An interview with Bryan Hill, Interxion’s Director of Marketing & Business Development

Editor’s Note: Streaming video changed the entire equation of bandwidth demand in the submarine cable industry.  In the course of just a few years, the industry went from a bandwidth glut to racing to meet the insatiable consumer demand for video over the Internet. 

As new cables with previously unheard of design capacities now enter service, a frequent question is, “Will video alone fill them up?”  We have pointed to other upcoming high-bandwidth applications, such as the Internet of Things and virtualization.  But one application that has not received enough attention is cloud gaming. 

To better understand how cloud gaming could impact bandwidth demand and the submarine cable industry, SubCableWorld reached out to Bryan Hill, Interxion’s Director of Marketing & Business Development, Digital Media.  He brings an insider’s perspective to this promising market that we in the submarine cable industry need to be aware of.  His comments follow:

Mr. Hill: Let me start with a little background.  The gaming industry has evolved through several different eras.  Most have involved a heavyweight hardware device, such as a console or a high-end PC where the processing for the gaming was done.  These are expensive.  A console can cost as much as $600, while a gaming PC can be as high as $3,000.  

Then there is cloud gaming.  Cloud gaming has really been around since about 2014.  What cloud gaming allows you to do is to play high-end, complex, high-production games with no processing required on the customer’s device.  So you no longer need this heavy computing capability residing in a console or high-end gaming PC.  You don’t have to spend $3,000 on a PC or $600 on a console anymore.  It allows you to play on any device whatsoever, as long as the device has the ability to render inputs and video.  That provides complete portability. 

One of the implications from this is that because the processing is done within the cloud gaming platform, it requires high-density computing, usually collocated with the network.  You need these heavy computer networking clusters that require lots of connectivity and they need to be close to the customer. 

Latency is absolutely one of the critical factors here.  It depends on the game that the customer is playing, but in order to reflect the user experience that a gamer expects having played at the high-end PC level, you probably need a maximum of between 20-30 ms round trip latency from the cluster to the end user.  The reason for that is that no one wants to play a fast-paced game that is sluggish.  If the game plays in slow motion because of high latency, gamers will lose the quality of the experience. 

Another issue is network stability, and we’re talking zero packet loss here.  From a game-playing perspective, network stability is even more important than latency.  Imagine a game that is being played for hours – heavy gamers can play up to seven hours in a session.  There’s a huge amount of time and a tremendous commitment by a very engaged audience.  So imagine that you’re playing a game where you’re fighting aliens and the cluster is doing all of the processing.  You come around the corner and see an alien and you’re going to shoot it.  And you press the button.  That impulse gets transmitted from the controller, to the router, to the exchange and to the carrier network.  So what happens if that input does not get registered – it gets lost somewhere along the way?  Or what happens when you turn around the corner and the clusters are rendering the video to you and a packet gets missed and you can’t see part of the target as it gets rendered?  It ruins the gaming experience and it is that experience that is so important to the customer. 

So network quality, network stability, choice of networks, flexibility to make sure that within that latency range every packet gets delivered, the use of multiple providers, multiple peers, public peering – that’s all very important.  The resiliency of networking, making sure you have multiple routes to the consumer so that there is no packet loss whatsoever, is critical. 

It’s the very early days for the evolution this sector, there’s no doubt about it, but cloud gaming has some momentum behind it.  Last year, Google announced Stadia, Microsoft announced xCloud and EA announced Project Atlas.  Cloud gaming is coming and the big players are behind it. 

So let’s look at potential business models.  Let’s take the subscription model.  You’re not going to continue to subscribe to a service where you fear that your gameplay isn’t either reflecting your existing gameplay that you would have played on a console or high-end PC, or you fear that you could be playing for hours and then suddenly you lose, through no fault of your own.  Those elements are critical.  So with cloud gaming you’ll pay more. While the hardware requirements will be far less with cloud gaming (you will pay much less upfront), you will end up paying more for your gaming.  You could pay $600 for a high-end console that will last you about three years. Then a few games at $60 apiece -- that’s around $20 per month. In the cloud, however, you will use your old hardware -- but you’re going to pay up to $40 per month in subscriptions for both games and data.

