Will Google's Stadia Game Streaming Platform Be A Dud?
from the stream-this dept
On November 19, Google is expected to finally launch the company’s long awaited game streaming platform, Google Stadia. Stadia is being heralded as the vanguard of a new push to eliminate your local game console, and shift all of the computing and processing power to the cloud. The shift to game streaming is likely inevitable, the only problem is that Stadia may be a little ahead of its time. And, like so many Google projects (like Google Fiber), game developers are apparently worried that Google may waffle on its commitment to the project:
“The biggest complaint most developers have with Stadia is the fear is Google is just going to cancel it. Nobody ever says, ‘Oh, it’s not going to work.’ or ‘Streaming isn’t the future.’ Everyone accepts that streaming is pretty much inevitable. The biggest concern with Stadia is that it might not exist.
Granted, that same developer then proceeds to point out there’s plenty of projects Google hasn’t waffled on:
“if you think about it like that, that’s kind of silly. Working in tech, you have to be willing to make bold moves and try things that could fail. And yeah, Google’s canceled a lot of projects. But I also have a Pixel in my pocket, I’m using Google Maps to get around, I only got here because my Google Calendar told me to get here by giving me a prompt in Gmail. It’s not like Google cancels every fucking thing they make.”
Having watched Google promise massive societal transformation with Google Fiber only to have Alphabet bean counters suddenly cripple the project without admitting as much, the worries still aren’t entirely unfounded. But while Google’s ability to stick with ambitious projects is a worry, there are more pressing concerns facing the project’s success. For one the launch lineup is fairly pathetic. There are only going to be twelve titles at launch, most of which (including three games from the Tomb Raider series) have already been out for years. As such, many view this as more of a proof of concept and a paid beta than a serious commercial launch.
But the biggest problem for Stadia, as we’ve mentioned previously, is America’s shitty broadband connections.
Thanks to limited competition and negligent regulators, ISPs have imposed monthly usage caps as low as 150 GB on the nation’s broadband lines. Stadia, according to estimates, can consume upwards of 15 gigabytes per hour at 4K resolution. Yeah, you can scale back the service to lower resolutions, but that defeats the idea of Stadia as any kind of symmetrical replacement for traditional game consoles.
ISPs, for their part, have spent years pretending that these caps are a good idea, despite the fact that even the industry has admitted they serve no real technical purpose outside of charging you more money for the same service. They’re glorified price hikes only made possible by a lack of competition and regulatory capture. To try and make the limits seem generous, ISPs love to measure them based on how many emails you can send or web pages you can browse. Here’s AT&T’s breakdown from the company’s website:
Whether Google is the company that will dominate the space or not, one thing is clear: ISPs are going to need a new schtick, and a lot of consumers are going to be surprised by the fact game streaming burns through broadband caps like popcorn shrimp.
Maybe Google succeeds in the space, maybe it doesn’t. Maybe Google sticks with the project, or maybe like Google Fiber Alphabet execs get cold feet and hang up on developers mid-stride in a year from now. Either way, game streaming is likely inevitable. From Sony and Microsoft to Verizon (net neutrality and zero rating should prove interesting in the case of the latter), there are any number of companies eyeballing this space. Who’s going to come out on top is far from clear, though what is clear is we’re going to need better broadband for the idea to gain widespread commercial appeal.
Filed Under: broadband, stadia, streaming, video games
Companies: google
Comments on “Will Google's Stadia Game Streaming Platform Be A Dud?”
16GB per hour for 4k -- right?
The Vice article says 16GB per hours, not 1TB, right?
In a word, yes.
In a few more words, Google doesn’t care all that much whether this succeeds. They try a lot of things, they fail a lot of times, and every now and then something pops out a winner. Good on them for trying but it’s an idea way ahead of its time (and not even close to the first time). The bandwidth simply isn’t there to support it.
Re: Re:
It’s a bit of a self-fulfilling prophecy. People are afraid of what happens if it’s a dud, because Google’s shown they are not willing to support a dud even if people depend on it. The example of Google Maps is ridiculous. If it disappears tomorrow, I’ll go to Bing Maps, OpenStreetMap, old paper maps, whatever. It would be a minor inconvenience.
So Google’s strategy here is just to paint their detractors as irrational, desite Google’s long history of doing exactly what they fear? That’s not the strategy of someone who wants to win. If they intend to win, how about double your money back if the service gets shut down? That’ll cost them nothing if they don’t shut it down, and even if it peters out and goes on life-support, that amount of server time and bandwidth will cost less and less every year. Or they could guarantee that their server-side platform and all game code will be placed in escrow, to be released if it shuts down.
But no, this company is just saying "trust us", it’s not going to be like Google Video, Google Health, Google Talk, Google Hangouts, Google Spaces, Google Allo, Orkut, Google Code, Google Plus, Google Hire, Google Fiber Louisville… and you certainly wouldn’t be banned from Stadia with no explanation or recourse, no customer service options at all. Where does the general public get these crazy ideas?
The bandwidth is there on Google’s side and will eventually come to the consumer side. Probably would’ve been better to start with a non-US launch. Not Australia either. Maybe Sweden (their bandwidth was legendary, 20 years ago anyway). Hell, if they hadn’t bungled Google Fiber so badly it could’ve been the USA. Oops.
