit the max diminishing returns and it's just not going to require a lot to shit out movies with some QTEs.

There's going to be a few years while hardware catches up to shit like HDR and (hilariously) godrays, but it will get worked out. Honestly a lot of the other bullshit games have with photo-realism is they can't optimize textures and the accompanying normals for dogshit.

Nevermind that if Stadia and similar offerings (such as from Microsoft) work, then say good bye to long patches and install times, the game will just instantly load.

Thing is, patching only REALLY sucks for people like me: who are stuck on rural DSL (25 over 1mb/s) in which Streaming will do nothing to resolve. It's not even just the bandwidth, the ISP network consistently just fucking dies for what reason, I can only guess. My packet loss makes anything online useless for 2-3 seconds at a time.

So, if I played Single-player games to compensate: the idea of streaming them becomes laughable.

Back even on shitty Comcast, I could pull 100MB/s. Some patches were done downloading by the time I got up to make a drink. Entire GAMES (in all their bloated uncompressed glory) might make me watch youtube for 30 minutes, if even.

And my mean latency with Comcast in most online games pushed 100ms because we're in bizarro universe. The only good thing I could think to come of this is since all the client/server side stuff SHOULD be relatively instant, bullshit due to choke/loss should be reduced, but it won't be gone because the input is still vulnerable to loss and the servers themselves are going to have delay.

[Jub]

I think that the APUs, and possibly chiplets and 3d stacked computer cores, will probably solve the issue of hardware prices before Stadia takes enough of a market share to kill consumer level gaming PCs and Consoles. PC requirements still haven't really risen in a meaningful way over the past half decade or so and my old hardware is holding up just fine at the resolutions I play at. I'd need to upgrade to go to 2 or 4k but I also don't exactly care about getting there until a mid-range card/APU and a $300 32" plus screen can get me there.

[Vendetta]

This whole concept seems like a solution in search of a problem. Does anybody actually want this?

Yes. In fact, this comes up often. To many people, PC hardware feels like a rat race, while many people would rather not spend a few hundred dollars on console hardware.

On the other hand, that market niche of people who would rather not spend a few hundred dollars on console hardware every 5-7 years are also probably not the people who would spend $40 a month on fast enough internet for Stadia.


The other problem Google are going to have is Microsoft.

Microsoft have a streaming solution in the pipeline, XCloud. They have a game library and subscription business model, Game Pass. They have more Azure datacentres. They are establishing connections with other hardware platforms (they're already putting Xbox Live on Switch, XCloud is not an unreasonable next step).

[Ace Pace]

Vendetta, I'm more rooting for Microsofts solution than Stadia, for exactly the reasons you mentioned, along with far more experience running a game business.
However,

On the other hand, that market niche of people who would rather not spend a few hundred dollars on console hardware every 5-7 years are also probably not the people who would spend $40 a month on fast enough internet for Stadia.

You underestimate just how many people around the world have far better internet and far worse hardware prices. A simple example is my locality, where due to taxes, PC and console hardware is marked up nearly 100% in some cases, while internet is dirt cheap.

Similar examples exist throughout all of eastern Europe.


I think what most people think of as deal breaker latency, is completely wrong. These same people also think that graphics really matters or that most people play competitively.

[Ace Pace]

Ars Technica has a good article on Stadia and id, which has some interesting technical information.

WHAT ABOUT THE SPEED OF LIGHT?
While users can't be sure where the Google data center serving their Stadia experience will be, Land said that server will always be within a 64 km radius. With data traveling at around 200,000 km per second along fiber cable, that means a direct signal could be sent in under 320 microseconds (or just 0.32 ms).
"We know those [latency] speeds aren't reality for the open Internet," Land continued, citing myriad issues related to switching, congestion, packet loss, video encode/decode, last-mile delivery, and consumer AV hardware. But these are "all things that can be addressed and are continuously being addressed by the companies involved in building and maintaining the Internet," he added. In any case, the speed of light does not seem to be at the core of the latency problem.

Google has also had to launch from a standing start when it comes to the platform services side of running a gaming company. The first version of the service id saw was literally just a streaming video window with the ability to read inputs, Land said. With id's support, though, Google has now added Achievement tracking, identity and profile management, leaderboards, parental controls, save games, stat management, matchmaking, and the kind of social graph that developers and players expect from a modern game platform. "Just because you're Google doesn't mean you get [these] things for free," Land noted.

Now that they're convinced the technology works, Land said id is optimistic that streaming could let high-end games reach a market of mobile phone, tablet, and laptop users that's potentially ten times the size of the current console and gaming PC market. The company also sees big benefits in the reduced friction of users not having to download game files and the added security of the game binary not being exposed to the end-user.

[bilateralrope]

I think what most people think of as deal breaker latency, is completely wrong.

All I know is that previous attempts at game streaming failed due to latency complaints and I haven't seen anyone explain how their attempt at game streaming is going to solve the speed of light problem.

Sure, Google are probably going to design their first party games to work with the real world latency of Stadia. But that doesn't do anything about currently existing games.

The question is: Will players like the compromises that Stadia games require ?
Are there genres that Google will avoid ?

Ars Technica has a good article on Stadia and id, which has some interesting technical information.

