The End of the Lap Warmer: What Ditching x86 for ARM Would Save the World

Switching from x86 to ARM would cut laptop and data centre power use substantially, since ARM chips deliver similar performance on far less energy. The result is cooler machines, quieter fans, longer battery life, lower electricity bills, and a meaningful drop in global carbon emissions from computing at scale.

You know the moment. You are forty minutes into a meeting that could have been an email, your laptop is perched on your knees, and somewhere around the third slide about “alignment” you realise your thighs are slowly being sous vided. The fans are screaming like a small jet trying to take off from your lap, the battery icon has gone from green to a sort of anxious orange, and the person next to you is politely pretending not to hear it. You shift the machine onto the table, then back onto your lap, then onto the table again, performing a little dance of thermal management that nobody budgeted for.

Next time this happens, I want you to sit there, gently roasting, and imagine a different world: one where the computer on your lap is cool, silent and still on 70% battery at four in the afternoon, where the ATM down the road is run by a board the size of a credit card, and where the server room hums a little more quietly. Machines like that already exist and many of them run on ARM, so I built a model of what it might be worth to the planet if every x86 Windows laptop, desktop, kiosk and server on Earth moved to a functionally equivalent ARM machine running Linux, macOS or another ARM compatible operating system. The headline, with every assumption published further down so you can argue with it, is this:

An illustrative full transition scenario (central case) suggests annual operational electricity savings of about 74 TWh worldwide, within a range of roughly 29 to 168 TWh. Realisable savings depend on workload eligibility, measured power differences and adoption at normal replacement cycles; on central assumptions, about 36 TWh is realistic.

1. Where x86 came from, and why it is faintly ridiculous

To understand what is cooking your legs, we need to go back to 8 June 1978, when Intel released the 8086: a 16 bit processor with about 29,000 transistors, a clock speed of 5 MHz and the ability to address one megabyte of memory, which it could only see through 64 KB windows at a time. It was a perfectly respectable chip for its day, roughly contemporary with the Sony Walkman, and nobody imagined that its instruction set would still be running a large share of the world’s laptops, desktops and servers nearly half a century later.

The 8086 did not even make it into the machine that made it famous: the original IBM PC used the 8088, a cut down variant with an 8 bit external data bus, and the success of that PC, and of the clones that followed, locked the industry into the x86 instruction set. From that moment the overriding commercial priority was backwards compatibility, so every generation since has been extended, patched and bolted onto rather than redesigned. A modern x86 processor still, in principle, starts life pretending to be a 16 bit chip from the disco era before being coaxed up into the present day.

timeline
    title Nearly fifty years of backwards compatibility
    1978 : Intel 8086 launches with about 29,000 transistors
    1981 : IBM PC ships on the 8088 and locks in x86
    1985 : Intel 386 takes x86 to 32 bit
         : Acorn ARM1 first silicon
    2003 : AMD64 takes x86 to 64 bit
    2007 : iPhone puts ARM in every pocket
    2020 : Apple M1 puts ARM in the laptop
    2024 : Apple M4 with 28 billion transistors

My favourite piece of archaeology is the A20 gate. Early PCs had a quirk around a single memory address line, and for a long stretch of history the mechanism used to control it was routed through the keyboard controller. Yes, the keyboard controller. Imagine explaining to a new engineer that the way you unlock memory is by asking the thing that handles your typing, and then imagine that this was considered a perfectly sensible solution for years.

Then there are the instructions themselves, which vary in length from one byte to fifteen, so the front end of the processor has to work out where each instruction even starts before it can do anything useful with it. Modern x86 chips cope with this by translating those instructions internally into simpler micro operations, which is a bit like a restaurant that takes orders in Latin, translates them to English in the kitchen, and then cooks the food. It works, and the engineering is genuinely brilliant, but somebody is doing the translating, and the decode machinery costs some silicon and some power.

In fairness, and specialists will rightly point this out, ARM is not exactly a teenager either, since its roots go back to Acorn in the mid 1980s, and older ARM instruction sets mix 16 and 32 bit instructions. It is the 64 bit A64 instruction set used by modern ARM laptops and servers that has fixed 32 bit instructions, which makes decoding simpler. More importantly, the instruction set alone does not decide how efficient a chip is: process node, design targets, how tightly the hardware and operating system are integrated, and how hard the vendor tunes for battery life all matter at least as much. ARM’s advantage is better described as heritage than physics; it grew up in phones that had to run all day on a tiny battery, and that design culture shows.