Re: Re: Re:
"this company is just saying "trust us", it’s not going to be like Google Video, Google Health, Google Talk, Google Hangouts, Google Spaces, Google Allo, Orkut, Google Code, Google Plus, Google Hire, Google Fiber Louisville"
I’m at a little bit of a loss as to why shutting any of these things down was a bad thing for Google to do. Video was shut down because it made no sense long term for Google to be hosting a YouTube competitor, and most of the others were shut down because they had essentially failed in the marketplace. If they had been individual companies rather than Google projects, they’d likely have been shut down because they weren’t able to make money ahead of competitors.
I understand annoyance and bitterness when Google decide to shut down a project because they theoretically have enough revenue to support hundreds of money losing services, but there’s not been any decision you listed that didn’t make sense, nor any that didn’t have competing services to move to.
"and you certainly wouldn’t be banned from Stadia with no explanation or recourse, no customer service options at all"
So, like every other online platform? I regularly see XBox gamers posting queries on Reddit where they’re told they’re temporarily blocked but don’t understand why and don’t understand the questions they’re asked to rejoin, for example. If this is your complaint against Google, what’s the actual difference to other platforms?
"Hell, if they hadn’t bungled Google Fiber so badly it could’ve been the USA"
The sole exception to the above list is the fibre rollouts. But, if you look at what actually happened there, it wasn’t "Google bundled the rollouts", but rather "existing monopolies did everything in their power to prevent Google from effectively rolling out on a large scale, so they decided to stall and look toward other tech that those monopolies can’t stop".
As for the other concerns, maybe another country might have been better, but all these companies are still very US-centric at the moment. Sony’s streaming gameplay was North America-only for a while, and Microsoft’s current beta is only US, UK and Korea.
I understand caution, but if your complaint is essentially that Google have a habit of shutting things down that lose them money and they might act the same as other platforms when problems arise, it’s a pretty weak argument.
Re: Re: Re: Re:
I think we should pay more attention to Google Fiber as a teaching moment. If a disruptive ~$900B company with a vested interest in breaking the ISP stranglehold on content consumers CAN’T get a toehold in the market, we should step back and say "holy shit something is wrong with this market" not "lulz the Google is always quitting things".
Re: Re: Re:2 Re:
Exactly. The American broadband marketplace is a bunch of colluding monopolies and duopolies who have managed to lobby for laws that prevent local governments from offering their own internet access and keep themselves as gatekeepers for new competitors.
If you see Google failing to break that at an individual city level with billions of dollars in investment and conclude that Google is the problem for not wishing to keep throwing that investment away, you might need to step back a bit.
Re: Re: Re: Re:
It will probably make sense to cancel Stadia too. The point is not that they shouldn’t have canceled those things, but that they cancel a lot of projects, and this could well be one of them.
Re: Re: Re:2 Re:
The point is that that have a history of cancelling projects that have not reached expectations or are losing too much money, as every other company does. I’m yet to see an example of a project that was unjustified in cancellation outside of "hey I liked that one".
Stadia may fail and die, but if it does it will be no less justified than when Nintendo or Microsoft cancels their less profitable ventures.
Re: Re: Re:3 Re:
You continue to miss the point – at least my point. I am not making any claim that any of Google’s project cancellations were unjustified, nor that cancelling Stadia would be unjustified. My point is Google has a history of cancelling projects, whether justified or not. This is probably because Google is more willing than most to take risks, and so many of them don’t pan out. To someone deciding whether to invest in a new, unproven Google project, the reasons do not matter, because at the start of the project there is no way to tell if this is going to be one that lasts a long time like Maps and GMail, or one that gets the plug pulled in a year or two.
Re: Re: Re: Re:
It’s hardly even a "trust us".. I mean asking for 130$ "up front" to be allowed to rent 12 titles going forward? It’s practically asking for donations
Re: Re: Re:2 Re:
They’re giving you a piece of hardware along with a package if games that typically retails for much more than that, along with access to the infrastructure to use them. Are you saying they need to give the non-physical product away for free?
Re: Re: Re:3 Re:
No I’m saying the hardware doesn’t do anything for you, only for them.. Whether you "own" the hardware or they do makes no difference whatsoever since all the hardware does is enable you subscribe to their service, so that’s what you are paying your 130$ for. 130$ for the ability to subscribe to their game rental service.
Re: Re: Re:4 Re:
I’m not getting your point here. Are you saying that Google should give physical hardware away for free? That they should make customers buy other hardware first, even though the hardware on offer is significantly cheaper that other options??
Re: Re: Re:3 Re:
And what you get for the service at the moment is so paltry that I have a hard time grasping that people will think it’s worth the up front cost they are charging for the hardware… You could probably permanently buy most of that offering for less than 130$ and play it on the hardware 95% of their market already has.
Sure it could get better later, but you can also sign up later if it does.
Re: Re: Re:4 Re:
The only part of Stadia I might use is the controller. It looks cool.
Re: Re: Re:4 Re:
"And what you get for the service at the moment is so paltry that I have a hard time grasping that people will think it’s worth"
Nobody’s forcing you to buy it. If it’s not for you, fair enough. If other people think it’s worth it, they buy it. What’s the problem?
