When Microsoft unveiled Windows 11 in June 2021, most of the excitement around Snap Layouts, a redesigned Start menu, and Android app support lasted about as long as it took people to run the PC Health Check tool. What followed was one of the most contentious hardware decisions in Windows' history: a mandatory requirement for TPM 2.0, paired with a strict list of supported processors. Millions of PC owners discovered, often for the first time, that a chip they'd never heard of now stood between them and the newest version of Windows.
Five years on, the numbers make clear this was never a minor footnote. Analyst firm Canalys estimated that roughly 240 million PCs worldwide couldn't meet Windows 11's requirements — machines that were still working, still useful, and suddenly locked out of the upgrade path. That single statistic has shaped nearly every debate about Windows 11 since: is this a genuine security upgrade, or a hardware refresh cycle dressed up as one?
TPM 2.0: Security Requirement or Hardware Barrier?
First, a clarification that gets lost in a lot of the online arguing: TPM 2.0 was never an enterprise-only rule. Microsoft's official Windows 11 minimum specifications list it alongside Secure Boot-capable UEFI firmware, 4 GB of RAM, and 64 GB of storage — requirements that apply to every consumer PC, not just business fleets.
The Trusted Platform Module itself is a small security chip (or, increasingly, a firmware-based equivalent like Intel PTT or AMD fTPM) that stores encryption keys and credentials in a way that's isolated from the main operating system. It underpins BitLocker drive encryption, Windows Hello facial and fingerprint login, and measured/Secure Boot, which checks that a machine is booting trusted code rather than something a malware author slipped in.
Microsoft has been consistent and specific about why it drew this line. David Weston, the company's director (later partner director) of Enterprise and OS Security, put it plainly when Windows 11 launched: TPM 2.0 provides "security backed by a hardware root-of-trust," calling it a "critical building block" for protecting user identities and data through Windows Hello and BitLocker. In a later interview with CRN, Weston explained the deeper rationale for making it mandatory rather than optional, as it had effectively been in Windows 10: "What we learned from 10 is, if you make things optional, people don't turn them on. They assume that if it was necessary, it would be on." Microsoft's stated goal was to make advanced protection the default rather than something only security-conscious users would bother enabling.
There's a threat-model argument behind this too. Weston has pointed to the scale of attacks Microsoft observes — the company has cited processing roughly 65 trillion signals a day and blocking well over a thousand password attacks per second — as justification for raising the baseline rather than leaving hardware-backed protection as a checkbox few people ticked.
From a pure security-engineering standpoint, that argument holds up. Security researchers generally agree that hardware-rooted protections are meaningfully harder to defeat than software-only equivalents. The friction is entirely on the other side of the ledger: a security chip doesn't ask whether your seven-year-old laptop can still browse the web, edit documents, and run a video call just fine. It simply draws a line, and everything on the wrong side of it becomes, in Microsoft's own support terminology, "unsupported."
Windows 11 and the Hardware Upgrade Cycle
Windows 11 officially launched on October 5, 2021, introducing not just the TPM 2.0 requirement but a specific list of supported CPU generations — eighth-generation Intel Core processors and AMD Ryzen 2000-series chips and newer, both of which happened to be roughly the point where TPM 2.0 support became standard on motherboards.
That timing wasn't incidental, according to Weston, who told Dark Reading the CPU cutoff was less about pure security and more about overall platform reliability: "This is not a security-centric floor. This is a general experience [CPU] floor to make sure that Windows 11 meets expectations." Every processor Microsoft certified for Windows 11 already included TPM 2.0 support, he noted — the CPU list and the security chip requirement were two sides of the same decision.
The practical effect was still the same regardless of the reasoning: a large population of machines that ran Windows 10 comfortably were, overnight, outside Microsoft's official upgrade path. Microsoft does distinguish between "unsupported" and "impossible" — registry workarounds exist to force an install on ineligible hardware — but doing so forfeits official support and, in many cases, future feature updates.
The Adoption Numbers, Five Years Later
The slow, uneven climb of Windows 11 adoption is itself a data point in this debate. For years, uptake lagged well behind previous Windows releases specifically because of the compatibility wall. As late as December 2025, Windows 11 held only around 50–63% of the Windows desktop market by various measures, even after Windows 10's official retirement.
That changed fast once the October 14, 2025 end-of-support deadline actually arrived. StatCounter's figures show Windows 11 climbing to roughly 72–73% of global Windows desktop share by February 2026, with Windows 10 dropping into the low-to-mid 20s. Microsoft told investors on its FY2026 Q2 earnings call that Windows 11 had crossed one billion active devices in 1,576 days — faster, in raw device count, than Windows 10 reached the same milestone, though Windows 10 did it without a hardware compatibility wall in its way.