2. Why unified design is the way to go

The bigger shift is not just the instruction set but the shape of the whole machine. A desktop with a discrete graphics card is a collection of separate parts with separate pools of memory: system RAM on one side of the board, a graphics card with its own memory on the other, and a lot of time and energy spent shuffling data between them across an expansion bus. Shared memory is not new or unique to ARM; x86 laptops with integrated graphics have shared memory between CPU and GPU for years, and AMD ships x86 chips built around the same idea. What Apple Silicon and other modern ARM systems on a chip did was make a high bandwidth, tightly integrated design the default rather than the exception, with the CPU, GPU and neural engine on one piece of silicon working on the same data without copying it around.

flowchart LR
    subgraph PC["Desktop with a discrete graphics card"]
        direction TB
        CPU1["CPU"] <-->|"copy across the PCIe bus"| GPU1["GPU"]
        CPU1 <--> RAM1["System RAM"]
        GPU1 <--> VRAM1["Graphics RAM"]
    end
    subgraph SOC["ARM system on a chip"]
        direction TB
        CPU2["CPU"] <--> UM["One pool of unified memory"]
        GPU2["GPU"] <--> UM
        NPU2["Neural engine"] <--> UM
    end
    PC ~~~ SOC

The fair comparison is the newest Intel against the newest Apple Silicon. Intel’s latest laptop chips are the Core Ultra series 3, codenamed Panther Lake and built on Intel’s own 18A process; Apple’s newest is the M6, its first 2 nm chip, which shipped in the Mac mini in September 2026. The numbers below come from Notebookcheck, which measures whole system power on an external display, and its M6 analysis tested the M6 against several Panther Lake machines in the same way. Where Notebookcheck tested more than one Panther Lake chip, I have used Intel’s best result in each row, so the comparison is as kind to Intel as the data allows.

🟢 marks the better result in each row and 🔴 the worse one; the second column says whether a higher or lower number wins.

type: hbar
title: Work done per watt, latest generation
subtitle: Cinebench 2024 points per watt of measured system power (higher is better)
x: Single core, Multi core (default limits), Multi core (Intel capped at 20 W)
Intel Panther Lake (best result tested): 5.1, 20.0, 24.6
Apple M6: 15.2, 32.5, 32.5
colors: #e34948, #1baf7a
source: Notebookcheck measurements on an external display; Intel figures are the best Panther Lake result in each test, from the Ultra X7 358H, Ultra 9 386H and Ultra X9 388H

Apple’s M6 wins six of the seven measured rows. It is faster on both speed tests, draws about 40% less power while running a single core benchmark, and does about three times as much work per watt on a single core and about 1.6 times as much across all cores. Even when Intel’s chip is deliberately held to a frugal 20 W limit, the M6 still does about 30% more work per watt. Intel’s one win is idle power, where lower is better and its newest chip draws about 0.4 W less than the M6 Mac mini, and that matters for office machines that spend much of the day waiting for someone to type. The gap that matters most for the world’s installed base, though, is less about the newest chips and more about the five to eight year old x86 machines still sitting under desks, which is why the model compares ARM against both the old machine and a modern x86 refresh.

For the machines actually being replaced, Apple’s own figures for the last Intel Mac mini (2018, Core i7) against the first Apple Silicon one (M1, 2020) are a useful guide: 19.9 W against 6.8 W at idle, and 122 W against 39 W at maximum. In the data centre, AWS says its Graviton based EC2 instances use up to 60% less energy than comparable instances for the same performance, which is a vendor’s best case for workloads that suit it rather than a general rule. None of these is a measurement of an average office anywhere.

3. How the model works

The method is deliberately simple, so anyone can check it or swap in their own numbers. For each class of machine I multiply how many there are, how much power each would save, and how many hours a year it runs. Servers are handled differently, because counting boxes is a mug’s game: I estimate how much electricity Windows servers actually draw and apply a percentage reduction, which means a production server, its DR twin and the QA box in the corner are all counted through the power they consume rather than through a guess at how many exist.

flowchart TD
    A["Machines by class"] --> B["Watts saved per machine"]
    B --> C["Hours a year"]
    C --> D["TWh saved"]
    S["Windows server electricity in TWh"] --> R["Percentage reduction"]
    R --> D
    O["Data centre overhead"] --> D
    D --> F["Full transition ceiling"]
    F --> G["Realistic: only eligible workloads move"]
    G --> H["ARM share: compared with a modern x86 refresh"]

The model produces three answers rather than one, because they answer different questions. The full transition ceiling assumes every machine moves. The realistic figure only moves the share of each estate that could plausibly migrate at normal replacement cycles, since plenty of Windows applications, drivers and peripherals will not move cheaply or at all. The ARM share asks how much of the realistic saving is really down to choosing ARM, rather than simply replacing old x86 kit with new, efficient x86 kit, which would capture a good part of the gain on its own.