"You could probably permanently buy most of that offering for less than 130$ and play it on the hardware 95% of their market already has."
Well, then they will fail if their offering is not worth the money. So?
Re: Re: Re:5 Re:
Yeah I’m not upset by it or anything, there’s no harm in it. I just don’t "get" it. The reason I said it’s more like a donation is that I don’t understand the value they are offering at this point. I get that there is a cost involved; I’m not saying they should give it away. I just don’t see the value to the customer especially at the beginning so I don’t quite get how they plan to sell it.
Re: Re: Re:6 Re:
"I just don’t see the value to the customer"
Well… this is where we wait and see. The thing is, the gaming market is a wide things with many niches. Just because you don’t see the value in something does not mean that there aren’t a number of people who will. What’s a mystery to you might be the sweet spot for others.
The example that immediately comes to mind is the Wii. Before its release, it was mocked as a grossly underpowered console that couldn’t hope to compete with the next gem Microsoft and Sony consoles, to the point where several major 3rd party publishers didn’t bother to plan titles for the platform. Within weeks, not only were those same publishers clamouring to get games on there as quickly as possible, it opened up gaming to an entire new set of people never previously considered as "gamers".
Now, I’m not saying that Stadia will have anything like that impact, and it could well flop. But, criticising Google for asking for money for hardware they provide is not particularly helpful. I’d get it if they were trying to charge the same as a new PS4, but they’re not. The point of a console vs PC is that there a small number of predictable configurations that they control, and I don’t see anything particularly wrong with the pricing here in terms of what you’re actually getting.
The only question is whether it’s compelling enough for people not already tied into the ecosystem of an existing platform, since many existing console gamers will just stick with what they have and wait for their current provider’s streaming options (I’m in the latter camp here – I’m very much tied into the XBox achievement system and am happy to wait for their streaming platforms to become available).
Re: Re: Re:7 Re:
I’m not criticizing for asking money for hardware. The hardware is just implementation details of how their service works and doesn’t matter to me. I’m actually more in the camp of "I can’t find an excuse to sign up right now".
I do see the potential value in the future if things go well and they get a bunch more games, and I could easily see myself signing up at that point but we are not in the future and anyone can always just sign up later if it does improve so I just don’t see the benefit in being an early adopter. It almost feels like I would be signing up for a kickstarter at this point putting my money down in order to help them get off the ground so the service has a chance to succeed. I feel like they should be offering a limited time thing or something extra for early adopters, but like you say it could be just me and maybe there are tons of people who think 130$ for the current offering for 3 months is worthwhile as it stands.
Re: Re: Re: Re:
Out of the list of Google services that had been shut down, those looked like the things most likely to have people who’d depended on them. The important thing is that for most of those, Google didn’t offer much help to the users they inconvenienced. Why couldn’t they copy every Google Video over to YouTube and have the old videos.google.com URLs redirect there forever? Or just make some of these services read-only and leave them up. I don’t recall them helping Google Plus users migrate all their old data and social networks to Facebook, or auto-migrating Google Code repositories to Github. Eventually that stuff was just gone.
Anyway, you’re making the same mistake Google made: thinking that each of these inconveniences is minor, is an isolated incident, and won’t cause the users to hold a grudge. It’s not hard to imagine that some of the heavier Google-users might have had a stretch of time where they were getting kicked off of several services a year, where every infrequent group-chat was via a service they’d never used before. Regardless of whether you or Google think users are right to view Google as undependable, it’s clear that some of them do.
One of the major concerns with Stadia is that the poor broadband connections provided by these existing monopolies will stop it.
Re: Re: Re:2 Re:
"those looked like the things most likely to have people who’d depended on them"
Google aren’t a charity and most of their software has competitors. They have no obligation to run a losing business just because you like it. Same with any other company.
They are also not obligated to do the work of migrating for you. I’m sorry if you tied yourself into a single service and are too lazy to work to move to somewhere else but it’s not their obligation to provide another free service to do it for you. It would be nice if they were more helpful at times, but after providing free services they don’t have to do more work for free when they realise your leeching is not beneficial to them any longer
This happens all the time, online and offline, with other companies and I never hear the same demands
"One of the major concerns with Stadia is that the poor broadband connections provided by these existing monopolies will stop it."
Then whine at them instead of the company that at least tried to get the infrastructure in place first
Re: Re: Re:3 Re:
That attitude is exactly the point. They know they have no obligations whatsoever to their users, and they’re not shy about taking advantage of that. The would-be users are starting to catch onto it, and are expressing a reluctance to spend time and money to be Google’s guinea pigs. Users have no obligation to Google either.
But hey, it’s all legal.
Who’s making demands here? People are saying they expect Stadia to be dead soon, and they’re not going to waste their time with it. That’s not a demand.
And people absolutely do express concerns about "fly-by-night" companies, and avoid dealing with them, in other lines of business. Google’s not a fly-by-night company, but their business units seem to have little more concern for their users than does a shady moving or roofing company.
Re: Re:
The bandwidth is fine for many people. I can get over 100 Mbps for example. It’s the data cap that’s the problem.
If it consumes 1TB per hour Google didn’t put enough engineering time into the game system.
Obviously traffic congestion is going to be a huge issue with that type of network configuration. I see no reason that a significant part of the load shouldn’t be offloaded to the clients computer like normal games do.
