In the news last week was a British Gas report with some rather dubious numbers about the amount of electricity (and thus money) being wasted by “vampire devices” that use power when plugged in but not being used. The Guardian had a nice summary of the criticisms of the latest round, and it was encouraging to not see the claims of a fossil fuel company taken at face value by at least one outlet, despite others regurgitating whatever press release got sent out. As pointed out by several critics, the framing of this as an issue (as with many climate-adjacent issues) of individual consumer behaviour rather than a result of larger forces beyond the control of individuals (e.g. manufacturing and design standards) is entirely on-point, even as it is clearly piggy-backing on the legitimate concerns around energy prices in the context of the European gas price shock resulting from the Russian invasion of Ukraine.
Terrence Eden on twitter linked to his blog post from last year (the last time these claims got aired) investigating the original study from 2013 that British Gas once again relied upon to calculate these figures, pointing out they are incredibly out of date and unreflective of the latest devices and standards. Citing the standby power use of a PlayStation 2 for instance is nonsensical in 2022, especially, as Eden pointed out, that Sony themselves publish up-to-date power consumption figures for their latest devices (which I wish I’d know about before I finished the book TBH!) though these are buried deep in the “Legal” section of the Sony website and not widely known about. Last week as all this was happening, I just so happened to be reading through the latest version of the initiative that governs these figures, the agreement being part of the reason why Sony even publishes them.
The initiative is called the “Games Consoles EU Self-Regulatory Agreement”, and I won’t go into the details of self-regulation but suffice to say most of what you need to know is right there in the name. This is not the EU trying to bring energy-profligate console manufacturers to heel, and is more like a simple mechanism to get them all to the table to see what stuff they already agree on and can codify into some (not even particularly binding) rules of the road. It is also not really a climate-focussed initiative, at least not primarily, which might work backwards from the known impact of the industry and/or a given constraint and say ‘you can use this much no more’ (and there are substantial barriers, practical and conceptual, to being able to do that yet anyway). Since the EU already has rules about “vampire power”, or the power consumed when switched off or in standby (any device made since 2013 and sold in the EU already has to use less than 0.5 Watts – that’s fuck all!).
The agreement (which can be found here) takes pains to emphasise that it aims to support innovation and development. Section 3’s “Commitments and Requirements” starts off the actionable portion of the agreement with the following commitment that all signatories will “use reasonable endeavours to… reduce the power consumption of Games Consoles to the minimum necessary to meet their operational specification while not limiting the industry’s ability to improve functionality and to innovate.” (10) There are a couple of other minor things that signatories to the agreement commit to do, like ensuring that at least 90% of games consoles sold meet the specified “power caps” (more on which in a moment), don’t alter their devices performance during testing to evade measurements (no Volkswagens here thank you), and lastly “foster open communication and active engagement with the European Commission, Member States and other relevant stakeholders regarding the energy and material efficiency of Games Consoles. This includes sharing expertise, experience, information, and best practice with the signatories of other ecodesign self-regulation measures.” (10)
This is all well and good, and seems pretty encouraging – and we should be encouraged that some of the biggest games companies in the world are party to this agreement, doing a small bit to reduce the large climate footprint of end user gaming (see chapter 6 of my book for just how large). But what we really need to understand about this agreement is the power caps themselves, which are fairly generous, but which only cover what might (legitimately) be considered wastedpower – power in excess of what is needed, and only needed in two specific modes of operation.
To eliminate this waste, the agreement specifies a maximum amount of power to be consumed by the device first in media playback mode (streaming or playing from physical media like a BluRay) and second in navigation mode (when the console is on, not playing a game, just showing menus). Because not all consoles are made equal, it splits these power caps into three – one for classic ‘HD’ capable gaming devices (1080p – mainly applicable to the Nintendo Switch these days, but also applicable to the Xbox Series S), UHD or 4K capable devices (PS4 Pro, Xbox One X, etc.) and 8K capable devices (both SKUs of the PS5, and the Xbox Series X).
Here is the table for Navigation:
What this shows is a 22% decrease in the power cap from 2014 to 2017, followed by a 28% decrease to 2019 (except for UHD Gaming Capable Consoles which for some reason remain a the same 2017 level) and no change in 2021, except for the introduction of an 8K capable category (to govern the PS5/Xbox Series X, which were released too late, I expect, to be included in 2020’s agreement). If the pattern were to continue, I would expect to see another 20-30% reduction in power consumption in 2022 (making this year a tier 7 – presuming a new one is added). I would expect (and hope) to see a figure of 40W in the HD category of console navigation in this year’s agreement, perhaps 50W for 4K devices. A figure of 50W in 8K devices, which tend to otherwise be more power hungry than their last-gen versions (in absolute terms) would be an achievement as well. We shall have to wait and see.
