Sections

Kicking off the streaming wars | Opinion

GDC 2019: Industry giants have decided that the medium's future lies in streaming - but success is far from assured even for enormous players like Google and Amazon

That streaming services will play some kind of role in the next generation of video games is already a given. The industry's biggest players have made that decision and stepped up the enormous investment in technology and infrastructure required to make it into a reality. Sony's PlayStation Now has been slowly evolving in the background with fairly little fanfare; now Microsoft's xCloud has taken an important step forward with a live demo, supposedly a lead-in to public trials later this year.

The other shoe whose dropping many pundits are eagerly anticipating belongs to Google, and that will hit the floor at GDC next week. The company has hinted at a gaming announcement more significant than anything it's done in the space so far -- hardware may be on the table but it's widely expected that a streaming service will be the main dish, with anything physical being in the form of controllers or streaming boxes to support the service. As the notion of a "Netflix for Games" service takes root, all of the major players are positioning themselves to make sure they get their slice of the pie.

On the surface, this puts Sony and Microsoft right onto a collision course with Google -- and it also raises big questions about what other market players will do to ensure they're not left out.

"Even if Apple's blind-spot around gaming keeps it away from serious engagement with this sector, Amazon looks like a shoo-in"

If Google and Microsoft -- two of the biggest cloud services providers in the world thanks to Google's Cloud Platform and Microsoft's Azure -- are using their infrastructure to these ends, surely Amazon -- whose AWS platform dwarfs them both -- can't be far behind? And, while it's not such a player in cloud services (aside from the infrastructure it's built for its own systems), Apple surely won't let Google steal such an important march on it in terms of services offered on their competing platforms?

I wouldn't actually bet much on that last prospect, but even if Apple's historical blind-spot around gaming continues to keep it away from serious engagement with this sector, Amazon looks like a shoo-in. And having four of the world's biggest tech companies (a line-up in which Sony actually ends up looking like something of a minnow) all competing for the same prize looks like a recipe for bloodied noses, bruised egos and ravaged bank balances.

Just as the video streaming market is getting ready for an almighty showdown as both Apple and Disney take aim at Netflix's market share, the games industry could also see its streaming services becoming a battleground for giant companies with far broader ambitions.

Ironically, the last time major industry players tried to make gaming into a key battleground for a broader conflict the most tangible result was the Xbox -- a console market entry designed in no small part to give Microsoft a beachhead in what it feared could become a major battle for consumer hearts and minds with Sony. Now both of those firms look set to end up in a direct conflict with even bigger companies for dominance of game streaming.

Or are they? Looking at the substance of what both Sony and Microsoft are doing in this space, something very striking emerges. Both companies are pitching game streaming as a "fill in the gaps" kind of service, and not a replacement for anything you currently do. At least for the foreseeable future, PlayStation Now and xCloud are both positioned as being a way to extend your gaming experience so that you can pass some time while you're commuting or away from home by engaging with your favourite games on a tablet, laptop or mobile device. They are snacks between meals, not alternatives to meals.

"There remain a lot of places where broadband speeds are not up to the task of providing high-quality game streaming"

If you're Google (or Amazon, or Apple), then the pitch has to be somewhat different. Google doesn't have any previous skin in the gaming space -- its function as creator of Android arguably gives it even less actual industry involvement than Microsoft did back in 2001, when at least it had some experience as a platform holder and publisher of Windows games. Amazon and Apple are even more distant, despite Amazon's recent gestures in the direction of becoming a games publisher and distributor.

For each of these companies, pitching a snack between meals isn't worthwhile; they have to somehow convince consumers that any streaming service they launch is really a meal in itself, a credible outright replacement for their Xbox, PlayStation or gaming PC.

That's a hell of a tough sell. The pared-back, filling-in-the-gaps nature of what Microsoft and Sony are gearing up to offer isn't down to a lack of ambition on the part of either company; it's a fairly realistic pitch for a streaming service given the combination of both technical and cultural obstacles standing in the way. There remain a lot of places, including plenty of wealthy first-world urban areas, where broadband and mobile speeds are not up to the task of consistently providing high-quality game streaming, or where data capping and charging remain in place.

Google teased its presence at GDC with an enigmatic image sent to the press

Google teased its presence at GDC with an enigmatic image sent to the press

5G and ongoing fibre optic rollouts will help a lot, but we're still quite a long way from being able to say mission accomplished on this front. Even 4K streaming of movies -- a far less technically demanding task -- is a pipe dream for many consumers at peak times such as evenings and weekends. Meanwhile, a large and influential part of the culture around games is focused on latency and visual quality (let alone the whole question of ownership itself). There is probably a pretty solid audience out there that's willing to consider streaming as a nice-to-have extra for long train rides or hooking up to the hotel TV on business trips. But willing to ditch their console or PC? not so much.

