“The technology works, but the infrastructure doesn’t”: The Mobile Mavens on the future of cloud gaming

• “Cloud gaming has always had a slightly awkward problem: it is an impressive solution looking for the right problem to solve.” – Stuart De Ville.
• “We’re trading one bottleneck for another: limiting streaming bandwidth and server capacity instead of making games themselves more accessible.” – Romy Halfweeg.

Stay Informed

Get Industry News In Your Inbox…

Sign Up Today

Cloud gaming has long aimed to change how we play by offering access to games without expensive hardware and letting players move between devices more easily. But despite continued technological development, it has yet to become as widely used as some expected.

Questions remain about the quality of the experience and whether the economics of streaming games can work at scale. However, rising hardware costs and future improvements in cloud infrastructure could create new opportunities for the model.

So where does cloud gaming stand today, and what does its future look like? We asked our Mobile Mavens for their thoughts.

In casual, it’s really not needed. It is way more efficient to download an executable and have it run locally. The processing power is rarely a bottleneck. Antstream, for example, could send all the data to locally execute one of their retro games within the first few seconds of a stream and likely require less system resources to do it.

In more hardcore and high-fidelity games, like AAA, where local processing requires big hardware outlay, it makes a lot more sense. Have people rent time on some shared remote hardware rather than owning something that sits idle 90%+ of the day.

But these kinds of games want the best possible visuals and the lowest possible lag. And those are two things that streamed cloud gaming sucks at. The lack of video buffering introduced by the low-latency need means that video bandwidth is often degraded and, worse, choppy.

Even in my own home, on the same Wi-Fi router, in the same room, Steam Remote Play from my gaming PC to my laptop is unplayable. I hear on wired connections this is much better, but that seems to defeat the point.

So unless there is a massive improvement in networking technology where speed and packet loss greatly improve, I don’t see streaming being a realistic competitor to locally processed games for enthusiasts.

I suspect that Microsoft have found that there are some outliers in their service who care less about the fidelity stuff and are playing exclusively via cloud. And as processing time, especially via GPUs, becomes increasingly sought after, I think a decision was made to begin limiting the service and freeing up those Tensor Cores to hallucinate some slop.

I think cloud gaming is still on track to become mainstream, but it depends on what “mainstream” means. As a way to occasionally play high-end games on a device you already own, it’s already working. As a full replacement for a gaming PC or console, it’s not there yet, and I don’t think it will be until connectivity is solved everywhere, not just at home.

I’ve been using GeForce NOW for a long time, and when it works, it’s genuinely one of the best gaming experiences available. It works brilliantly with my Steam and Epic libraries, and at home on a strong connection, it could replace a gaming PC for a lot of people.

Ironically, it’s weakest exactly where it should be most useful: away from home. I travel a lot and play a lot of Diablo IV, and hotel Wi-Fi is completely unpredictable, often enough that streaming becomes unplayable.

I actually bought a Switch 2 partly just so I could keep playing Diablo IV on the road without relying on a connection. That probably sums up the core problem better than anything else: the technology works, but the infrastructure around it doesn’t work consistently enough yet.

The economics is where it gets interesting, and where I think the industry hasn’t fully caught up yet. Cloud gaming lowers the cost of entry into high-end PC gaming dramatically because you don’t need an expensive rig, just a device you already own.

Combine that with ad-supported access, which GeForce NOW already offers and Microsoft is now testing with Xbox, and you get a model where advertising effectively subsidises the hardware, not the game itself. You still buy the games you want on Steam or Epic and cloud gaming just removes the hardware barrier to playing them.

That’s a meaningfully different model to something like Game Pass, and it’s why I think Steam ends up a beneficiary of cloud gaming going mainstream rather than a casualty of it, particularly as services expand into markets like India, where the hardware cost has been the real barrier to PC gaming, not interest in the games.

From where I sit, that also raises questions the industry hasn’t really worked through yet: cloud gaming sessions don’t behave like traditional installs: they can start on an ad-supported tier, move across devices, and never touch a conventional “install” event at all.

As ad-supported access becomes a bigger part of how people get into games, measuring that engagement properly, who’s actually playing, on what, and what drove them there, becomes a real challenge, and one that’s going to matter more as this F2P PC model scales. 

Not sure if this answers your questions necessarily, but something we have been thinking about is the increasing focus on cloud gaming in the face of escalating games hardware costs – namely consoles.

Will cloud gaming services grow more rapidly due to higher-priced hardware as consumers look for other content access points?

We think this is unlikely. Service and content availability, alongside the industry challenges aligned to this distribution approach, mean the potential to benefit from the hardware price dynamics is undermined. 