Are these sorts of reductions possible to sustain or will manufacturers start to reach diminishing returns? These devices are still nowhere close to the efficiency of mobile devices, which post wattages in the single digits. The introduction by certain chip-manufacturers of ‘efficiency cores’ into a DIE, as in Intel’s 12th Gen Alder Lake CPUs and Apple’s M1 chips, could be the kind of game-changer that consoles need to keep these numbers dropping. If, and only if, they can make their way into devices. But so far I haven’t been able to find a similar offering from AMD, who just so happens to be the manufacturer for both PS5 and Xbox Series X main chips. (Side note: Alienware’s Aurora R10, which used AMD CPUs, was one of the few products last year to get caught out by new Californian energy efficiency legislation, limiting overall system power consumption of certain devices. Whoops.)
Next is the table for Media Playback.
Here we see a nearly identical 20% drop in the cap from 2014 to 2017, with UHD (i.e. 4K) devices being absent from the 2014 numbers, only showing up after 2016. 90W for media playback is a pretty high figure, and bizarrely, the cap seems to have been increased for UHD (4K capable) consoles from 2016 to 2019. (Unless I am reading the table wrong) Granted, playback at 4K does involve a substantial technical feat of decoding and outputting to a screen lots and lots of pixels, one that perhaps has to rely more on the in-built GPU, but even so 110W is close to the power output of the PS4 Pro when in active gaming mode, which uses an average of 146 W (according to their own figures). Likewise, Sony’s figures for the three PS4 Pro SKUs never went above 60W for media playback anyway.
So there is an encouraging trend happening here, though there is almost always room for greater ambition. Most of these figures are, in the context of an individual household’s energy use, really not that high, but neither are they insignificant. What we really need to remember, however, is that these EU standards set the power consumption for millions of devices over the next decade. In aggregate the difference between 60W and 110W across a million devices is substantial, however any eventual impact of squeezing down these figures will be further affected by other factors – such as the sales figures of consoles, individual patterns of use, the emissions intensity of different national energy grids, and so on. As we saw in 2020, lots of people started played games during the pandemic, and indeed these and other lockdown related behaviours saw an aggregate drop in energy demand and emissions growth. But this should spur us on to more ambition, not less, especially if we are advocating for gaming as a lower-emissions switch in activity away from, say, travelling to a sports match (as some have done!).
So this is not straightforwardly a critique, rather it points toward the need to re-emphasise the urgency of action, and that these numbers should be the most ambitious they can possibly be, in light of the climate emergency. On that front, we know, again thanks to Joshua Aslan, that past projections of energy consumption for both Navigation and Standby modes (at least for the PS4) over-estimated their eventual power consumption. Inversely, we might say that these early projections expected less ability to adapt to efficiency measures from console hardware manufacturers (though Aslan’s charts here do start coincidentally in 2013, precisely the year the EU’s energy efficiency regulations first came into force).
Aslan concludes this section of the discussion by noting that “on average the power consumption of PlayStation 4 (excluding networked standby mode) has reduced at a rate of 1.64 W per month – this is approximately seven times the rate of reduction predicted by Webb (2014), at 0.22 W per month and almost double the rate predicted by Malinowski, Acharya and Radulovic (2015), at 0.89 W per month.” (85) Might the EU agreement still be under-estimating the ability of its signatories to reduce its consumption over time? Given that it’s Sony, Microsoft and Nintendo that are basically deciding these numbers on their own, it would seem a hard argument to make, however one possibility is that they themselves might not know precisely what is possible. With any luck, we should know in the next few months what they think is possible for the next couple of years ahead, as the 2022 meeting is due to be held before June where I’m sure these issues will be discussed. I’m trying to get myself invited to the meeting, as the process allows for outside stake-holders to attend and ‘ask questions’. I imagine I’ll have a few. Maybe we could convince a few journalists to come along as well, I imagine that might change the feeling of the room somewhat, help squeeze out some more ambition.
The main question I have for the signatories to the agreement, which I’ll leave you with for now, is why these caps currently leave out perhaps the single most important part of the games console and its energy consumption: the active gaming use phase of the device. If we consider the total energy consumption of these devices – which with this latest generation have increased in absolute terms, even as they increase performance by much more – then surely the active gaming phase of the device is of central importance and should come under this same regulation? There will be arguments to be had about the centrality of high energy consumption to the active gaming use-phase, but it would at least be great to get something on the record. Perhaps we can get some of these questions answered in Brussels (virtually, I hope).
Thanks for reading the first issue of Greening the Games Industry – a brand new newsletter written by Dr Benjamin Abraham, author of Digital Games After Climate Change.
I'm really excited to start this new project, as I think the time has come for the games industry to address its environmental footprint, and I see momentum for change building in all corners. I hope to use the posts here to keep readers informed about the latest research and developments, build optimism and excitement about the positive changes that are happening (and yet to happen), as well as providing some industry best-practice for how to go about reducing emissions in key areas.
I have half a dozen stories planned for the coming weeks already, touching on different parts of the games industry's environmental impact, what's known already and what we still have to find out, as well as what's been happening in the industry's green transition. These pieces will be informed by the research I (and many others) have been doing to help guide the industry to a net-zero emissions future. I'll be aiming to post here once or twice a week, with longer pieces like this interspersed with the occasional roundup of news and interesting developments.
I would love it if you came along for the ride, and if you want to support me in continuing to do this work, you can subscribe for as little as $7 a month.