🎮

LostInConsoles

Your Gateway to Retro Gaming Reviews

Investigations

Nintendo's Switch 2 Declares 2.0 W in Sleep. Its Own Spec Sheet Declares 5.

By Nia 12 min read Industry Hardware Console Economics AI Memory
Nintendo's Switch 2 Declares 2.0 W in Sleep. Its Own Spec Sheet Declares 5.

Two answers to one question

Nintendo publishes, in English, a UK-facing page headed "Information provided under the requirements of Commission Regulation (EC) No 1275/2008 and Commission Regulation (EU) 2023/826". The page is scoped to the console Nintendo calls Nintendo Switch 2 [BEE-001/BEE-001(-01)], and its power table reads:

Networked standby: Max. 2.0W Another condition equivalent with standby mode: Max. 0.5W The power consumption of the equipment in networked standby if all wired network ports are connected and all wireless network ports are activated: Max. 2.0 W

Nintendo publishes, in English, a second page for the same console: the technical specifications. Its power table reads:

Sleep mode (Wi-Fi connected; wired LAN connection not maintained): 0.5 W Sleep mode (wired LAN connection maintained): 5 W Power off: 0.5 W Estimated annual energy consumption: 14 kWh / 50 kWh

Both pages were live on 17 November 2026, and both were served in German and French with the same figures. [OBSERVED FACT]

Two point zero, and five. The second number is two and a half times the first, and the state it describes — asleep, wired network connection maintained — is precisely the state the first page exists to declare. Which is the number a European buyer of a Switch 2 is supposed to be able to trust?

State, in Nintendo's own words Declared draw Against the legal ceiling
Sleep mode, Wi-Fi connected, wired LAN not maintained 0.5 W Under it
Sleep mode, wired LAN connection maintained 5 W 2.5× the 2.00 W networked-standby ceiling
Power off 0.5 W At the 0.50 W off-mode ceiling; over the 0.30 W ceiling from 9 May 2027
Estimated annual energy, Wi-Fi sleep 14 kWh —
Estimated annual energy, wired-LAN sleep 50 kWh 3.6×

What the regulation actually says

Regulation (EU) 2023/826 has applied since 9 May 2025. Its Annex II, category 4, lists "Toys, leisure and sports equipment: … games consoles". Annex I, point 19 defines the term for the first time in EU law — equipment whose principal function is video game playing, and, explicitly, handhelds running from a battery are "considered to be a type of games console". The predecessor regulation named only "hand-held video game consoles" in its scope list.

Annex III then sets the numbers:

  • 1(a): off mode shall not exceed 0.50 W, and "two years after the application of this Regulation" — 9 May 2027 — 0.30 W.
  • 1(c): "The power consumption of networked equipment, other than HiNA equipment or equipment with HiNA functionality, in networked standby shall not exceed 2.00 W."
  • 3(a)(1): instruction manuals and free-access websites "shall include … the power consumption expressed in watts rounded to the first decimal place" for each low-power mode, plus the time in minutes it takes to reach it.
  • 3(a)(2): they must also state "the power consumption of the equipment in networked standby if all wired network ports are connected and all wireless network ports are activated."

Annex IV, which governs how a market-surveillance lab tests this, is not ambiguous either: for networked standby it connects a network port, connects a device able to send "the remotely initiated trigger that will reactivate the tested unit", lets the unit fall asleep, measures, "then the appropriate trigger is given to the unit through the network port and a check is made on whether the equipment is reactivated." [OBSERVED FACT]

So the disclosure duty is not decorative. It exists because a console that keeps a network port alive is a console that can be woken over the network, and Europe decided what that may cost.

The experiment Nintendo ran on itself

Look again at the two sleep rows. They are not two different products, two different tests or two different years. They are the same console, the same dock, the same attached Joy-Con 2 controllers, and the only stated difference is whether the wired LAN connection is maintained. The gap between them is 4.5 W.