4. Counting the world’s machines

For PCs, the cleanest anchor is Microsoft’s own: in January 2022 it reported more than 1.4 billion monthly active Windows 10 and Windows 11 devices. Windows still accounts for about 76% of desktop operating system traffic worldwide according to StatCounter’s September 2026 figures, and the market is refreshing at roughly 270 million new PCs a year according to Gartner. The central case therefore uses 1.4 billion Windows x86 PCs, split 60% laptops and 40% desktops, which is my assumption for the installed base rather than a published figure.

Then there are the machines nobody thinks of as computers. RBR counted about 3.09 million ATMs worldwide in 2021, and behind those sit self service checkouts, airport and ticket kiosks, queue systems, fuel forecourt systems and digital signage, a large share of which run Windows on a small x86 PC around the clock. Nobody publishes a global count of these, so the central case assumes about 15 million such devices including the ATMs.

Servers are where the uncertainty is greatest, so here is the chain in full. The International Energy Agency estimates that the world’s data centres used about 415 TWh of electricity in 2024, roughly 1.5% of global consumption, with servers accounting for about 60% of it on average. A growing slice of that server power now goes to AI accelerators, which are not running Windows, so I take only conventional servers and then the Windows share of those, and add a margin for the server rooms and branch servers that sit outside any data centre.

At an average of 350 W per physical server, that is the equivalent of roughly 21 million Windows servers humming away somewhere, from hyperscale halls to the cupboard behind reception.

5. How much power each machine saves

This is the input that drives everything, and it deserves honesty: there is no published, workload matched measurement of the world’s office machines, so these are assumptions informed by the evidence above rather than measurements.

The central desktop figure of 25 W is larger than the measured idle gap and far smaller than the gap at full load, so it is an assumption about real office use, which sits somewhere in between, rather than something that comparison proves.

6. What it adds up to

In the central full transition case the saving comes to about 74 TWh a year, worth roughly US$8.8 billion at an assumed US$0.12 per kWh, and equal to about 0.23% of the 31,779 TWh of electricity the world used in 2025 according to Ember. Put another way, it is about 18% of everything the world’s data centres used in 2024.

type: bar
title: Where the ~74 TWh comes from
subtitle: Central full transition scenario, annual electricity saved by device class
x: Laptops (840m), Desktops (560m), "ATMs, kiosks, tills (15m)", Windows servers (~64 TWh), Data centre overhead
y-label: TWh a year
suffix: TWh
Saving: 13.4, 35.0, 5.3, 16.1, 3.8
colors: #2a78d6
source: Illustrative model. Laptops 2,000 h a year, desktops 2,500 h, kiosks and servers 8,760 h. Overhead uses Uptime Institute's average PUE of 1.54, with half of it scaling with IT load.

The three answers sit quite far apart, which is the point of showing them together.

type: bar
title: Ceiling, realistic and ARM share
subtitle: Annual electricity saved worldwide under each scenario
x: Conservative, Central, Aggressive
y-label: TWh a year
suffix: TWh
Full transition ceiling: 28.6, 73.6, 167.8
Realistic at normal refresh: 8.8, 36.2, 115.4
ARM share vs modern x86: 3.5, 18.1, 69.3
colors: #2a78d6, #1baf7a, #eb6834
source: Illustrative model; inputs for every scenario are listed in section 12

7. How many homes, and how many power stations?

Numbers like 74 TWh are hard to feel, so let us translate them into homes. If a household uses something like 3,000 kWh a year (a deliberately round assumption, since real usage varies enormously from country to country), then the central full transition would cover the annual electricity of roughly 24 million households. Picture a village as around 400 homes and that is in the region of 61,000 villages kept lit for an entire year, using electricity we currently spend on keeping laptops warm, ATMs toasty and server rooms cool.

