Quick Thoughts: Is Cloud Gaming the 5G Killer App?
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai
FOR IMPORTANT DISCLOSURES 203.901.1633 /.901.1634
psagawa@ / email@example.com
November 26, 2019
Quick Thoughts: Is Cloud Gaming the 5G Killer App?
- Carriers need 5G to cope with explosive data growth in developed, competitive markets. Spectrum flexibility, increased cell capacity, and low-cost small cell expansion make it a no brainer. Residential broadband replacement and IoT for industrial applications are the new revenue opportunities.
- Beyond congestion relief, there is no obvious 5G killer app for many mobile subs. Faster connections and lower latency are unnecessary for most consumer apps, including streaming video.
- Cloud gaming is the exception. Gamers are extremely latency sensitive. The 10x reduction in latency from 5G to 4G facilitates streaming, eliminating processing and storage bottlenecks to more serious gaming on mobile platforms and enables multiplayer online games.
- 5G should also offer 75-80% better latency than fixed residential broadband, improving the performance of massive multiplayer games (e.g. Fortnite) on any platform. This could greatly increase the likelihood of households dropping broadband in favor of 5G.
- Gaming is a $150B market growing at >10% a year, serving hardcore 500M consumers worldwide (and more than 2B consumers including casual mobile gamers).
Carriers ARE going to implement 5G and will almost certainly end up spending more capex to do it than they are currently letting on. We have written about this (5G: Rising Global Carrier Competition to Drive Capex, The Three Phases of 5G: Coverage, Density and Applications, 5G: Why TMUS will Win). The reasons for this are rapidly growing demand for capacity, relative network expansion costs and competition. 5G offers as much as 10 times the capacity per cell site vs. 4G LTE and can tolerate much smaller individual cells at lower costs per cell site (achieved through implementing more of the processing elements in the cloud). While the initial build out of the new generational standard will be expensive, the cost of increasing network capacity to eliminate congestion will be MUCH lower vs. 4G, giving carriers ample incentive to move heavy usage customers to 5G as quickly as practical. If your rival beats you to 5G, they can offer these customers lower prices, higher (or no) usage limits, and much better network performance (Exhibit 1). In the US, that rival will be T-Mobile – presuming the merger comes through, and perhaps, even if it doesn’t.
Carriers will have ample incentive to move their customers over to their more capacious 5G networks, even though the benefits of the technology for consumers are very modest. The marketing blather goes on about faster connection speeds but fail to make much of a case for why anyone would really want to pay for that. 4G LTE download speeds average better than 35 Mbps in most major markets, more than enough to handle 4K video and even enough for 8K video, once the new Virtual Video Codec (VVC) compression standard
Exh 1: SSR Criteria for 5G improvements in critical functions over 4G
comes into wide adoption over the next couple of years. Note that only about 1% of the installed base of TVs support 8K video and that only a handful of smartphones even support it. If a wireless user is complaining about connection speeds, generally, they are complaining about congestion spots or slow response from an app server, not the capability of the network. 5G will help the carrier deal with the first problem, but the second will still persist with the new generation standard. In this context, Verizon’s millimeter wave implementation of 5G, which ignores broad coverage to offer extraordinary speeds (1Gbps +) in spotty coverage around small cell hotspots, is overkill.
The big benefit to wireless customers is unlikely to be connection speed, but rather, network latency. The 5G standard cuts out intermediate network processing points to cut the loop time between a server and the user by a factor of 10, from an average of 40ms for 4G LTE to less than 4ms. This is also dramatically faster than wired residential broadband networks, which average 15 to 30ms depending on the network type. For many applications, latency is relatively unimportant – the 30ms is about two frames of a video stream. The big exception is gaming.
We recently published an overview of electronic gaming (Game Streaming: Not Everyone Can Win) in the context of the pending launch of several cloud-based streaming game platforms. In that piece, we noted that the biggest hurdle to the adoption of streaming gaming was lag – fast games like fighting games or shooting games are compromised by any apparent lag between the player’s input and the action on the screen. At the same time, MMO (massive multiplayer online) games, like the massively popular Fortnite, can only be
Exh 2: Historical Trend in Gaming Sales by Category, 1970 – 2018
delivered from a cloud platform and are forced to compromise game play in order to facilitate large, geographically dispersed groups. The difference between 40ms or 30ms or even 15ms and 4ms is noticeable and significant. 5G could induce developers to create faster games for the cloud, drawing casual users to play much more sophisticated games, and spurring serious gamers to consider the cloud.
Gaming is already a $150B global industry – bigger than newspapers, periodicals, books, recorded music or movies while quickly gaining on the $230B Pay TV business (Exhibit 2). Anything that can attract gamers and their money is an interesting opportunity. We believe that high performance gaming may be the killer app for the consumer adoption of 5G, particularly in gaming heavy markets like China, Korea, Japan and the US (Exhibit 3). This could induce younger demographics to switch carries in search of better 5G coverage – hotspot-only buildouts like Verizon’s won’t cut it – and to cut the cord on their wired residential broadband in favor of low latency 5G mobile service. Note that we are skeptical of “fixed wireless” 5G residential service via millimeter wave spectrum, which requires an external antenna to bring the signal indoors, is vulnerable to disruption by weather, and cannot follow the subscriber out of their home. We believe that the real challenge to cable modems, fiber-to-the home and DSL will be from mobile 5G.
Who is this good for? Wireless carriers with the spectrum and cash to be first and biggest with 5G – in the US, that’d be T-Mobile. Cloud gaming platforms – our complete take is in our gaming piece (Game Streaming: Not Everyone Can Win) but Microsoft is at the top of the list. Independent game developers –
Exh 3: Key Elements Driving a Paradigm Shift in Gaming
the JJ Abrams of gaming is out there someplace champing at the bit to sell his/her work across 7-8 fiercely competitive platforms. E-sports streamers (Amazon and Google) and other businesses tied to the popularity of video games (watch out for privately held Valve and Discord). And of course, the players!
Who is this bad for? Spectrum or balance sheet challenged carriers – that is, Verizon and AT&T. Fixed line carriers – the entire cable industry plus Verizon and AT&T again. Game publishers, unless they have enough intellectual property in house to play the “Disney +” role in the game streaming wars to come. (N.B. – the downside for publishers is that creators will have a LOT more leverage).