
A game can look much the same on two screens while asking very different things of the equipment behind them. France’s ecological transition agency, ADEME, set out to measure those differences in its first comprehensive assessment of video gaming’s environmental impact in France, published on 22 October. Its findings give the data centre industry meeting on 9 November at Bercy a concrete example of what changes when a digital service moves more work off the user’s device.
What ADEME counted
The study looked at player equipment, networks and servers across 18 representative scenarios. It estimated the carbon footprint of video game use in France at 0.96 to 1.16 million tonnes of carbon dioxide equivalent, or around 3% of the country’s digital carbon footprint. That is an estimate for the sector in the study. Its value lies in including the equipment and infrastructure used for play in the same assessment.
Including the equipment changes the questions an assessment can answer. A console, a phone and a large television require materials and energy before anyone switches them on. Once a session starts, the display, processing hardware and connection all have a role. Looking at the electricity drawn during play gives useful information about that session. Looking at the equipment’s wider life helps explain what happens if the way people play affects how often they replace it.
Where a cloud session moves the work
In cloud gaming, a remote machine runs the game and streams it to the player’s device. The device still displays the picture and sends back the player’s inputs. This setup can make a demanding game playable on hardware that would struggle to run it locally. It also uses the network and remote servers throughout the session.
ADEME gives an idea of the range even within cloud gaming over Wi-Fi. It estimates around 128 grams of carbon dioxide equivalent per hour for one smartphone gaming scenario and nearly 250 grams for a home console scenario. Both scenarios rely on defined equipment and usage assumptions. They help explain why the words cloud gaming alone are too broad to describe a service’s footprint.
What a browser-based service would need to measure
An audit of browser-based gambling games would start by tracking activity during a session. The browser may download images, sounds and code, then render some of the activity on the device. Remote systems can handle requests while the session is running. A useful assessment would record the data transferred, the work performed on each side and the equipment used. ADEME’s figures describe video gaming scenarios. They provide a basis for learning, while the numbers for a particular gambling service would require its own measurements.
Even two sessions on the same browser could give different results. A returning visitor may already have some of the files used on a first visit. An older device can take longer to process an animation. Someone playing for a few minutes and someone keeping a page open for an hour may make different demands on the service. Defining a representative session gives an audit a basis for comparing design changes.
What will the new infrastructure host?
The French economy ministry’s announcement for the Bercy meeting describes a gathering of international project developers, industrial suppliers and energy-transition participants. A single facility can serve many businesses. To understand how much of its activity comes from a particular type of service, you need to know what runs there, for how long and how much work is shared. The video game study shows why activity type belongs in a discussion about new capacity.
Practical design decisions happen on both sides of the connection. A streaming service can adjust the data it sends to suit the screen and connection. A browser developer can investigate whether the same assets are repeatedly transferred. A hardware maker can assess how long a device remains useful. Each action addresses a different part of the footprint, and measuring only the server would give an incomplete view of its effects.
The same principle applies when an operator changes a service. A new feature could reduce traffic while making an older phone work harder. A before-and-after comparison should capture both effects. That lets a team see whether the work has decreased across the system or simply moved to a different part of it.
Better comparisons need the same starting point
The figures become more useful when comparisons specify the equipment, network, software and time spent playing. A small change in one assumption can change which part of the service uses the most resources. That is where the French study contributes beyond its national estimate. It shows the value of describing an ordinary session in enough detail to see where its environmental costs arise and which design choices could change them.