Cloud gaming can shift a lot of electricity use away from your home, but that does not necessarily make the total system more energy-efficient. A cloud session still needs a gaming server, data-center infrastructure, the network between you and the server, and your local streaming device.
The strongest widely cited lab research found cloud gaming used more total electricity than comparable local gaming in its tests. The important caveat is that those measurements are now several years old, so their exact wattages should not be presented as universal 2026 numbers.
What the Berkeley Lab research actually found
Lawrence Berkeley National Laboratory’s Green Gaming project tested a range of gaming hardware and cloud scenarios. Its cloud-gaming summary says cloud gaming required significantly more energy than similarly powerful local equipment across its tests, reaching roughly three times as much in the most extreme cases it identified.
The project attributed the difference to three loads:
- the remote gaming server;
- data-center cooling and electrical overhead;
- network energy between the server and player;
plus the electricity still used by the player’s local device and display.
That is useful evidence, but it came from an earlier cloud-gaming generation. Berkeley Lab’s published PC-cloud example estimated about 340W per user at the data center and 180W for the network. Those figures are study assumptions and measurements from that period, not current specifications for GeForce NOW, Xbox Cloud Gaming or every modern network.
Why a simple “cloud uses 3× more” rule is too strong
Cloud infrastructure keeps changing. Servers become more efficient, utilization can improve, networking equipment changes, codecs become more efficient, and data centers can reduce overhead.
At the same time, cloud gaming is also pushing higher resolutions and frame rates, which can increase server and network demand.
So the correct 2026 conclusion is not “cloud gaming always uses exactly three times as much electricity.” It is:
Cloud gaming adds server and network energy that local gaming does not need, so total energy can be higher even when the device in your room uses much less power.
Local gaming power also varies enormously
A local gaming PC is not a fixed 300–500W appliance. Our gaming PC power guide explains why a modest system can use far less while a flagship GPU build can use far more.
Microsoft’s current Xbox sustainability telemetry gives another useful reference point. Its global platform baselines report average gameplay power of about 122W for Xbox Series X and 61W for Xbox Series S. Those are console-at-the-wall averages, not cloud comparisons, but they show why the local device you compare against matters so much.
When cloud gaming can look efficient at home
If you stream a demanding game to a small tablet, thin laptop or TV app, the electricity measured at your wall may be much lower than a high-end gaming PC would use locally.
That is a real household-energy saving, but it is only part of the system boundary. The missing computation is happening elsewhere.
If your goal is your personal electricity bill, local client power matters most. If your goal is total electricity or carbon impact, server and network energy belong in the accounting too.
Carbon emissions are not identical to electricity use
Two gaming sessions can use the same amount of electricity but have different emissions if the electricity comes from different generation mixes.
Cloud providers can locate infrastructure on relatively low-carbon grids or procure renewable electricity. A local gamer may also have rooftop solar or a low-carbon grid. Conversely, either side can be powered by fossil-heavy electricity.
That means energy use and carbon emissions should not be treated as interchangeable.
What about manufacturing and e-waste?
This comparison becomes even harder when hardware manufacturing is included. Cloud servers are shared infrastructure, which could reduce the amount of high-end hardware needed in individual homes. But cloud systems also require large server fleets that are regularly upgraded.
Berkeley Lab explicitly noted that its cloud-gaming work had not fully evaluated the solid-waste side. That remains a reason to avoid sweeping lifecycle claims from operational power measurements alone.
What changes cloud gaming’s energy use?
- Stream resolution and frame rate: higher settings generally require more encoding, processing and network traffic.
- Server hardware and utilization: efficient, well-utilized infrastructure can reduce overhead per player.
- Network path: access technology, distance and network equipment affect energy demand.
- Client device: a TV, phone, laptop and desktop PC all consume different amounts while decoding a stream.
- Display: the screen can be a significant part of either setup’s electricity use.
- The game itself: rendering load varies substantially between titles and settings.
How to reduce gaming electricity use
For local gaming, right-size the GPU, cap unnecessary frame rates, use sensible power management and let the system sleep when inactive. Our energy-efficient gaming PC guide goes deeper into those choices.
For cloud gaming, avoid selecting a resolution or refresh rate your display cannot benefit from, and use an efficient client device when practical. Do not assume that turning down an in-game graphics preset will necessarily reduce the cloud service’s total electricity in a predictable way; service-side scheduling and hardware behavior are not under the player’s direct control.
Bottom line
The best evidence we have shows an important warning: moving game rendering to a data center can increase total operational electricity because the network and remote infrastructure are added to the client device.
But the popular “cloud gaming always uses three times more energy” claim is too absolute for 2026. That figure came from earlier Berkeley Lab scenarios and tests. The real answer depends on the local hardware being replaced, modern service efficiency, network conditions, streaming quality and where the electricity comes from.