WHAT ABOUT THE SPEED OF LIGHT?
While users can't be sure where the Google data center serving their Stadia experience will be, Land said that server will always be within a 64 km radius. With data traveling at around 200,000 km per second along fiber cable, that means a direct signal could be sent in under 320 microseconds (or just 0.32 ms).
"We know those [latency] speeds aren't reality for the open Internet," Land continued, citing myriad issues related to switching, congestion, packet loss, video encode/decode, last-mile delivery, and consumer AV hardware. But these are "all things that can be addressed and are continuously being addressed by the companies involved in building and maintaining the Internet," he added. In any case, the speed of light does not seem to be at the core of the latency problem.

You missed this update:

Update: Land tells Ars this only applies to id's testing, not more generally for Stadia. Ars regrets the error

[Ace Pace]

That didn't exist when I quoted

I think in the Microsoft case, this assumption is reasonable. Microsoft has enough data centers that the majority of the target audience is close enough to a data center, probably a few hundred km radius. In that case, the majority of lag would come from sources other than distance.

[Steel]

That didn't exist when I quoted

I think in the Microsoft case, this assumption is reasonable. Microsoft has enough data centers that the majority of the target audience is close enough to a data center, probably a few hundred km radius. In that case, the majority of lag would come from sources other than distance.

The whole speed of light thing seems like a complete red herring. Who cares why the lag is there, as long as it is there.

I should be able to get a round trip ping from Europe to East coast US in 40ms at 0.5c, but it is more like 200ms typically. So 80% of that lag is due to non-speed of light factors, and these factors will not be going away any time in the near future.

The minimum level of lag that this would add (given any realistic internet connection) would make any 3d games vomit inducing.

[Jub]

The whole speed of light thing seems like a complete red herring. Who cares why the lag is there, as long as it is there.

I should be able to get a round trip ping from Europe to East coast US in 40ms at 0.5c, but it is more like 200ms typically. So 80% of that lag is due to non-speed of light factors, and these factors will not be going away any time in the near future.

The minimum level of lag that this would add (given any realistic internet connection) would make any 3d games vomit inducing.

These are the best case numbers... That 200ms ping is one way, toss in the same trip for your inputs and you could be looking at 400ms reaction lag under current real-world circumstances. Add on shitty US internet, the fact that you can't play any games if you have a service outage, and monthly fee overload being a real concern and this is dead in the water.

[TheFeniX]

The whole speed of light thing seems like a complete red herring. Who cares why the lag is there, as long as it is there.

I am way out of the loop considering my heyday, but the Internet has become slower as it has become faster for multiple reasons. Ace Pace probably has a better view on it, but my opinion is basically: traffic shaping due to the monumental way in which Internet data and has changed.

Notably: streaming. Latency (outside insane amounts) is unnoticeable for the vast majority of Modern Internet traffic. Technology concerning streaming (some of it multicasting (I assume) WRT actual streamed content)) and the massive amounts of bandwidth for general video streams is way different than of the "old" Internet where most of it was simple send/receive stuff. This has added way more demand on the hardware itself and processing takes time, no matter how little. And every hop along the way adds a bit more. This adds up in the end. Physical distance has always meant a lot less online versus logical (hops, congestion, etc) and it's seemingly become even worse.

These are the best case numbers... That 200ms ping is one way, toss in the same trip for your inputs and you could be looking at 400ms reaction lag under current real-world circumstances. Add on shitty US internet, the fact that you can't play any games if you have a service outage, and monthly fee overload being a real concern and this is dead in the water.

Ping doesn't work that way. It's round trip.

[Jub]

Ping doesn't work that way. It's round trip.

*facepalms* I can't believe I derped that hard... That was an especially rank brain fart that I really ought to have caught.

[TheFeniX]

It's fine because even 100ms latency on inputs is going to be atrociously terrible. While there IS some inherent latency in input detection, when you hit a key on your keyboard, move your mouse, etc: it's salvageable. Even for a dbag like me, playing a optimized enough game: I can't notice it. My inputs feel like they are happening instantly.

Now, that input registry is going to happen no matter WHAT is processing it. In fact, if my dummy terminal (mom's Dell whatever) is ALREADY on the low end (read: high) of input latency and THEN that data has to be sent, processed, and the results sent back? Oof. Now, TBF: a lot of the middle man WILL be cut out here. But the bullshit isn't going away, just the TYPE of bullshit you will see is going to change.

However, I find the problem is going to be compounded here. Even for casual bullshit. Tablets (read: touch screens) can be/are notoriously shitty RIGHT NOW WRT input latency AND the precision of the screen. So if I'm going out and buying a garbage tablet to even play Candy Crush 54.0 and all that data is going to be processed online. So... shit input latency + network latency = broken tablet.

Thing is, there's inherent benefits to having the software run locally on hardware good enough to at least run it barebones and then using THAT as some kind of streaming terminal. But not having an instance of the engine running on your local PC means EVERYTHING has to be processed online before you get any response. And that's sounds fucking terrible to me.

[bilateralrope]

That didn't exist when I quoted

I think in the Microsoft case, this assumption is reasonable. Microsoft has enough data centers that the majority of the target audience is close enough to a data center, probably a few hundred km radius. In that case, the majority of lag would come from sources other than distance.

Where can I find all the Microsoft data center locations ?
If this page shows their locations, that's a lot of the world that's out of range. Remember that a data center in Australia is not close enough for someone in NZ.

The majority of humans within range is not good enough if game streaming takes off to the point that devs stop making games you run locally, as that would mean that anyone who doesn't live close enough to a data center becomes unable to play video games. And I do expect AAA publishers to decide to go streaming only if someone figures out a business model that makes it more profitable than games you run locally.

The game streaming business model is my other big worry about game streaming.