Re: Re:
I also don’t think streaming is the likely inevitable future. There’s a reason every consumer that could ran away from the IBM cloud after the 1970s when it was called a mainframe. Microsoft and Intel spent the next few decades eating IBM’s lunch because IBM kept doubling down on the belief that home consumers and small business would want to return to their mainframe (cloud). There are still large organizations for which the mainframe’s from IBM made sense and are still in use today but it is not known to be a home consumer friendly business.
Both my parents worked at IBM and both of them continued at other jobs that continued using IBM mainframe products and they aren’t known to scale down conveniently.
I wish google the best of luck on its project though.
Re: Re: Re:
At the consumer level, there’s no way people could have afforded the computer time or the connections for that. Home computers were hundreds of dollars, maybe thousands for the pricey ones, and IBM’s not going to talk to you for those amounts.
On the business side, the home computers were cheap enough that people could "sneak" them in—a manager could approve an expense that size without having to justify anything to a budget committee. One couldn’t "try out" IBM as an experiment.
None of that seems relevant to Stadia…
Sun tried to bring back mainframe computing as client/server computing, around Y2K, and failed. Now that it’s branded as "the cloud", it’s really popular with businesses in particular (when the internet connection fails at my workplace, they send everyone home…). But consumers are willing to use it too, such as for video streaming services. We might have predicted those would fail due to ISP bullshit, but they’re wildly popular.
If Stadia fails, I expect it to be because of Google’s reputation, or maybe gamers being averse to not owning things. On the technical side, even with American ISPs, it’s probably good enough (for enough people to make it a viable market).
Re: Re: Re: Re:
When using a mainframe remotely the old terminology was "slave terminal" and they were cheap. They were basically a monitor, modem, and keyboard.
You could run one from anywhere you could lug it to a phone line.
Re: Re: Re:2 Re:
Yes, there were acoustic couplers in the 1970s (direct-connect modems were newly legal, but too expensive). Was there anywhere to connect to in those days, if you weren’t affiliated with IBM, a university, or a national research lab? BBSes didn’t exist till 1978, and online services for the public weren’t widely available till the mid-1980s.
Re: Re: Re:3 Re:
IBMs main business line is computer services. They sold it and still sell it everywhere. It was not confined to just IBM in any way.
Anyway, the point is that decentralized computer architecture and centralized computer architecture competed directly for decades and people don’t like centralized computer architecture because individual consumers and small businesses don’t like losing control of their documents and entertainment.
That is the lesson of great lesson of IBM losing to Microsoft, Intel, and now Apple. It’s still a great company but they used to be the undisputed industry leader before they made the bad technology bet on centalized mainframes over decentralized PCs individuals could control.
Re: Re: Re:4 Re:
You seem to be confusing the point. You seem to be addressing enterprise-level tech and consumer level tech at the same time, which is never going to be a valid concern. Yes, lots of people hate centralised services, meanwhile companies like Citrix have made their entire worth on providing such things. It’s all about the needs of the customer, and different sectors have very different needs.
"individual consumers and small businesses don’t like losing control of their documents and entertainment"
Which is why cloud services for those things have never taken off. /s
You do realise you’re saying this on the same week that Disney+ was brought to its knees because too many people were rushing to use it at launch, right?
Re: Re: Re:5 Re:
You seem to be confusing the point. You seem to be addressing enterprise-level tech and consumer level tech at the same time, which is never going to be a valid concern. Yes, lots of people hate centralised services, meanwhile companies like Citrix have made their entire worth on providing such things. It’s all about the needs of the customer, and different sectors have very different needs.
There is a very large market for cloud services, however consumers still prefer things they can keep at home based on all my experience. Also, IBM and Microsoft/Intel absolutely did compete for home consumers and small business.
You do realise you’re saying this on the same week that Disney+ was brought to its knees because too many people were rushing to use it at launch, right?
… and there are absolutely no ways to save copies of streamed video if you feel like it. /s
Not the same thing.
Re: Re: Re:6 Re:
"There is a very large market for cloud services, however consumers still prefer things they can keep at home based on all my experience."
SOME consumers. Others are happy to be replacing their physical media with cloud services for various reasons, from catalogue range to hardware issues to kids destroying the DVDs every time they get their sticky fingers on them. Your broad brush is the problem here.
"Also, IBM and Microsoft/Intel absolutely did compete for home consumers and small business."
They also worked together at many points in time. So? The point is that it was the advent of better suited tech that caused businesses to jump from IBM mainframes, and IBM weren’t interested in the consumer market long term.
"and there are absolutely no ways to save copies of streamed video if you feel like it"
There certainly are. However, very few people who just subscribed to Disney+ will be doing any such thing.
"Not the same thing."
Why is a cloud service not the same as a cloud service for your argument? You have to be specific here, because your argument is all over the place.
Re: Re: Re:
Never mind that MS and others are back to pushing edge computing (likely forced if you want to continue to use their products), while their clouds are meant to hold all the data and applications. Whatever serves their interests in extracting rent while bearing as little of the overhead as possible.
Re: Re: Re:
"There’s a reason every consumer that could ran away from the IBM cloud after the 1970s when it was called a mainframe"
Yes, because most people didn’t need that kind of service and the PC met their needs better when the option appeared.