The adoption curve also isn't uniform. Regionally, StatCounter data shows North America and Europe adoption exceeding 75%, while parts of Asia-Pacific and other emerging markets — where hardware refresh cycles are naturally slower — sit closer to 60–68%. Enterprise data tells a similar story: Lansweeper's scans of managed business endpoints found Windows 10 still running on roughly 17–23% of scanned Windows clients in mid-2026, with healthcare and pharmaceutical organizations carrying the heaviest holdout share of any sector at around 23%, followed by consumer/retail and manufacturing. Gaming hardware, unsurprisingly, moved fastest of all — the Steam Hardware Survey showed Windows 11 crossing 50% adoption among PC gamers roughly ten months before it crossed the same mark on StatCounter's all-desktop measure, reflecting that gaming rigs tend to run newer, TPM-compliant components.
For anyone who still can't or won't move, Microsoft's Extended Security Updates (ESU) program is the main safety net. The consumer ESU track (which can be claimed for free using a Microsoft account, or paid for otherwise) runs through October 2026. Commercial ESU pricing starts at $61 per device in year one and doubles annually for up to three years — a structure explicitly designed to make migrating cheaper than delaying.
What About Windows 12?
Here's where it's important to separate speculation from fact: there is no official Windows 12 announcement, and no published Windows 12 hardware requirements, from Microsoft as of this writing. The company has continued developing and shipping feature updates to Windows 11 rather than confirming a successor.
That means any claim about Windows 12 mandating a specific CPU generation, TPM version, RAM floor, or dedicated AI hardware (Copilot+ style NPU requirements, for instance) should be read as speculation, not fact — however plausible it sounds given Microsoft's current AI push.
What Windows 11's rollout does offer is a precedent. If Microsoft ever ships another major version with a stricter hardware baseline, the pattern of the last five years — a rocky launch, years of lagging adoption, a scramble once the predecessor's support clock runs out — is the most concrete evidence available for how such a transition might unfold. Until Microsoft says otherwise, that's a reasonable expectation to hold, not a headline to write.
The 4 GB RAM Controversy
A separate claim that circulates periodically is that Microsoft quietly lowered Windows 11's minimum RAM requirement from 8 GB down to 4 GB. That's not accurate. Microsoft's published minimum has listed 4 GB of RAM since Windows 11's original requirements sheet, with 8 GB listed as the recommended amount for a usable long-term experience. Nothing was walked back — the two figures have simply always existed side by side and get conflated.
The more useful question isn't about what changed, but about what "minimum" actually delivers in practice. A machine can meet Windows 11's stated 4 GB floor and still struggle the moment a modern browser, a background sync client, and a video call are open simultaneously. If Windows 11 is engineered to function within that minimum, as Microsoft's spec sheet implies, the more defensible criticism is why ordinary, non-edge-case usage on 4–8 GB systems so often feels resource-constrained rather than smooth.
Windows 10 Has Already Reached the End of Its Normal Support
Windows 10 officially reached end of support on October 14, 2025. Microsoft no longer ships routine feature updates, non-security bug fixes, or free technical support for it. Consumers can enroll in Extended Security Updates for continued critical and important security patches, currently slated to run through October 2027 for those who take the free enrollment path — a program Microsoft extended after initial criticism that the original one-year window was too short.
That leaves owners of Windows 11-ineligible hardware with a narrower set of choices than they had in 2021:
- Buy new, Windows 11-eligible hardware.
- Stay on Windows 10 under Extended Security Updates, on borrowed time.
- Install Windows 11 through an unsupported workaround, accepting the loss of official support.
- Switch to Linux, or a lightweight option like ChromeOS Flex.
- Move to a different platform, such as macOS, entirely.
Could This Push Users Toward Linux?
Every Windows hardware controversy tends to produce the same reflexive response online: "just switch to Linux." The reality is more nuanced but not baseless. Linux distributions built for constrained hardware — Lubuntu, antiX, and similar lightweight options — can genuinely extend the useful life of a machine that Windows 11 has ruled out. Google's ChromeOS Flex offers a similar path for less technically inclined users, effectively repurposing an ineligible Windows PC into a fast-booting, low-maintenance Chromebook-like device.
Linux's real edge here is choice — the ability to pick a distribution and desktop environment matched to a machine's actual capabilities, rather than being tied to one vendor's hardware baseline. But it's not an unqualified upgrade. Windows still leads decisively in commercial software compatibility, PC gaming (Steam's own survey data puts Windows at roughly 94–96% of the platform's gaming population), and enterprise device management tooling. Linux adoption on the desktop remains in the low single digits by most measures, and driver support, proprietary application compatibility, and a real learning curve remain genuine obstacles for non-technical users.
The E-Waste Argument
This is where the TPM 2.0 debate stops being an abstract UX complaint and becomes a measurable environmental one. Canalys's 240-million-PC estimate — about one-fifth of the entire Windows 10 installed base at the time — has become the anchor figure for nearly every piece written on the subject since 2023. Canalys illustrated the scale vividly: stacked together, that many laptops would form a pile roughly 600 kilometers taller than the distance to the moon.
