Monday, 2 July 2012

Why game video streaming is too expensive

The news about GaiKai being acquired by Sony reminded me I needed to write up this post.

Don't expect to be playing lots of cheap games on GaiKai after the Sony acquisition. The economics of video streaming of games are tough.

Whilst you can buy cheap time of a virtual PC on a network from a myriad of cloud companies- at Google I/O their Compute Engine was announced to offer a basic virtualised CPU at 14.5c per hour.

But, this cost needs to be qualified when thinking about a game streaming service.

Firstly, a virtualised CPU is shared between lots of customers processes. Whilst you get an average performance you paid for, peak performance varies greatly. That makes for occasional stuttering that doesn't matter in a business application, but would be bad in a FPS game.

Next up is the impact of GPU performance. These services render the 3d on the server. That means that you need a GPU in the server to do so, and most commodity servers just don't have GPU's with any umph. Fortunately, Nvidia's announcement of a virtualisable GPU for servers will help, but as yet the costs and loadings aren't visible.

It's worth remembering too that a cloud hosted game demands more of the CPU/GPU than the same game does when running on a home PC- as it must compress the video from the game with very low latency so it's small enough to send to the consumer.

So having their own servers (or deals to put GPU's into 3rd party racks) isn't too bad?

Well it would be, but there's a compounding issue: the speed of light. Unlike ordinary cloud services, games need fast response times, too fast for servers to be located far from their users. That means that whoever owns GaiKai needs to equip enough servers close to their customers to cover PEAK loads.

So, round Christmas, where perhaps 50% of PS3's might be operating in peak hours, there'd need to be the capacity to service them from servers with GPU's close to the customer.

It's very hard to get to a final answer- as lots depends on the deals you can do and the scale you can reach, but in my opinion it'll be well above $1/hr to provision the service at 720p, and maybe twice that again if you want to go to 1080p.

That has implications- if the average AAA title has 20 hrs of gameplay then that's a big dent in the game economics, unless you believe that the consumer will be happy to pay an extra $20/game to save on buying a new console?



Friday, 29 June 2012

Game downloads getting slower

Moores Law gives us an expectation that the computing power in our consoles, PC's and mobile devices will double every couple of years or so. Naturally game developers want to use all that power to give us the best possible game experience- so they use that new memory to the best of their ability.

Unfortunately, networks don't get faster at Moores Law, for example the UK's average internet speed has taken around 6 years to double.

Roger Walkden suggested we take a look at a sample game to see how each generation has changed the download time. Here are the results:


Whilst there's a bit of jumping up-and-down here- there certainly isn't a trend to shorter download times here!

How long will my game take to download?

Time
Every time we want to download a game from most of the online sites we face an unknown time before it will be ready. Many sites hide the size of a game until you have committed to the purchase, and so we have to google the size first.

We thought it would be helpful to provide a handy tool to calculate the download time. Just put in the speed of your connection and the size of the game...

Thursday, 21 June 2012

Console quality graphics on tablets?

IThere's lots of speculation over whether tablets and mobile devices can replace consoles.Whilst that may prove to be part of the story, we've been thinking about how that could arise.

One neglected part of the story, masked by the announcements of upcoming graphics processors (GPU's) for mobile devices, is that there are physical constraints associated with a device being handheld and battery powered when it comes to the performance of the GPU.

In particular in a console or desktop power is cheap and there is space for cooling fans. The result is that fixed gaming GPU's consume 50W-500W+.  Above (courtesy of TechRepublic's excellent teardown) is a recent XBOX 360 cooling arrangement- note the big piece of aluminium and the fan to take away all that power.

So, not surprisingly this doesn't fit into a svelte tablet or that it would flatten the battery in minutes.

In practice, my belief is that a tablet can afford to dissipate a maximum of 5W, or about 1/20th of what a console can afford. (The whole Ipad 3 consumes about 7.3W peak).

Professor Moore provided us with a "law" of thinking about how power consumption changes over time as semiconductors develop which has proved reliable over decades. Roughly, consumption halves every 2 years.

So, that means that at 1/20th of the allowable power dissipation, tablets will be 4.3 iterations of Moore's law behind what console technology can do at any one time. That equates at historical rates that have been consistent since the 80's at around 8.5 years.

Therefore we should not be surprised that today's latest tablets, launching 7 years after the XBOX 360, would be fairly close on graphical performance.

However the industry is not standing still. The next generation consoles are likely to have graphics of a much higher quality- exemplified beautifully like this: https://www.youtube.com/watch?v=UVX0OUO9ptU. These wont be able to be rendered on portable devices for perhaps another 7 years.

The exception to this is if the rendering is done on the server and streamed to the players over the internet in the style of OnLive, GaiKai and their ilk, but the challenges there are significant too... I will talk about that in a future post.

So, any claim that consoles will die off is predicated on (as far as I can see) only two scenarios:

1. Some unanticipated step-change in games techniques means that the effective speed of improvement is much faster than expected (although consoles would move then too!)

Or

2. Users pick convenience and portability over graphical performance and choose not to move to the new generation of hardware.

My view is that we'll find that both platforms co-exist, and that the most graphically-intense experience will be limited to consoles and PC's for foreseeable future.





Stories on mobile devices replacing consoles:
http://www.pspgamesthemes.com/psp-news/mobile-devices-could-replace-consoles-says-carmack.html
http://www.itproportal.com/2012/06/20/why-your-next-gaming-console-could-be-smartphone/
http://www.tomshardware.com/reviews/android-ios-game-developer,3169-8.html
http://androidandme.com/2011/08/news/qualcomm-says-dedicated-gaming-consoles-will-be-replaced-by-snapdragon/
http://www.popsci.com/gadgets/article/2012-04/pocket-console

Monday, 18 June 2012

In case anyone doubted it- users hate long game downloads

We asked a bunch of Steam users what they think of download times. This was their response:

Do gamers have faster internet connections?

There's a body of opinion that live HD video streaming over the internet is the solution to the game industry's pressures. We took a look to see how that option lies on today's connections.

We compared the speeds reported by Akamai's "State of the Internet" report with Steam's dashboard.

These are the results:



Our conclusion would be that there are few significant differences.

Most importantly, however you measure it, ~75% of internet connections are sub 5M bps. That's harsh news for those who think video streaming is the answer for HD gaming.