There are people for whom a Stadia setup makes more sense than constantly upgrading a PC, buying a new console every few years that may or may not be backward compatible with the games they want to keep playing, etc.
As with most things, the future is not all or nothing. Some people will continue to buy physical media and having the best local experience, some people will prefer Stadia.
"There are still large organizations for which the mainframe’s from IBM made sense and are still in use today but it is not known to be a home consumer friendly business."
Yes, after inventing the PC market, they got outplayed by competitors and eventually decided that they’d be better off retooling with open source technology and concentrating on large big iron projects, than try to compete in the consumer grade commodity market. They were never really cut out as a consumer level organisation, and there’s nothing wrong with them realising that.
'... why am I not buying the games again?'
The main problem I see cropping up even if they do get a worthwhile lineup is definitely the data cap one, in that a service like that will use up so much data that people will either very quickly find the service unusable as their connection is throttled to a crawl, or paying so much in overage fees that they might as well have bought the games.
In a country where high-speed, unlimited data connections were a thing this might have had some potential, but in the US it’s almost certainly DOA for any number of reasons, chief among them the abysmal internet service most people pay for that all but ensure the service will be useless for all but a handful of people.
Re: '... why am I not buying the games again?'
Yeah, data caps are obviously a bottleneck. But it may help bring the issue to a head. Few people currently consume enough to anger their ISP overlords, but as the overages become more widespread people may come to understand how much they’re really being charged for internet connection.
I haven’t used a streaming game yet, but I assume they’ll favor downsampling rather than buffering stutter and lag on deficient connections. The game should still be playable at 1080p.
Not Yours
The biggest issue I have with this is it’s yet another way of showing we don’t own what we buy. This is a blatant attempt at making damn sure that you don’t own even the slightest bit of a game – you’re renting not just the game, but the game machine the game runs on. You own the TV and the controller, and google owns the rest. Wanna mod the game? Sorry, it’s not yours. Wanna keep playing after Google inevitably shuts it down. Sorry, all you ever got was a stream of video. Nothing to hack to make work later (even more of an issue than always connected games!).
Buffering
I have yet to be somewhere i can start a movie without "buffering", so i have serious doubts about gaming on a service like this… It can take several seconds for a movie to buffer, so unless i start seeing movies and tv shows that can stream with no noticeable buffering i cant see gaming working any better
Re: Buffering
Stadia’s going to be an acid-test for connection quality. There’s no papering over temporary dropouts; any delay over 5-10 ms will be noticed.
The lengthy buffering for video only means your connection’s bitrate is not significantly greater than the video’s. Video providers push buffering onto everyone, because that’s easier—it can paper over temporary network hiccups at either end, but that doesn’t mean they’re actually happening.
Re: Re: Buffering
Any noticible buffering… that includes that less than 2 ms load that happens on most anything that isnt cached locally… And im this sort of content doesn’t benefit from CDN, so there’s that too
Yeah, that time to go from controller to server means at least a round trip on what you do impacting what you see… so yeah, i would expect instantaneous playback to be a working thing long before gaming as a service to be a serious competitor to a local system.
Im not sure if you realized that this is more of a latency issue… Your bandwidth capacity is meaningless if you have a terrible latency (ping)
Id love to see it actually work… Then would expect video to work a bit better. As it stands now, got plenty of bandwidth myself and things are great when im watching content thats able to come from a close by cdn, but damn if i can’t.
Re: Re: Re: Buffering
2 ms is fine. Only the most serious gamers argue about latencies less than 10 ms (e.g., buying low-latency TN monitors and running them with FreeSync above 60 Hz). If a video ever stops to buffer in the middle, you’re probably screwed for Stadia.
Google will have to distribute the content widely, in some form. Call it a CDN or not, but the game will be running on a server near you, or will be unusable.
Yeah, but unlike traditional online gaming where it’s almost entirely latency that matters, the bandwidth usage here is significant enough to matter too. Especially if 2 or 3 people in the household are using these, or if someone else is streaming video or downloading whatever.
Re: Re: Re:2 Buffering
Yeah, bandwidth will also be an issue, but without decent latency, you wont get far enough to worry about bandwidth…
Latency on an led monitor is not the same as latency on a network… The are similar in that they are a time measurement, but one is the time for a packet to traverse the network, the other is the time it takes for the led to transition to another color without the previous being present… Most of the early led had some ghosting issues, but these days thats mostly gone.
Id love to see google build more datacenters for this, but have my doubts that it will ever be widely available. I expect a DC filled with game servers to run stadia are going to be significantly more expensive than a traditional cdn.
But it would be so awesome if it works, they keep it and make it available near me.
Re: Re: Re:3 Buffering
John Carmack wrote an article about latency that you may be interested to read. The LCD monitor delays were not all based in physics; some were caused by image processing (sometimes unnecessary processing, like passing an image through the scaler when no scaling is required).
All sources of latency can matter. They matter more for certain types of games like first-person-shooters, and some of the really competitive people are clinging to CRTs and running them at 180 Hz or more.
