Wednesday, July 11, 2012

The only market for cloud gaming

In my previous blog i gave a short but I think solid analysis of why  so called "cloud gamin' cannot succeed in the markets  OnLive has been pursuing.  Interestingly enough, there is one market that OnLive's solution *might* just fit.  I blogged on this quite a long time ago, but I think it deserves some new attention.

The issues with OnLive is that, fundamentally, its a  more expensive way to get games to customers then digital download and produces inferior quality gameplay.  The expense is because it requires as much hardware at the data-center end to run a game as it would if it was local to the player PLUS it also requires the player's receiver hardware and the bandwidth charges for game data.  The impact on gameplay quality is due to the compression they must use to both control bandwidth costs and cross the last mile to the user's home, as well as the inherent latencies of internet communication.

There is one, and only one, market I know of where these negatives can be mitigated and where OnLive might offer a compelling offset to all these negatives.  That is over cable TV service,

There are a prefect-storm of reasons why this market makes sense.  The cable operators own the connection from their data center to their customers.  That means they can allocate as much bandwidth as they feel they can afford to give to this service. They can also prioritize the packet communication and give preference to the game packets thereby minimizing latencies.

The customers for this service are not those users that have a game PC at home.  It will always be better for them to play locally.  But it is a service they could offer through the cable box to those without such a computer.  Cable boxes are built to "throw away" and be as cheap as possible.  The cable companies assumption has to be that they wont get many of them back and so their cost has to be amortized rapidly in the service cost.  A more expensive box means more expensive service which is bad for business.  For this reason the cable companies attempt to hold as  much of the smarts as they can of the system back at the data center.  Data center equipment can be used and reused until it is obsolete.

In this environment it COULD make sense to operate an Onlive back-end in their local data center (what they tellingly call the "head-end").  It still means a much faster obsolescence curve then the cable TV companies like for head-end equipment, but if they could get enough monthly for the game service it just might pay off.

3 comments:

Unknown said...
This comment has been removed by the author.
Unknown said...

Some of the cost could be offset by selling the same service to more people than you have equipment for, each user needs a display box, but you only need enough server side hardware to cope with the number of simultaneous users. I don't know what ratio is needed to make the cost less than normal, or what ratio is reasonable to assume though.

Unknown said...

The flip side of this, though, is that you need your service to be always available. If people can't play when they want or have to wait on lines, you lsoe customers.

Since PCs are notoriously finicky, you actually have to have enough hardware to cover your peak load PLUS enough spare capacity to cover machines that fail. Add to that the fact that each"head-end" of the cable system serves a single neighborhood, and you could actually need MORE equipment then just giving everybody 1 PC.