TMR is closing the gap on Hall-effect. It hasn't caught it.

Five tracker readings in a row, all pointed the same direction, is not noise. On thock's Trends Tracker, TMR (tunnel-magnetoresistance) switches have posted an "up" score every single week from W38 through W42: 55, 63, 71, 85, 90. That last number is the second-highest score on the entire 2026-W42 snapshot — trailing only Hall-effect/Rapid Trigger, which sits at a ceiling-pinned 100. Every one of those five weekly rows has credited its movement to the same single piece of coverage: Keychron's ceramic-TMR Q16 HE 8K launch. That's one product driving a category-wide number for over a month. It's worth asking what's actually moving, and whether TMR is doing anything more than riding one vendor's press cycle.

Hall-effect and rapid-trigger sensing already won the broader magnetic-switch conversation this year, holding a flat score near 100 for weeks running rather than still climbing toward it. TMR arrives into that landscape as the second wave, not the first — a technology trying to carve out room inside a category Hall-effect already defined. The question this piece asks is narrower and more answerable than "will TMR win": is the five-week climb a real second act, or just one brand's launch calendar dressed up as a trend.

Five readings, one shrinking gap

Line the two scores up and the shape of the story gets sharper. Hall-effect/Rapid Trigger was already near its ceiling at W38 — 97 — and has crept up one or two points a week since, landing at 100 in W42 per the same snapshot. TMR started that window at 55, a full 42 points behind, and has gained ground every week since: 63, 71, 85, 90. The gap is now 10 points. That's catching up from behind, not overtaking — Hall-effect hasn't lost a single point in this window, it's simply run out of room to climb further on a 0–100 scale. TMR is the one doing the work.

What a tunneling junction is reading that a Hall sensor isn't

Both Hall-effect and TMR switches work the same way at the mechanical level: a magnet rides down the stem, and a fixed sensor on the PCB reads its approach instead of waiting for a physical contact to close. Where they differ is in what that sensor actually measures. A Hall sensor reads a voltage that the magnetic field generates directly across a semiconductor element. A TMR sensor instead reads the electrical resistance of a tunnel junction — two magnetic layers separated by a barrier a few atoms thick — and that resistance changes as the approaching magnet rotates one of those layers. Vendors marketing TMR switches claim that arrangement resolves smaller changes in magnet position at the same travel distance than a Hall sensor can. That's a vendor-stated claim at this point, not an independently verified one, and thock's own deep dive into TMR's sensor mechanism covers the physics — and a direct rebuttal from Hall-effect's own category leader — in far more depth than belongs in a trends piece.

Who's actually moving

A single vendor's launch anchoring every tracker row is a fair reason for skepticism, but TMR isn't a single-vendor story — it's just that the tracker's linkage hasn't caught up to the second data point. Cherry's gaming sub-brand XTRFY got there first: on July 3, 2026, it shipped the K5 Pro Compact around a MK Crystal TMR switch, which thock covered at the time as a legacy MX manufacturer hedging into a category it had already lost ground in. That was a catch-up move from a brand whose mainline MX switch line was bleeding tracker share to magnetic switches generally. Keychron's September 7 Q16 HE 8K launch is a different kind of signal entirely — it's the category's own Hall-effect standard-bearer choosing to bet on the technology that's supposed to supersede the one it built its own dominance on, from a position of strength rather than catch-up.

That two brands at opposite ends of the market pressure spectrum both landed on TMR independently is more interesting than either launch alone. It suggests the appeal isn't limited to a vendor trying to manufacture a headline. Keychron's TMR bet isn't even its only hedge on switch technology this year — the V6 Ultra HE's Nova Socket lets a single PCB run mechanical or TMR switches per key, the same unwillingness to commit to one sensor technology dressed up as a hardware feature instead of a launch.

Closing the gap, not crossing it

None of this should read as a call that TMR is about to replace Hall-effect. thock's own look at the Hall-effect/Rapid Trigger plateau made the case in August that a flat line at the top of the tracker's scale is what "already won" looks like, not what "starting to lose" looks like — and nothing in the five weeks since has changed that read. Hall-effect hasn't ceded a single point this window; it's gained two, from 98 to 100. A five-week climb on one emerging sensor technology, anchored almost entirely to one brand's launch cycle, is a real signal worth naming — and it's worth remembering that a tracker-score gap closing isn't the same thing as a capability gap closing; Wooting's own rapid-trigger firmware lead over the hardware that's since caught up to it is the clearest example on this site of that distinction mattering. It is not a sustained, multi-brand verdict yet, and the honest version of this piece says so plainly: it's too early to call this a shift away from Hall-effect. Keychron's Kickstarter enthusiasm can cool the way any single-product bump can; XTRFY's TMR bet still hasn't produced a second SKU in the three months since, by this site's own count. The tracker row's reliance on one article is itself evidence the category's coverage is still thin.

What would change that read is a third brand committing to TMR independent of either Keychron's or Cherry's launch cycles, or Keychron following the Q16 HE 8K with a second TMR board the way it did twice over with Hall-effect before that technology became the category default. Either would turn a five-week climb into the start of a pattern instead of one company's press cycle measured five times. The Trends Tracker carries that row forward week to week; whether the gap keeps closing, stalls short of Hall-effect's ceiling, or reverses is a question the next several snapshots will answer on their own schedule, not one this piece needs to guess at today.