"The lack of noise around the commercial impact of PS Now tells you a lot about how consumers have responded"

That creates a fairly tricky economic environment for streaming services. PS Now is presently offered as a $19.99 per month (or, more reasonably, $99.99 per year, though that whopping 58.3% discount tells you a fair bit about how much the service actually gets used by annual subscribers) subscription independently of PS Plus, which has almost certainly not been working very well for the service. PS Now was specced out initially as a full-bore Netflix-for-games offering, and the sheer lack of noise around its commercial impact thus far tells you a lot about how consumers have responded.

We don't know how xCloud will be priced, but it's far more likely to end up being an ancillary service that plugs into an Xbox Game Pass subscription -- a few bucks extra a month to be able to stream the games you can already access through Game Pass to your thin client (mobile phone, tablet etc.) as well. Sony at some point is also going to have to rationalise the existence of PS Plus and PS Now as separate services, and will likely end up with a broadly similar model.

Without the ability to do that -- with no existing subscription to tag a hypothetical Google or Amazon streaming service into -- the new entrants into the market are going to face an uphill struggle. Their entries will presumably end up looking very much like PS Now in terms of pricing and offering, even as Sony -- if it's serious about making this work -- will be looking at radically reworking PS Now's value proposition by the time the PlayStation 5 rolls around.

The new services will need to convince rightly dubious consumers of the technical strengths of the streaming approach and get them to fork over a significant fee. Existing players can make far less extravagant technical claims and market streaming as a cheap add-on to existing game subscription fees, not dissimilar to how Netflix charges a few extra dollars each month to use more devices or stream 4K content.

This isn't to say that the entry of players like Google into the market won't shake things up -- it will, of course -- but even if what it announces at GDC next week really is an all-singing, all-dancing streaming service for games, Microsoft and Sony will still be the players to beat. It's worth bearing in mind what happened the last time a huge industry giant set its eyes on this industry; the Xbox may have its admirers but it was hardly a commercial knockout, with Microsoft only really finding its feet with a second generation of hardware.

Neither Google nor Amazon is likely to do much better with their first attempts, especially when their likely point of entry is a technology that still has a long way to go in convincing consumers of its value.

Related stories

Stadia director of product: "Eventually all games will be in the cloud and we'll feel great"

Andrey Doronichev responds to concerns about service discontinuation, ISPs, features in Reddit AMA

By Rebekah Valentine

Blizzard unleashes Google's DeepMind AI against European Starcraft II players

Experimental AlphaStar agent will play anonymously, but only against gamers who opt into test matches

By James Batchelor

Latest comments (10)

Nick Parker Consultant 4 months ago
We must keep our eyes on the prize; the long term streaming competitive landscape post 2022 will be more attritional. Google and Amazon should not be underestimated as new hardware entrants. They have robust brands and deep experience in understanding consumers and the fact that they would be new to market like Xbox was, as you mention in the penultimate paragraph, that does not translate into early iteration failure; the PlayStation was launched in 1995, a new disruptive hardware market experience from another strong brand of the time, Sony, and we all know how that turned out.
What will be interesting, as well, will be the games catalogues each will offer; Amazon and Google would invest in their own first party range but, being more generic e-commerce platforms, will want to major on PC, PS and Xbox titles to stream if Sony and Microsoft open up their streamed catalogues as they do for physical product. That leads onto exclusivity of streamed games - third parties don't care where their games are sold/streamed, they only care about maximising eyeballs and access for their games but will Sony and Microsoft be more selective as they have to defend their own more profitable streaming services?
0Sign inorRegisterto rate and reply
Jeff Kleist Writer, Marketing, Licensing 4 months ago
This is what I will say, streaming games to my bedroom TV on my wired gigabit house network has noticeable lag.

Itís fine for strategy games, point and click adventure games, telltale games, casual and card games, but I never play Apex legends on it
2Sign inorRegisterto rate and reply
Bob Johnson Studying graphics design, Northern Arizona University4 months ago
Not going to pan out.

They can't overcome latency. We all use the internet daily. And can see the latency of our connections for our selves. Imagine hooking your input device and screen to a pc or console across that same internet connection. And yet we're supposed to believe we'll get a great experience out of that?

Not going to happen.

What I'm wondering is why the recent impetus to getting these services off the ground? OnLive failed not too long ago.