The commercial factors that are battering the console market are also being felt across cloud gaming infrastructure.

But, in the cloud gaming model, the cost of the hardware shifts from the end user to the service provider. As infrastructure costs escalate, it seems inevitable that streaming-per-hour costs will also increase. Subscription service prices will need to go up, or other monetisation methods will need to be incorporated, to accommodate these costs.

There is also the general lack of availability of compute capacity across data centres thanks to the explosion of AI, and this means expanding cloud gaming to new markets or to serve larger audiences will be difficult. 

Essentially, the higher costs of streaming content are a fundamental issue that makes the scaling of the cloud gaming business challenging. Ad-supported offers are a crucial potential avenue to offset costs to the end user more passively – something Microsoft is actively exploring. 

Ampere estimates the pure-play cloud gaming market at $620m in 2025 – around 0.3% of global spend on games content and services.

From an indie developer’s perspective, I think cloud gaming has always had a slightly awkward problem: it is an impressive solution looking for the right problem to solve.

The technology itself is genuinely exciting. Removing the requirement for expensive hardware could make high-end games accessible to a much wider audience, and the ability to move between devices without worrying about installs, downloads or hardware specifications is genuinely attractive.

For developers, there is also an interesting possibility in reaching players who might never have owned the hardware traditionally required to run a particular game.

But the experience still depends on something developers have very little control over: the player’s internet connection. We can optimise a game enormously, but we can’t optimise away latency, bandwidth limitations or an unreliable connection.

For certain genres, that might be a minor inconvenience. For a competitive game, a precision platformer or anything where responsiveness really matters, it can fundamentally change the experience.

The economics are probably the bigger question, though. We’re used to thinking of digital games as having almost zero marginal distribution cost, but cloud gaming changes that equation. Every time somebody plays, somebody is paying for the hardware, bandwidth and energy required to stream that experience. If a player can play for hundreds of hours without buying additional hardware, that is fantastic for accessibility, but it creates a very different cost structure for the platform.

So I don’t think the introduction of usage limits is necessarily evidence that cloud gaming has failed. If anything, it demonstrates that the industry is still working out what a sustainable business model looks like. There is a difference between a technology being technically possible and a service being economically viable at enormous scale.

I’m also not convinced cloud gaming necessarily needs to become the primary way everyone plays games. It may ultimately be more valuable as another layer of the ecosystem. Being able to try a game instantly, play something on a low-powered device, continue a game while travelling, or access hardware-intensive experiences without owning a gaming PC could all be incredibly useful.

For indies, that accessibility is probably the part I’d be most interested in. We spend an enormous amount of time worrying about whether players have the hardware, storage space and willingness to install our games. If cloud technology can remove some of those barriers without creating new ones, that’s potentially very powerful.

I suspect the future of cloud gaming will therefore be less about replacing consoles and PCs and more about making the boundaries between platforms increasingly irrelevant. If the industry can solve the economics and the latency problem, it has a very interesting future. If it can’t, it may remain a useful complement to traditional gaming rather than the revolution it was once predicted to be.

It seems like we’re trading one bottleneck for another: limiting streaming bandwidth and server capacity instead of making games themselves more accessible.

For high-end games, cloud streaming makes sense because it lets them run on devices that otherwise couldn’t handle them.

But plenty of games don’t need that level of compute, or could be optimised much further.

Players value being able to start playing quickly and easily, and now that server costs push cloud gaming toward usage limits, I think we’ll see more focus on optimising games so they can run locally on lower-end devices too. 

I’ve been wondering lately if cloud gaming may be about to have a moment due to rising hardware costs. 

So far the promised rise of cloud gaming has been hindered by 2 factors that I can see 

First, it turned out cloud gaming mostly attracts core players, rather than the more occasional players wanting access to top-end hardware. But core players often already own a powerful, pricey PC or HD console, and therefore have little use for cloud. 

Second, what cloud gaming does is trade the need for powerful local hardware for a heavy reliance on a fast, low-latency broadband connection. But low ping and high bandwidth aren’t yet available with consistent performance for a lot of people, even at home. And all the while, smartphones mean we’re all carrying devices with powerful rendering capacity in our own pockets! The trade-off never really made sense to me. 

Now back to 2026, high-end PCs, graphics cards and RAM have become extremely expensive and hard to come by. Perhaps some core players will find it difficult to justify a $3000 investment in a rig that they’ll only use for a few hours a day. Thus, cloud gaming may become an efficient way to pool unused compute across multiple players. 

That is, of course, if platforms allow cloud to become a cheaper alternative to buying hardware by not putting too many usage limitations. Xbox introducing monthly cloud gaming limits for Game Pass subscribers isn’t a great signal from that perspective.