From the gamer’s point of view, it’s all the same. They press a button on their controller, and it takes some time before they see the result. There’s USB/Bluetooth latency on the controllers, OS latency to encrypt that and build a packet, delays from store-and-forward network devices (cut-through devices impose much less latency), then the speed of light in fiber (high-speed traders go with radio for a 50% speed-up), then the server has to process the input, draw a frame, and do all those network steps in reverse, and finally some data reaches the monitors and speakers.
There’s a kind of "uncanny valley" effect with latency. If it’s almost-but-not-quite good enough, people will just feel Stadia is ineffably worse than a game console. Few will be savvy enough to call their ISP and know what to say; Google will need to get good telemetry and deal with the ISPs directly. There’s a good chance they’ll need to deal with router manufacturers too, to ensure routers can maintain low latency even while streaming unrelated 4K videos.
FYI, Google doesn’t just use its own datacenters. Some Google hardware exists at ISPs. Stadia boxes might end up there too.
Re: Re: Re:4 Buffering
The LCD monitor delays were not all based in physics
You make a number of credible and convincing points in your post. However I have one question. Is not non-physics based LCD screen manufactured by this thing?
https://cdn.arstechnica.net/wp-content/uploads/2019/10/unicorn-rainbow-slinky-800×450.jpg
Re: Re: Re:5 Buffering
ughh
That’s supposed to be:
Is the non-physics based LCD screen manufactured by this thing?
I need to install the grammar checker again.
Re: Re: Re:5 Buffering
Yeah. There are physical timing limitations to the crystals, and then the integrators, thinking they can use unicorn magic to improve image quality, fuck up the latency. And a panel that could physically run with 2 ms latency ends up taking 5 times as long because of the firmware.
Re: Re: Re:4 Buffering
So, um, none of that matters to starting a movie… all in all, I would consider it a far easier task, but for me we not even close to the ‘uncanny valley’ effect for a movie, it still an incredible delay from button press to go…
From the movie watcher’s perspective, the cause of the delay is the same…
That whole argument was beside the point of the original post… shouldn’t a movie be more or less able to start somewhat similar to a movie put into a tray before we can expect the same sort of effect for a gaming system?
Re: Re: Re:5 Buffering
Kind of. You need a connection on which a video could start instantly (within tens of milliseconds). That you see buffering now doesn’t mean your connection is incapable of better; the video provider might be lazily applying the same buffering settings to good and bad connections, and the buffering itself depends more on bandwidth than gaming might. (Linear video playback is insensitive to latency, within reason.)
More specifically, what you want for gaming is enough downstream bandwidth to stream video at your preferred resolution and quality (5-10 Mbit/s minimum) and probably below 30-50 ms latency.
Re: Re: Buffering
5 to 10 ms? Lemme check my Spectrum high-speed broadband… let’s ping google.
Hmm, minimum latency of 44ms, average of 125ms… yeah, streamed games would be perfect over Spectrum broadband. 😉
Re: Re: Re: Buffering
That’s horrendous. Light can travel about 8800 km through fiber (2/3 c) in 44 ms. The lower end of that latency is probably within the uncanny valley, good enough for non-latency-sensitive casual gaming if there’s not much deviation. But look at the maximum! That packet could’ve done 6 loops around the planet.
Re: Re: Re:2 Buffering
Yeah, but it’s pretty common for cable modem. I check it periodically, and it’s never any different. Only +/- 5ms or so from what I posted above. Cable modem is fast, but has horrible latency. I got better latency on DSL, but the price/bps is horrible for DSL. Wish I could get fiber…
Re: Re: Re:3 Buffering
I just pinged Google on my cable connection and got 10ms min, 12ms max. I can’t tell from that which of us is the outlier, but at the minimum I can tell 125ms ping is not universal for cable internet.
Re: Re: Re:4 Buffering
See this page:
https://tools.ietf.org/id/draft-white-tsvwg-lld-00.html#rfc.section.2
"Finally, the third [column], ’99th Percentile,’ gives an indication of the maximum [round-trip] latency that a customer would commonly experience in real usage scenarios."
You’re getting best-case latency for DOCSIS 3.1 Active Queue Management, which has 99th-percentile latency ~100 ms. Low Latency DOCSIS 3.1 promises to lower all numbers to 0.9 ms (or 4.8 ms without Proactive Grant Service). That would be the round-trip latency to your ISP, not to Google.
JoeCool’s numbers match DOCSIS 3.0 Early Equipment (or, presumably, older DOCSIS versions).
Re: Re: Re:5 Buffering
Cool, good to know. Thanks!
Re: Re: Re:5 Buffering
Yeah, clearly NOT DOCSIS 3.1. Ran the ping for a while again and eventually got a 31ms min latency! Wow! Doing really good this afternoon. 😀
new numbers..
400g limit.
avg Sd movie 700-800meg CD
Is about 4-7 moves on 1, 4gig DVD
so, about 400-700 movies, in low res..
Lets Double to quadruple the size for HD and above.
so we 400-700 and /2 and /4…
200-350
100-175…MAYBE..
For 1-2 people thats not to bad..
I get about 20 YT per day, 600 per month at 20-30 min avg..
1 hour would be 300, a 2 hour movie 150..at 1080p
Now divide that for a family of 4..
Re: new numbers..
I bet the telcos would love to find some way to go back to per-minute charges, for any kind of streaming. People "streaming games" from out-of-town BBSes in the 1980s quickly found out how ridiculous those charges were.