Is it because they just can't ignore the fact that there are billions of mobile users and only 100 million console users? And if there is at all a chance that they can sell console games to some of those billions of mobile users that they need to throw money into it even if people say it's a not a great experience?


My only other guess is Net Neutrality has something to do with it. That eliminated that and to me that means an internet provider can provide QoS traffic aka your provider could prioritize traffic for more $$$$ aka reduce latency of something like a game streaming service. But that's just a guess based on latency being a problem and net neutrality being struck down and nothing more.
1Sign inorRegisterto rate and reply
Show all comments (10)
Klaus Preisinger Freelance Writing 4 months ago
Take an average hit game in Germany. It has 1 Million players and they want to play in 1080p. That is 8MBit in Bandwith to make it reasonably unblocky. Even then, maybe disable a few particle effects. 8MBit also sounds reasonable for the consumer, since 8MBit is something most people have and are used to get when watching Netflix. 1 Million watching? Not a problem.

Now all we need is the same 7 Terabit of bandwidth to serve this one million people. In contrast to Netflix, we cannot cheat our way around this with proxies and mirrored servers running from inside ISP networks. Each gaming stream is unique. For this very reason we now officially have a problem.

DECIX is the biggest internet exchange point in the world. It claims to be able to handle 45TBit per second. Our brand new business model just assumes we get 15% of that and more in the future, because 1 Million people clearly isn't the limit. Never mind all the serverfarms you need to collectively output those 7TBit worth of streams, plus all the gaming compatible server hardware there.

Actually, let's do mind those. Your million player hit streaming game, effectively just put out an order in the neighborhood of 56 typical datacenters. Each serving around 17.000 people. One may be inclined to shrug it off and point towards new super-high speed fiber technologies, but one would be wrong. You see, each of these fine people playing the game also causes 150W of power draw computing the game, which your average Netflix stream does not. That would be 150 Million Watts, not counting the air conditioning on top of that. Mind you that from here on, when I talk about scaling things up beyond 1 Million players for our Datacenter, the mission is no longer about securing a better fiber connection, the mission is about building a nuclear power plant to keep up with the power draw.

At this point, you can pick up the phone, call Germany's biggest Datacenter Hetzner Falkenstein and kindly ask them to hand over their entire capacity and then build five more of these datacenters, so they can handle your game for 1 Million players. Fun fact, I already base this on their stats about their power draw and backup capacities and not on their fiber connection, where I magnanimously assume an adequate upgrade to manifest because tech industry.

Why stop there though? Naturally, you want an Apex Legend sized hit with 50 million players. Put all their PS4s together and you would have a cube with roughly 454ft length. That is where the path to streaming success will lead you. A datacenter, the size of a Pyramid, only as a cube and powered by a nuclear power plant twice the size of Palo Verde Arizona; plus air-conditioning, never forget. Maybe some things work better when remaining highly decentralized. Like, one in each living room sort of decentralized.

If only it were just the latency. Although ironically, the nuclear power plant will have some. So when people get home and spike the demand, a correct loading screen should actually say "lowering rods, powering up nuclear power plant, starting game". If that is not marketing gold, nothing is.
2Sign inorRegisterto rate and reply
Jeff Kleist Writer, Marketing, Licensing 4 months ago
@Klaus

If I recall correctly, each blade of PSNow is actually 4 Complete PS3s in much smaller form factors. So not quite the epic pyramid :) to be honest Iíd be way more concerned about the heat and power. Netflix serving up movies is patty cake by comparison. Even Iíd they have brand new highly efficient silicon youíre talking 100w per user easy.

Streaming is mostly chasing the mobile users and the millennials that donít have TVs. I think youíre going to see it used for rental or demos of new games as well. Ubisoft is certainly experiments with that in Japan
0Sign inorRegisterto rate and reply
Klaus Preisinger Freelance Writing 4 months ago
@Jeff
4 PS3 units in a small form factor in a datacenter? Who are we competing here with? Flee markets and thrift stores? Mobile gamers seems to be happy with the visuals they get on their Skinner boxes. Who is left to be served? Core market gamers too cheap to buy a console and interested enough to subscribe to a service that basically racks up the same costs, only over time? When talking about the advantage of monthly pay, do those streaming services ever realize that every electronic store in the world accepts monthly payment schemes, making the financial advantage of the "cloud hardware rental service" moot? Are we just too courteous to call them out on that?