Power stations are where this gets fun. A large 1 GW nuclear reactor running 90% of the time produces about 7.9 TWh a year, so the central full transition is worth the output of about nine of them, and the aggressive case about 21. The usual caveat applies: these figures are annual averages, energy saved divided by the hours in a year, rather than a constant reduction or a dependable contribution at peak, because office machines save power during office hours while servers save it around the clock. Lower electricity demand can reduce generation requirements; the capacity benefit depends on when those savings occur. The comparison below is illustrative.

type: bar
title: Large nuclear reactors' worth of output
subtitle: Annual saving divided by the output of a 1 GW reactor running 90% of the time (~7.9 TWh a year)
x: Conservative, Central, Aggressive
Full transition: 3.6, 9.3, 21.3
Realistic at normal refresh: 1.1, 4.6, 14.6
colors: #2a78d6, #1baf7a
source: Illustrative model; reactor equivalents describe energy over a year, not firm capacity

8. The ATM that thinks it is a desktop

This is the part of the argument I find most compelling, because it is so obviously absurd once you look at it. Inside a typical ATM or self service kiosk there is, quite often, a small form factor x86 PC running a full desktop class Windows installation, whose entire job is to show a few screens, talk to a card reader, a PIN pad and a cash dispenser, and send messages to a host. It is the computing equivalent of using a delivery truck to post a letter.

A Raspberry Pi class ARM board can do that job comfortably. Even being generous and allowing 15 W for the board, its display controller and peripheral interfaces, it sits well below a typical x86 kiosk PC drawing perhaps 50 to 60 W around the clock, and the board itself costs a fraction of the PC it replaces.

type: hbar
title: The brain of a kiosk does not need to be a desktop PC
subtitle: Compute module power draw, illustrative
x: Small form factor x86 PC, Raspberry Pi class ARM board (generous)
suffix: W
Power draw: 55, 15
colors: #1baf7a
source: Compute module only; dispensers, displays and printers unchanged

flowchart LR
    subgraph Today["Typical kiosk today"]
        direction TB
        W["Small form factor x86 PC<br/>full desktop Windows<br/>~55 W, 24/7"] --> X1["XFS drivers"]
        X1 --> P1["Card reader, PIN pad,<br/>dispenser, printer"]
    end
    subgraph Tomorrow["ARM kiosk"]
        direction TB
        R["Raspberry Pi class board<br/>hardened minimal Linux<br/>~15 W, 24/7"] --> X2["Device services"]
        X2 --> P2["Same card reader, PIN pad,<br/>dispenser, printer"]
    end
    Today -. "swap the brain,<br/>keep the peripherals" .-> Tomorrow

The caveat is that ATM software stacks have historically been built around Windows, through the CEN/XFS standard that lets applications talk to dispensers and card readers, so this is a vendor and certification journey rather than a weekend project, which is why the central case assumes only half of these devices can realistically move.

9. Desktops carry the bulk, servers punch above their weight

At global scale the biggest single prize is the humble office desktop, simply because there are hundreds of millions of them. Servers are only about 1.5% of the machines in the central model, yet they account for about 27% of the full transition saving.

type: stacked
title: Servers are 1.5% of the machines but 27% of the saving
x: Share of machines, Share of savings
suffix: %
y-max: 100
PCs: 97.5, 65.9
ATMs, kiosks, tills: 1.0, 7.1
Windows servers incl. overhead: 1.5, 27.0
colors: #2a78d6, #1baf7a, #eb6834
source: Central full transition scenario; servers counted as 350 W host equivalents

The reason is simple: your laptop spends most of the year asleep in a bag, while a server runs every hour of every day and drags an air conditioner along behind it. So the warm laptop on your knees is the visible symptom, the desktop under the desk is the biggest single line, and the server room that nobody visits is where each migration pays back fastest.

That server prize is also the hardest to claim. The major cloud providers have made ARM fairly painless for many Linux workloads, but that success does not transfer automatically to Windows workloads; an estate built on Windows Server, SQL Server on Windows, .NET Framework applications and vendor packages has to be replatformed before ARM is even an option, which is why the central case assumes only 40% of Windows server power can realistically move. There is also a quieter saving hiding in the same rooms: a lot of DR, test and QA capacity sits nearly idle, and idle x86 servers still draw a surprising amount of power. Consolidating or switching off that capacity saves energy whatever the processor is, so it belongs in the business case but should not be credited to ARM.

10. The economic case, with its costs

The electricity is the first line of the business case: roughly US$3.4 billion to US$20 billion a year across the full transition scenarios, or US$1.1 billion to US$14 billion on the realistic figures, at an assumed average of US$0.12 per kWh that will be too high in some countries and too low in others. Average PUE does not guarantee that facility overhead falls in proportion to IT load, since cooling and power distribution have fixed elements, so the model assumes only part of the overhead scales down.

