A retrospective review of a dead watch is a strange exercise. You cannot recommend it, you cannot test it against a 2026 reference device because it will not turn on, and the cloud that hosted every interesting bit of data has been gone since the last day of 2016. What you can do is look at what people said about the Basis Peak when it was new, compare those reviews against what the device actually was, and pull out what the experience tells us about how wrist wearables grew up.

We owned two Peaks at this office between 2014 and 2016, one in matte black and one in the brushed stainless trim. Both went back during the recall. The notes below are partly ours and partly stitched together from the contemporary press, archived support threads, and the Reddit posts from owners that are still indexed.

The case: a tank with a touchscreen on top

The Peak’s body is the part most reviewers got right. It was a slab of stainless steel, 41 mm wide, with a square 1.26-inch OLED touchscreen recessed under a Gorilla Glass face. The matte version had a black PVD coating that scratched if you looked at it; the polished stainless option held up better but looked, fairly, like a small steel paving stone on your wrist.

It was waterproof to 5 ATM, which mattered. The original Basis B1 you had to take off in the shower; the Peak you wore swimming. Engadget’s Edgar Alvarez, in his November 2014 review, called it “borderline indestructible.” Brad Molen at the same outlet called the build “the best part of the watch.” That tank reputation is the thing Peak owners still remember most fondly. We have a Peak that fell off a bike at 18 mph onto pavement; the screen is unmarked.

The trade-off was that it was heavy and it was hot. At roughly 73 grams it was about twice the weight of a Fitbit Charge HR (which was the direct competitor), and the steel back trapped heat against the skin during summer runs. The straps were proprietary, which became its own problem when the rubber inserts started cracking. Replacements stopped being sold the moment the recall happened.

The touchscreen, and the cold-weather problem

The Peak’s interface used four edge gestures: swipe up for sleep data, swipe down for activity, swipe left for heart rate, swipe right for notifications. There were no buttons. The implementation was the single thing most reviewers complained about.

The screen worked fine indoors. Take it outside in winter and the response went soft. Joanna Stern, then at The Verge, called the touch sensitivity “the most frustrating part of the watch” in her December 2014 review. We saw the same thing wearing one in Boston in January 2015: gloves off, swipe right, nothing happens, swipe again, you have somehow ended up in the heart-rate screen instead of notifications. Intel never fully fixed this. The November 2015 firmware update that later became infamous for the overheating was, partly, an attempt to improve touch responsiveness in cold conditions.

This is the thing that dates the Peak more than anything else. Modern Garmins, Apple Watches, and Fitbits all use capacitive layers and software that have largely solved cold-weather lag. In 2014 it was a real engineering problem and the Peak shipped with it half-solved.

The optical heart-rate sensor: fine for life, bad for HIIT

The Peak measured heart rate continuously, twenty-four hours a day, before that was normal. The Fitbit Charge HR launched a few weeks later with a similar promise. The Apple Watch did not ship with continuous HR until later firmware updates in 2015.

For resting and sub-threshold cardio the Peak’s HR was good. Lauren Goode, then at Re/code, ran the Peak against a chest strap on steady-state runs and reported readings within a few beats per minute. Ars Technica’s Casey Johnston had a similar experience for everyday wear.

What the sensor could not do was track rapid heart-rate changes. During interval training, kettlebell circuits, or any session with sharp upward and downward swings, the optical signal would drop, lock onto a stale value, or report a steady 110 bpm while you were clearly above 160. This was a hardware-and-firmware limitation common to nearly every wrist optical HR sensor of that generation. Mio, Fitbit, and TomTom all shipped versions of the same problem. Intel happened to market the Peak as a workout watch, and reviewers held it to that bar. Engadget’s review made the point bluntly: good for runs, frustrating for anything that spikes your pulse.

Cold weather made it worse. Reduced peripheral perfusion plus the touchscreen lag plus a sensor designed for warmer skin meant a January HIIT class on the Peak was, basically, a wasted session. Our deeper write-up on this is in Basis Peak heart-rate accuracy.