Re: Re: new numbers..
until someone can get gigabyte Data up and running..
We have a Big problem they Dont want to fix..
These are not the builders of the infrastructure of the net.. they are the bill collectors. they fix things as they break.
We have telephone poles over 40 years old Still standing.
so we have a few corps responsible for the Ends of the lines, that are Cable/Sat/Telephone/cellphone/.. all communications across this nation, and they Have been paid to increase and update the End lines..
Oh! and they get REAL upset(pissy) when a state or Town installs its own. the Most fun you can have is to NOT have any of their devices. Because that means they dont get any money form you. so they cant pay off the City/state/ Congress/regulators.. by raising your Prices.. And they do that, because they dont want to loose any Money they give to the Boss’s/owners/…
It all comes out of our pockets, when they payoff someone to Do their bidding.. And Why most of those $19.99 for the first year and $100 after that for 2 more years…Contracts, are Worthless. It will double, then triple..
where's google fiber when you want it for google stadia?
a twofer match
I couldn’t use it if I wanted, 3mb dsl. 3 miles away its 20mb. Which is top for our town around 50. Rural America is still being screwed by the telcoms.
Re: Re:
Interesting word choice.
Tell me, if you were the CEO of Uber Telecom, Inc. would you approve a $30m+ project to serve 50 people knowing it would take over 1000 years just to break even?
The Math?
I don’t get the math. It costs me upwards of $2000 to $3000 for a top of the line gaming rig that can run current games at 60+ FPS (ideally 120 FPS) at up to 4k Resolution. Google needs to put in the range of 2x my hardware per person playing at the same time at least in order to make the entire project work. Even with huge discounts I cant see an FPS game running at 120 FPS being doable.
Lighter games, like Civ 5 or Cities Skylines would perhaps be ok. Those games do not need light speed FPS to be great can work with the service, but I don’t think it can run Crysis.
Re: The Math?
Most people won’t be playing most of the time, but the service is likely to be significantly peaky. "Everyone" will play during the evenings. Luckily, "evening" is different in every time zone, and businesses will be inactive during that time. So my best guess is that this is an attempt to duplicate the CompuServe stategy: the home users occupy Google’s data center after the businesses leave. And then the remaining late-night capacity can be sold cheaply for batch jobs (like scientific tasks).
If this hypothesis is correct, we should see more availability of GPUs on Google Cloud servers.
Hotels will love it
This will be perfect for Hotels with an incentive to get good internet. Once Starlink is working, bandwidth for systems like this will no longer be a problem. Yay progress.
Re: Hotels will love it
You need very low latency too though, what is the latency going to be like with Starlink?
Re: Re: Hotels will love it
Too much.
https://en.wikipedia.org/wiki/Starlink_(satellite_constellation)
Re: Re: Re: Hotels will love it
That is latency for geosync communications satellites. Due to the lower orbit of starlink estimates are that the latency will be approximately 25-35 milliseconds, which should be acceptable for most games.
Re: Re: Re: Re:
I’m curious, did you just honestly not read the first part of that sentence where it said that was for current geo-synchronous satellites? And did you also not read the very next sentence:
Re: Re: Re:2 Re:
Apparently… my bad.
Re: Re: Re: Hotels will love it
I’m gonna assume that the article was edited after I read is and that my reading comprehension skills aren’t really that poor. That’s the ticket.
Re: Re: Hotels will love it
The closest satellites will have an altitude of 340 km. They’ll rarely be directly overhead, so let’s say around 1000 km to go up and back down. That’s about 3 ms of latency. But data in space travels 50% faster than fiber, so space-based routing could more than make up for the up-and-down latency, maybe giving ping times below 10 ms total.
In theory. This is guesswork with best-case numbers, and Elon Musk’s goals and expectations haven’t always been realistic in the past.
Google services are in perpetual beta state with 50% chance they will be dropped in less than 10 years. Sometimes they force services on you if you use one or more of their other services, then drop those after lunking all your accounts and hoovering your data. Why am i not shocked that some people are not overexcited for Stadia?
Beyond that, why force pointless traffic over the network? Certainly, broadband is pretty sad in the States and could be astronomically better, but i still don’t understand how it is sensible (other than from someone’s business perspective) to generate loads of network traffic for things that can be done locally.
Re: Re:
"Google services are in perpetual beta state with 50% chance they will be dropped in less than 10 years."
So are most other peoples’ services, at least Google are honest about it.
"Beyond that, why force pointless traffic over the network?"
Define "pointless". Why is this less valuable than whatever you like doing?
"i still don’t understand how it is sensible (other than from someone’s business perspective) to generate loads of network traffic for things that can be done locally."
I heard the exact same argument from people arguing against Netflix streaming being a viable long term investment. How did that turn out, again?
Re: Re: Re:
They’ve been doing streaming for all of 12 years. If you want "long term" results, check back in a few decades. We haven’t even had a market crash since their streaming service became popular.
I am in the cancel camp
I think it will have moderate success and will be a well done product. I also think the lack of enough homes with sufficient bandwidth will be the biggest reason it only has moderate success. My fear and the reason I won’t be purchasing is that this moderate success won’t be enough success for Google and they will cancel Stadia.