That matters for the obvious objection — that the 5 W figure might include charging the Joy-Cons or the battery. Both rows describe a console in sleep. In both, the controllers are attached. If trickle-charging were the explanation for 5 W, it would have to show up in the 0.5 W row too. The only variable Nintendo names is the LAN port. [INFERENCE]

The footnote on the annual figure is blunter still. 50 kWh is the "estimated annual electricity consumption when the wired LAN connection is maintained during sleep mode", and 14 kWh is the figure with Wi-Fi. Nintendo is telling its customers that leaving the network port live costs them 36 kWh a year more than letting it sleep on Wi-Fi: three and a half times the energy across the year, on Nintendo's own arithmetic. [OBSERVED FACT]

That arithmetic also disposes of the most convenient explanation: that 5 W is a download burst. Reconstructed with Nintendo's stated two hours of gaming a day and sleep for the rest of the year, the wired figure is 54 kWh — close to the 50 kWh declared — and the Wi-Fi figure is 17.9 kWh against 14 kWh declared. The small overshoot in both cases is consistent with rounding in Nintendo's method. What does not survive is the burst hypothesis: if 5 W applied for an hour of downloads a day rather than around the clock, the annual figure would land near 20 kWh, not 50. The 36 kWh gap between the two annual figures is, across 8,760 hours, 4.1 W of continuous extra draw — almost exactly the 4.5 W difference between the two sleep rows. Nintendo's own published annual numbers only close if the 5 W state is treated as a resting state. [INFERENCE, from Nintendo's published figures]

The previous generation did not do this

This is not a fixed property of Nintendo hardware. The Switch 1 spec page still publishes its numbers, with the same two-condition split:

Nintendo Switch (HAC-001) Wired internet setting off Wired internet setting on
Sleep mode 0.3 W 2.2 W
Power off 0.3 W —
Estimated annual energy 6 kWh 21 kWh

The Switch 1 already sat 10 per cent above the 2.00 W ceiling in the wired state, at 2.2 W. The Switch 2 more than doubled it, to 5 W, and raised the annual figure from 21 kWh to 50 kWh. The direction of travel, on Nintendo's own published data, is the wrong one. [OBSERVED FACT]

The other two signatories publish something different

The same disclosure duty binds Nintendo's two competitors, and both are named in the same self-regulatory filing system. What they publish is not the same kind of document.

Sony's PlayStation ecodesign page carries a measured table per model, with a test date. For the current disc-drive PS5, model CFI-71, tested 05/02/2026:

PlayStation 5 (CFI-71), Sony's labels Draw
Low power Rest 0.30 W
Rest (Supply Power to USB Ports) 4.65 W
Rest (Stay Connected to the Internet) 0.84 W
Rest (All) 4.71 W

The like-for-like comparison with Nintendo's 5 W row is Sony's "Rest (Stay Connected to the Internet)" — a console asleep with its network connection alive — at 0.84 W, roughly one sixth of Nintendo's declared figure. Sony's 4.71 W "All" row includes supplying power to the USB ports, which is a different function from maintaining a network connection, and its page says the "'low power use' setting is the default standby setting on PS5" — the same default-versus-option structure Nintendo describes. [OBSERVED FACT]

Microsoft's 2025 Product Compliance Report, dated 15/01/2026, does not carry the numbers itself. At item 1.6 it points readers to an Xbox support page, "learn-about-power-modes", for "information on console power consumption of navigation, media play, gaming, regulatory standby mode(s) and other user-enabled low power conditions." We fetched that page on 17 November 2026 and again through the Internet Archive's 8 November 2026 capture: both return a 2.9 KB shell reading "You need to enable JavaScript to run this app." No wattage is present in the served HTML, and we were not able to read one before publication. [OBSERVED FACT — what we could retrieve, not a finding about what Microsoft discloses to a human with a browser.]

The voluntary agreement handed standby to the regulation

All three companies are signatories of the Games Consoles Voluntary Agreement, recognised by the European Commission in 2015. Nintendo's compliance page says its information is provided "to meet the self-regulatory requirements of the games console voluntary agreement", and the agreement's own site publishes every signatory's yearly Product Compliance Report and the independent inspector's annual report.

Those filings contain the quiet admission that makes the discrepancy consequential. Footnote 1 of Nintendo's 2025 report, dated 06/02/2026, states:

"From 9 May 2025, requirements for reporting on the default regulatory standby mode(s), such as networked standby, standby, off mode or another equivalent condition(s); their relevant APD time(s); and their power consumption(s), are specified under Commission Regulation (EU) 2023/826 and are not in scope of the Games Console Voluntary Agreement."

The inspector's 2025 report, issued 15 May 2026, confirms the alignment and records that it reviewed the filings and that "no non-compliances or issues were reported." It also explains the scope of the agreement's own power caps: they "apply to in-scope consoles consuming more than 20W in Active Gaming mode." Nintendo declares the Switch 2 at 19 W in active gaming — under the threshold. [OBSERVED FACT]

Put those two facts together and the Switch 2's low-power behaviour is governed by exactly one instrument: Regulation (EU) 2023/826, whose disclosure page says 2.0 W and whose spec page says 5 W. The voluntary agreement, which used to be the forum where such numbers were argued about, has formally stepped out of the way. [INFERENCE]

The 2027 deadline nobody has priced

There is a second date in the same annex. From 9 May 2027, off mode must fall from 0.50 W to 0.30 W. Nintendo declares 0.5 W for the Switch 2 powered off, and 0.3 W for the Switch 1 before it. That is compliant today, exactly at the current ceiling — and it leaves Nintendo under six months to change the hardware or its firmware.