But in addition, you’ll play more. With the cloud powering things, you can go fully mobile. Particularly with the advent of 5G, phones and tablets will suddenly become miniature, portable consoles with the ability to play the same games you once stayed home to play. Even at home, you can use that older TV you no longer use -- with the cloud it becomes a game-worthy console. And DIYers can bet we’ll see you using kits to build your own consoles, too.

So there are a couple of defining factors about where cloud gaming service providers will deploy cloud gaming platforms.  What do you need when you deploy these platforms?  You need a sufficient number of end customers where there is Fiber-to-the-Home (FTTH), because you’re talking about, realistically, a connection of at least 15 Mbps.  Ideally, you’ll want more.  You could probably get a good gameplay at 15 Mbps, but you’ll probably want more like 25 Mbps.  It needs to be a big enough gaming territory in terms of players and there needs to be sufficient FTTH.  In some parts of the world, that’s already established.  In other parts they still have a way to go.  Cloud gaming is a long game, forgive the pun.  It requires proximity to end users, relatively low latency, high degrees of network stability with zero packet loss, dense connectivity models, and a territory with a sufficient number of gamers with FTTH or other high-bandwidth access.  Of course, 5G is going to be a real game-changer in this regard. 

Interxion works with all types of gaming companies.  Whatever the gaming service is, we take care of everything that is done inside the data center.  We have 52 data centers in 13 cities and 11 countries across Europe, with over 700 connectivity providers.  In order to underpin the services, to provide the quality of services that the end users require, you need Tier-III+ data centers in multiple locations.  You need dense connectivity providers, -- CDNs, carriers, IXs, ISPs.  Low latency, flexible, diverse, network connections to the end users so that they get the highest quality of service. 

We see submarine cables playing an important part within the evolution of cloud gaming.  We see the increasing amount of capacity these cables can deliver, from the number of fiber pairs to the way transmission technology has improved.  But from the cloud gaming perspective, we also see submarine cables opening up new opportunities for cloud gaming companies to deploy in multiple territories.  For example, from Marseille, you can just hop on the cables to go to Italy and Spain, to Africa, the Middle East and South and Southeast Asia. 

Submarine cables also offer low-latency options.  For gaming, this comes back to the need for high levels of network stability and having more options.  You can have a quick route to enter or test new markets, but you can also have diversity by using both terrestrial and submarine cables as part of your cluster-to-end-user architecture.  So we see submarine cables definitely being increasingly part of the mix.  It’s happening now and will only increase in the next couple of years.  We see submarine cables as an important component part of that flexible architecture for gaming companies. 

The other part of this, and this applies to submarine cables, is that you’re going to need a ton more data.  Data for gaming is not like video data.  Streaming benefits from data compression -- but compression is hard and doing it well takes time.  When Netflix does compression, they do it in advance. Yet when it comes to the two-way live action that gaming requires (e.g., when that alien is coming and you need to shoot it now), even a second delay is not okay. So, data will be at a premium. Expect new ISPs and data providers to roll out higher tiers of data plans for gamers.

In cloud gaming, everything is done at the cluster within the data center where the connectivity providers’ equipment resides.  These games are high production.  Gamers really want at least HD and preferably 4K.  We’re talking about tens of Gigs of traffic per user per hour.  Even with HD, one hour of gameplay can generate something like 90 Gigs.  Our connectivity customers are very well aware of the needs but they’re also aware of how much valuable traffic there is. 

There are some similarities between cloud gaming evolution and the evolution of video and the OTTs.  When Netflix first came out, there was a lot of hype that this was the wave of the future, but then there were a lot of articles that said nobody would cut their cables.  It was never going to happen.  In the same way that people are saying nobody will give away their high-end PC or gaming console for cloud gaming.  It’s going to take years, for sure.  New people are going to game like they’ve never gamed before.  People will game even more because gamers will love the portability that comes from cloud gaming.  It’s going to come down to what the OTTs came down to.  It’s going to come down to portability of content, quality of service, which is even more critical when it comes to gaming than video.  And it’s going to come down to finding the right business model.

Bryan Hill

Bryan Hill