The other lines are real but harder to size, and I have deliberately not put numbers on them. A credit card sized ARM board costs a fraction of a kiosk PC, and AWS markets its Graviton instances on better price performance than comparable x86 instances. Licensing may fall where full Windows desktop and server licences are no longer needed, but Linux estates still need security tooling, patching, monitoring and support, so any licence saving needs to be priced against an organisation’s actual agreements. Against all of this sit the costs: application rewrites and testing, vendor certification, dual running during migration, retraining and new support skills.

flowchart TD
    M["Move eligible workloads from x86 Windows to ARM"] --> E1["Electricity<br/>US$1.1bn to US$14bn a year realistic"]
    M --> E2["Capital<br/>cheaper boards and ARM cloud instances"]
    M --> E3["Licensing<br/>needs pricing evidence"]
    M --> C1["Costs<br/>migration, testing, certification,<br/>dual running, skills"]
    E1 --> S["Lower demand and emissions"]

11. The social and environmental case

At Ember’s 2025 global average of 458 g of CO₂ per kWh, the central full transition avoids roughly 34 million tonnes of CO₂ a year, with the realistic central figure closer to 17 million tonnes. The actual figure depends heavily on where the savings happen, since a kilowatt hour saved on a coal heavy grid avoids far more carbon than one saved on a hydro or nuclear grid.

type: bar
title: Carbon avoided each year
subtitle: Million tonnes of CO2 at the 2025 global average of 458 g per kWh
x: Conservative, Central, Aggressive
suffix: Mt
Full transition: 13.1, 33.7, 76.8
Realistic at normal refresh: 4.0, 16.6, 52.9
colors: #2a78d6, #1baf7a
source: Illustrative model; carbon intensity from Ember

The timing matters too. The IEA expects data centre electricity use to more than double to about 945 TWh by 2030, driven largely by AI, at the same moment grids everywhere are being asked to electrify transport and heating. Every terawatt hour that the world’s existing laptops, desktops, kiosks and servers stop wasting is a terawatt hour available for that growth, or for the hundreds of millions of people who still have unreliable power or none at all. And at the level of the individual, a laptop that genuinely lasts a full working day on battery keeps working through an outage, which matters most in exactly the places where the grid is least reliable.

12. Every assumption, so you can argue with them

13. Before anyone sends me an angry email

The model deliberately bundles three changes that a careful analysis would separate: the move to ARM, the move off Windows onto operating systems that manage power more aggressively, and the consolidation that usually comes with a migration. The “ARM share” figure is my rough attempt to separate the first from the rest, by comparing against a modern, efficient x86 refresh rather than against the old machine.

This is also an operational electricity calculation. It does not account for the energy, emissions and materials involved in manufacturing replacement devices, and with well over a billion PCs in play, a full lifecycle view could easily show that throwing away a working machine early does more harm than good. The sensible reading is that ARM should be the default whenever a machine is due for replacement anyway, not an excuse to fill the world’s landfills with perfectly serviceable laptops.

Finally, the installed base is an estimate. I have anchored it to Microsoft’s device count, published ATM counts and the IEA’s data centre figures so that it can be checked, but nobody publishes a reliable count of Windows x86 machines in use, or of how much of the world’s server power runs Windows, and better inventory data would move this estimate more than any refinement of the wattages.

14. So, is this the end of the lap warmer?

Probably, though not in a dramatic global switchover; it will happen one replacement cycle at a time, as organisations discover that the new machine is silent, cool and cheaper to run. An illustrative full transition suggests annual operational savings of about 74 TWh worldwide, within a range of roughly 29 to 168 TWh, which is about nine large nuclear reactors’ worth of output in the central case. What is realisable depends on which workloads can actually move, what the power difference turns out to be when someone measures it properly, and how quickly organisations adopt ARM at normal refresh; on central assumptions that is closer to 36 TWh, still enough to keep tens of thousands of villages lit every year.

So the next time you are sitting in a meeting with your thighs gently browning, take a moment to appreciate that you are personally hosting a small tribute to 1978, along with a billion or so other people. Then, when it is time for a new machine, a new kiosk or a new server, consider letting the lap warmer retire with dignity. It may not single handedly save the world, but it is one of the easiest favours we could do it.