Analysts at the firm were blunt about why refurbishment isn't a clean escape valve. As Canalys put it in its original research note, the channel's growing capacity for circular-economy resale doesn't help much "if partners will not be able to refurbish and resell PCs unsupported by Windows 11" — a device without a supported OS path is a much harder sell, even in working condition. Ish Jessop, an analyst who has written on the topic, made a related point to TechNewsWorld: those 200-million-plus devices "don't become obsolete after Win10's end of support," and plenty of consumers keep using unsupported machines regardless, which raises separate concerns around unpatched security exposure rather than eliminating the waste question entirely.
More recent estimates have pushed the number higher. A report cited by PIRG (the U.S. Public Interest Research Group) put the ineligible-device count closer to 400 million once the full run-up to Windows 10's end-of-life is accounted for, translating to roughly 1.6 billion pounds of potential e-waste — with electronics recycling researchers separately estimating the Canalys-scale figure alone at over a billion pounds. However the exact number lands, the pattern across independent estimates is consistent: this is not a rounding error, and academic researchers have started treating it as a case study in the environmental costs of software-driven obsolescence, alongside broader debates over right-to-repair and extended producer responsibility.
Microsoft's public position has stayed close to Weston's original framing — that the hardware baseline exists to improve performance, security, compatibility, and reliability across the board — and the company doesn't dispute the environmental critique so much as treat it as a tradeoff against a security posture it considers non-negotiable. Whether that tradeoff was struck in the right place is, at this point, less a technical question than a values one.
Windows Has a Long History of Criticism Over Updates
None of this happens in a vacuum. Windows is a genuinely massive software platform running on an almost incomprehensible range of hardware and third-party software combinations, and update-related bugs, driver regressions, and feature churn have dogged nearly every release cycle for a decade. That doesn't make every release "unfinished," but it does explain why a segment of the user base has come to expect a familiar rhythm: release, discover problems, patch, introduce new changes, discover new problems. Whether that's a fair characterization of any single update is debatable — but the underlying fatigue driving the perception is real, and it colors how skeptically people receive Microsoft's next hardware mandate, whatever it turns out to be.
Privacy and Microsoft's Increasing AI Focus
Layered on top of the hardware debate is a parallel, related one about how much Windows now reaches outward — cloud account integration, telemetry, and a steadily expanding set of AI features, from Copilot's system-wide integration to Recall's local activity indexing on Copilot+ PCs. Some of this is genuinely useful; AI-assisted search, accessibility tooling, and coding support have real, demonstrated value. The recurring criticism isn't that AI is inherently bad for Windows, but that users increasingly want more granular control over what's enabled, how deeply it's woven into the OS, and what data it touches by default — control that TPM- and NPU-gated features tend to remove by tying them to specific certified hardware tiers rather than leaving them opt-in everywhere.
The Bigger Problem Is Hardware Longevity
Strip away the individual arguments about RAM minimums, CPU lists, and security architecture, and the TPM 2.0 controversy is really about one underlying question: how long should a piece of working hardware reasonably last before software support decides otherwise?
Microsoft's position — that modern threats justify a modern hardware-backed security floor, and that supporting every legacy chipset indefinitely carries real engineering and security-testing costs — has genuine technical merit, and security researchers broadly agree with the underlying premise even when they question the rollout. The counter-position — that a computer capable of five or seven more years of everyday use shouldn't be rendered obsolete by a policy decision, with the environmental cost measured in hundreds of millions of discarded devices — also has real weight, and it's the argument that analyst firms, environmental researchers, and a meaningful share of the Windows user base have converged on.
Both can be true at once. The real work is in the balance between security, performance, compatibility, sustainability, and consumer choice — and Windows 11's launch showed just how unevenly that balance can land when a hardware line gets drawn.
Final Thoughts
Microsoft's TPM 2.0 requirement wasn't simply a scheme to force new hardware purchases — the security rationale, articulated repeatedly and specifically by Microsoft's own security leadership, holds up on its technical merits. But the five years since launch have also demonstrated, in very concrete numbers, the cost of tying OS support that tightly to a hardware generation: hundreds of millions of devices sidelined, a genuinely difficult e-waste problem, and years of slower-than-expected adoption that only resolved once Windows 10's support clock ran out and forced the issue.
As of late 2026, Windows 12 remains officially unannounced, so any predictions about its requirements are still speculation, not fact. What's no longer speculative is that Windows 11 established a stricter hardware baseline, Windows 10 has ended normal support, and Microsoft's approach to hardware-gated security has permanently changed how a lot of consumers think about how long a PC is supposed to last.
For people on modern hardware, none of this registers as a problem. For people who owned a perfectly good machine in 2021 and watched it get quietly reclassified as obsolete, the underlying question hasn't gone away: should an operating system vendor decide when your hardware is done, or should you?
That debate will likely outlast the TPM 2.0 controversy that started it.
