Cloud gaming five years from now

Cloud gaming forecasts at one year are mostly noise (the press releases haven't dropped yet) and at ten years are mostly fiction (the technology curves are too uncertain). Five years out is the sweet spot — long enough that the slow-moving structural shifts dominate over quarterly product moves, short enough that we can ground predictions in technologies that already exist in early form today.

Here's what we think the cloud gaming landscape looks like in 2031.

By 2027–2028, every major cloud gaming service has migrated to AV1 as the default codec. The bitrate savings are too large to ignore and the hardware decode landscape has filled in. By 2031, AV1 is the legacy default the way H.264 is today.

What replaces it on the bleeding edge: the Alliance for Open Media's AV2, which is in spec work as of 2026 and should ship in production silicon around 2029. Expect another ~25% bitrate improvement at iso-quality, and another ~5 ms of encoder latency overhead. The latency budget tightens; the bitrate budget loosens.

Today we still talk about cloud gaming providers as global brands. By 2031 that framing is partially obsolete. GeForce Now wins in Europe, Xbox Cloud wins in North America, a regional Asia-Pacific player (we'd bet on Tencent's hosted offering, but Sony Asia is in the conversation) wins in much of APAC, and Latin America is still underserved.

Customers will subscribe based on where they live, the way they currently subscribe to cellular carriers — and most people will tolerate this without thinking it's strange. The era of 'one global cloud gaming brand wins' is already mostly over.

If you spend $800 on a gaming PC today, you can play most games at 1080p/60. In 2031, that same $800 instead buys you a four-year subscription to a high-end cloud service with no upgrade obligation. The economics tip definitively past the indifference point.

What survives at the consumer end: high-end enthusiast PCs (because the cloud has a latency floor you can't beat for esports), low-end laptops/Chromebooks (because they don't compete on local gaming anyway), and handhelds (which become primarily cloud devices with the local-rendering option as a fallback).

The mid-range desktop gaming PC, $700–$1,500 range, becomes a niche category by 2030 the way mid-range tablets did in 2020.

Microsoft's strategic direction is already clear: Xbox is becoming a service brand, not a hardware brand. By 2031 we'd be surprised if Xbox is still selling discrete consoles outside of a handheld form factor.

Sony is the more interesting question. PlayStation has the highest-margin first-party software in the industry tied to its hardware install base. They have the strongest incentive of any platform holder to keep consoles around. We think they will — and that PS Plus Premium grows from 'streaming option for the catalog' into 'streaming-first tier' that's actually competitive with GeForce Now by 2029.

Nintendo will continue doing exactly what Nintendo does, which probably does not include cloud streaming in any meaningful way.

Every analyst piece on cloud gaming includes a paragraph about consolidation. We don't buy it. The structural moats are too service-specific for M&A to be a clean fit:

NVIDIA isn't going to buy a competitor, because their entire product strategy is built on being the GPU supplier for everyone else. Microsoft and Sony are platform companies that own their first-party software; they're not buying each other. Amazon Luna keeps under-investing rather than divesting because cloud gaming is a strategic pixel for AWS rather than a stand-alone bet.

What we do expect: Boosteroid acquired or absorbed into a bigger player by 2028. That's a single regional service with limited expansion runway and a thin balance sheet relative to the others.

The single biggest change between today and 2031 that we don't think the industry is taking seriously enough: cloud gaming on TVs. Specifically Samsung, LG and Roku TVs that ship native cloud gaming apps and a paired controller in the box.

The TV-as-console category will matter more than the PC-via-cloud category by 2029. Casual gamers don't want a gaming PC and don't want a $500 console. They want to press the gaming button on the remote and play. Whoever owns the TV integration story wins the largest single demographic in gaming, which is non-enthusiasts who currently play almost nothing.