To the developer that was quoted in the article: You are right they haven’t cancelled every project, but they have canceled enough that people have become wary. It is one thing to use a free product that is cancelled, but a product pay for is in a whole different league.
let me get this straight
If this is marketed to the casual gamer who:
-already is barely playing video games as it is because time/non-interest
-already is only spending minimal money on video games as it is because time/interest
why does anyone (marketing dept. I’m looking at you) think that this same niche is going to be willing to pay indefinitely (subscription) to play video games?
Genius . Greedy . Google
Re: let me get this straight
That’s a bit of a weird comment. First you say that this will never sell (because, apparently, you understand more about Google’s target market than they do). Then, you’re saying this is all about greed.
Which is it? Are they throwing money away or are they greedy?
Re: Re: let me get this straight
Usually I see comments from "casual gamers" who consider this a possible deal, but if you’re already only a casual gamer with a fleeting interest (and time budget) in gaming, why sign up for a monthly SUBSCRIPTION for games?
Who’s going to pay monthly for a form of Entertainment in which they only have a (self admittedly) limited interest?
I wish Google would have continued to disrupt the ISP monopolies instead. Then, with a low priced local ISP, Stadia could have been the killer app.
Genius . Greedy. Google
Re: Re: let me get this straight
Oh yeah, to answer your question, PaulT. Google IS greedy, from a data-collection-without-my-permission-get-rich-by-selling-my-data-and-not-break-me-off-any point of view.
In the end, while I’m glad we’re getting more options in the gaming world, I’m not a fan of Gaming As a Service/$ubscription.
Re: let me get this straight
You do realize Stadia standard is free (No monthly fee) right?
Perfect Storm
I’ve thought for a while that the 4k hype was overblown because it’ll take a while to actually get content that can take advantage of the format. But video games have the potential to generate 4k content dynamically, greatly expanding the format demand.
I was down on content as service vs content ownership for multiple media: video, music, and now video games. But I’ve come around on all of them for all-you can eat subscriptions rather than buying a la carte. (xbox game pass is still downloading the games though)
So the time seems ripe for an explosion in demand for streaming game services. People are looking for 4k content and people are comfortable not owning the titles. As long as the lag is acceptable and hardware is cost effective it could take off like wildfire, and not just as a google service.
I imagine what most users will see is downsampling when ISPs fail to maintain sufficient bitrates; like encrypted traffic, streaming games are CDN immune. As long as the game streamers can communicate that this issue is because your ISP oversold their capability, I think we’ll see the telcos scrambling to catch up with their claims of the last few years. Of course it’ll create an even greater peering asymmetry, so we’ll also see ISPs shaking down streaming providers for even more while captive customers don’t get what they paid for.
Data usage estimates
Just as a little factoid: The game Red Dead Redemption 2 for the PC clocks in at 112 GB, and is only available via download. The first patch received was another 9GB.
This month I also redownloaded World of Warcraft at ~60GB, and Fortnite I think is around 50GB.
And if I actually download my entire Steam library I would need about 2TB.
It,ll be a flop in america ,where most broadband is capped,
it might work ok in countrys like japan or korea where the internet is fast and
they have a generous .
I don,t see the point of streaming games , i can join the epic store and get loads of games free or very cheap without using up my broadband allowance .
Maybe its 5-10 years early,
streaming music is very popular because smartphones are fast and powerful
and mp3, music doe not use alot of data .
Where is the casual gamer who does not have a ps4 or an xbox1.
But just wants to stream games on his pc or laptop?
And also owns a 4k hdr monitor .
This is a very narrow market.
its, like a company selling jeans but only for people who are
over 6ft 5inch in height .
Yes console games are now over 100gig ,but you can buy em on a disc,
and you only need to download em once .
Re: stop it
Now just shut up and listen, for once in your life:
Google reallllly needs this money!
again, options are the best option
When it comes to entertainment, options are what we really want.
-we don’t all have to watch chick flicks
-we don’t all have to buy physical media
-we don’t all have to rent physical media (redbox)
-we don’t all have to rent digital media (iTunes)
-we don’t all have to stream digital media
We have options…options are good.
So did anyone think of the issue with the Internet of Trash?
While I’ve seen some people already mention that not everyone will be willing to switch to game streaming because having a copy of the game on your PC means you will be able to play it if the internet is down and it’s got single-player I would like to point at something else as another problem for streaming: DRM.
Streaming games is such an improvement to DRM to control who has access to a video game it’s not even funny, but any streaming service that works will be copied by the AAA market for the purpose of better DRM and then when the DRM inevitably causes problems for paying customers (which it will as every other DRM before it and for the same reasons and you can add caps and throttling by the ISPs on top of that) and the fact that pirates will figure out how to rip the games out of the stream (recently Diablo 1 got recreated without the use of the original source code for an example of how such a rip might come about even if the source code for the game gets locked up) it will put more people toward ownership of the game.
And that is before we get into the fact that since the games are on servers DDOS is now an option for people to block games they don’t like and that while Google and Alphabet are actually decent at cybersecurity the fact that they will be working with content corporations on this means that there will be shit equivalent to Content ID and other crap like that from Youtube.
Get ready for an Object lesson in why Game (Program) Streaming will ever be the thing that is just out of reach of truly working as advertised.