So yes, delivering 7TB of data per second is a solved problem. Computing 7TB/s of data herein lies the problem. The per user footprint of Netflix is maybe +3W (all-inclusive) compared to idle at best; file-serving is not computing. The power footprint of a gaming stream is easily +150W (just the game), plus air-conditioning, which can get as high as another +100W due to current datacenters preferring packing density (hence the file-serving specialization), over swallowing up space (typical computation adaption, not just when super-computing, but also when opening your PC case on a hot day). So your game streaming data-center has one of two overheads, cooling costs (cmp. MMO server farm), or dissipation space.

Even if silicon gets more efficient, this will be used to make better graphics, not get the footprint down. Look no further than PC and console graphics, which developed in leaps and bounds from one generation to the next, mainly because it was a race to hit 300W TDP.

Imo, it is a race for the patents, that is the reason for the goldrush. It is a technology that makes sense, if you do not think too hard about it and as a result it seems reasonable to invest in the race for patents surrounding it. But due to the size the video game market has grown, the scaling approaches absurd proportions.

Then again, we live in a world where somebody once walked the streets, saw all the cars standing outside day in day out, slowly rusting exposed to nature and thought: 'you know what we should add to that? A chemical that starts a fire when it comes into contact with water'.
0Sign inorRegisterto rate and reply
Bob Johnson Studying graphics design, Northern Arizona University4 months ago
there wouldn't really be an increase in power used because you're just trading, for example, a ps4 in the home for one in the data center.


I don't think bandwidth is a problem. A lot of people in the US now have access to gigabit internet connections even if they don't need one. And gigabit would be overkill.

I just think that latency is the problem and what separates streaming games from shows on Netflix. Netflix can buffer. Games can't. Yes they can build more data centers in more locations to cut down on latency. If you live next to a server latency much less an issue.

And yeah agree that the cost of streaming games is not likely to be cheaper over the long run. Opposite. Renting stuff is more expensive than owning as a rule of thumb.

I'd like to know more about how well received the streaming of Resident Evil 7 to the Switch experience was in Japan. There was a lot of news about it coming to Japan. Much harder to find any information on 1st hand experience with it.

Edited 1 times. Last edit by Bob Johnson on 16th March 2019 4:01pm

0Sign inorRegisterto rate and reply
Alex Barnfield Lead Engineer, 17-BIT4 months ago
There's real room for games to take cues from Netflix. I like the subscription model; I like an absence of physical media, I like not having to download and install my movie before I start watching it. But games aren't movies; it feels like there should be ways to accomplish this without the latency issues introduced by just streaming the final image.

Edited 2 times. Last edit by Alex Barnfield on 18th March 2019 9:35pm

0Sign inorRegisterto rate and reply
Jeff Kleist Writer, Marketing, Licensing 4 months ago
@klaus

The millennials in many cases donít own a TV, and they much prefer subscription models. Donít forget, this is a generation raised on piracy, the only way to get them to pay for something is to make it easier, and close enough to piracy they donít mind

Theyíre also saddled with crippling student loans and terrible job prospects, but thatís another issue.

So the lack of ownership and the sub model donít bother them. However If the experience sucks, that will. Itís why Iím. A big believer in gamepass. And yes, I wait with glee for the day the Internet is out and they have nothing. :) Iíve been ďoldĒ in these regards since I was very young :) I also agree with you about tThe patents. Just as with Kinect the technology is applicable to a wide range of applications.

Edited 1 times. Last edit by Jeff Kleist on 17th March 2019 1:11pm

0Sign inorRegisterto rate and reply
Klaus Preisinger Freelance Writing 4 months ago
@Jeff
Right now:
$400 console, $60 per year Xbox Gold, $80 per year Game Pass

Future:
2-Year MS Contract $27 per month all inclusive.

It worked wonders for smartphones within that very same demographic, creating one of the most overpriced category of devices ever conceived. In my opinion, that will be the change. The entire IT industry is already addicted to this type of pricing, even if it means weird risks are emerging in places where the actual costs are fronted. The rest will be added subscriptions for other publishers, normal $60 games, f2p and microtransactions. And if you do not have a display device, then that is an easy fix as well.

As far as consumers readjusting their preference in brands due to economic pressure is concerned, expect a massacre. Because, in a nutshell, China has long since noticed there is no use in fighting U.S political power head on, you have to fight U.S. economic power coming from corporations and the resulting lobbying first. That is how the Huawai situation got so ugly, that is why Northstream 2 got ugly. There is no need for politicians to shake hands, when an economic war is waged behind their back. The rise in protectionism is not a sign of wanting to go back to the past, it is a sign of being at the losing end of an economic war in the present; with no millennials to blame for it, btw.
0Sign inorRegisterto rate and reply

Sign in to contribute

Need an account? Register now.