Sony's page shows what the other side of that line looks like: its "Low power Rest" state is already measured at 0.30 W. Rest is not off, and the two are not directly comparable — but the direction of the design work is visible in the published numbers, and it is not visible in Nintendo's. [INFERENCE]

The strongest case that this is nothing

The fair reading of Nintendo's position is on its own page, and it is not frivolous:

  1. The 5 W state is opt-in. Nintendo says of the wired connection: "You can choose to maintain connection to the internet or prioritise low power consumption (default for Nintendo Switch 2)." The console as shipped sleeps at 0.5 W. The high figure belongs to a setting most owners never touch.
  2. It may not be "networked standby" at all. The regulation defines networked standby as a condition in which the equipment can resume a function "by way of a remotely initiated trigger from a network connection". If the Switch 2 in that state only completes downloads already queued rather than answering a remote wake, Nintendo could argue the state is outside the definition — or that the console is performing a function, not standing by.
  3. "Max. 2.0W" may be the limit restated, not a measurement. Every figure in that block is a round regulatory number — 2.0 W, 0.5 W, 20 minutes. It reads like a declaration of conformity rather than a measured value.

Each answer creates a worse problem than the one it solves.

If the state is not networked standby, then it is a standby condition offering a reactivation function, and the ceiling is 0.50 W — which 5 W exceeds tenfold. If the console is performing a function in that state, then Nintendo is publishing an annual energy figure (50 kWh) for a state it says is transient. And if "Max. 2.0 W" is the legal limit restated rather than the measured consumption, then the page does not discharge Annex III 3(a)(1), which asks for "the power consumption expressed in watts" — the actual number, which is the entire point of the duty. [INFERENCE]

On the first answer, the opt-in defence has a specific weakness. Annex III 1(c) caps the consumption of networked equipment in networked standby. It contains no carve-out for modes a user must enable. Annex IV, the measurement method, tests the port-connected condition and checks reactivation — the worst case the product offers, not the friendliest setting in the menu.

What would settle it

Three documents would end the argument, and none of them is public:

  • Nintendo's technical file for BEE-001, which Annex III 3(b) requires to state the category of equipment, whether it is HiNA, and the consumption for each mode.
  • A market-surveillance measurement under Annex IV, with the LAN port connected and a remote trigger attempted.
  • Commission guidance on whether a user-selectable high-power network mode counts as a networked standby condition under 2023/826. We could find none published.

Until one of those exists, the honest position is this: Nintendo has published two numbers for the same low-power state of the same console, they differ by a factor of two and a half, and the higher one is the one that appears on the page a shopper reads.

Method, and what this does not prove

Every figure above is quoted from a document published by the company it concerns or by the regulator, fetched on 17 November 2026. No console was measured for this article; we did not have one.

What this proves. Nintendo's EU-facing compliance page declares a maximum of 2.0 W in networked standby, including the all-ports-connected case, for the Switch 2; Nintendo's EU-facing technical specifications declare 5 W in sleep with the wired LAN connection maintained for the same model, with 50 kWh a year instead of 14. Both statements are live and public. Sony, under the same duty, publishes 0.84 W for the equivalent state. [Confidence: 95]

What it does not prove. That the Switch 2 is unlawful. That depends on a classification — whether the wired-LAN sleep state is "networked standby" as defined — which only a market-surveillance authority or a court can settle, and on whether Nintendo's technical file shows a different measurement basis than its consumer page. [Confidence that the 5 W state falls within the 2.00 W cap: 75]

What it does not prove, part two. That Microsoft is or is not compliant. We could not read its published figures. We are not suggesting it is in the same position as Nintendo; we are saying we do not know, and that a disclosure routed through a JavaScript application is a disclosure a journalist, a researcher or a crawler cannot check.

The compliance record is otherwise clean and public: the independent inspector found no non-compliances in the 2025 cycle, and Nintendo filed its report on time. The contradiction described here sits between two of Nintendo's own web pages, in the one area the voluntary agreement has explicitly handed to the regulator.

Sources


All documents fetched on 17 November 2026. The archived Xbox support capture referenced in the text was retrieved from the Internet Archive's snapshot of 8 November 2026.

More Guides