Four days of battery, when you actually got four days

Intel claimed up to four days of battery life. In our use we got three and a half to four with notifications off and the watch face on tap-to-wake. Turn on persistent notifications and the always-on time display and you were charging every other day. This still beat every smartwatch on the market in late 2014 by a wide margin. The Pebble managed similar numbers with an e-paper screen; the original Apple Watch, when it shipped six months later, lasted about a day.

The charging clip was a magnetic four-pin contact that snapped onto the back of the watch. It worked. It was also incredibly easy to lose, and Intel charged $25 to replace it, which felt punitive even at the time.

The Peak’s battery is also the part of this story that turned the watch from “fine, niche product” into “recalled health hazard.” We will not relitigate the recall here. The recall page covers the CPSC sequence. The relevant detail for a review is that the November 2015 firmware update, the one that tried to fix the touchscreen and heart-rate issues, also changed charging behavior in a way that, in some units, caused the cell to fail catastrophically. The hardware was fine until the software changed.

The app, Habits, and the Pattern system

The Basis app was where the Peak tried to be more than a tracker. Instead of giving you raw step and HR data, the app surfaced “Habits”: automatically suggested behavior goals, scored on a Pattern points system. Wear the watch for twelve hours, get points. Sleep at the same time three nights in a row, get points. Move every hour during the workday, get points.

This was the most genuinely novel thing about the Peak, and almost no contemporary reviewer engaged with it seriously. Most reviews mentioned Habits in passing, scored the watch on heart rate and steps, and moved on. With hindsight the Habits framework looks like a clear ancestor of Oura’s readiness score, Whoop’s strain-recovery model, and the Apple Watch’s stand-and-move rings. The Peak shipped it in 2014. It just shipped it inside an app most reviewers found cluttered.

The app’s data export was always limited. You could view trends in the web dashboard and pull CSVs of activity, but the underlying physiological data (the raw GSR traces, the temperature curves, the per-minute HR samples) was never exposed to users. When Intel shut the servers down on December 31, 2016, anything that had not been screenshotted was gone. We make a habit of noting this on every wearable we cover; the Basis Peak is the canonical case study.

Sleep tracking, which was years ahead

The Peak’s sleep mode was the feature we still defend. It used four signals (heart rate, skin temperature, perspiration via GSR, and motion) to infer light, deep, and REM stages. Every other wearable in late 2014 was doing sleep tracking with motion alone, which was about as accurate as guessing.

In contemporary side-by-side tests against actigraphy and consumer-grade polysomnography, the Peak’s totals (total sleep time, time in bed) were respectable. Stage classifications were more uneven. Specifically, the Peak tended to overcount REM and undercount deep sleep, which is the same direction in which most modern consumer trackers err. There is a longer treatment in Basis Peak sleep tracking.

The thing the reviewers missed was that the multi-signal approach was the right idea. Oura built a company around it. The Apple Watch added wrist temperature sensing in 2022 and used it for cycle tracking and, eventually, sleep stage refinement. The Whoop strap, the Garmin Body Battery, and the Fitbit Daily Readiness Score are all downstream of the same insight: motion alone is not enough. The Peak got there in 2014 and got almost no credit for it at the time.

What the contemporary reviewers got right and wrong

Looking back at the original reviews from Engadget, The Verge, Wired, Ars Technica, CNET, and TechCrunch, there is a clear pattern.

What they got right:

  • Build quality was genuinely good. Nobody overstated it.
  • The four-day battery was a real, meaningful advantage over the competition.
  • Cold-weather touchscreen lag was a flaw, not a quirk.
  • Optical HR under load did not work, and reviewers said so.
  • Notifications: one-way, forgettable.

What most missed:

  • The Habits and Pattern system was the most interesting idea in the watch, and almost no review covered it in detail.
  • Multi-signal sleep tracking was ahead of the market on its merits.
  • The cloud dependency was a structural risk. Nobody flagged this at the time. Two years later it was the reason the device became unrecoverable.
  • The November 2015 firmware update was a warning sign in retrospect. Several owners reported the watch running hotter immediately after the update; very few reviewers re-reviewed the device.

The cloud-dependency point is the one we keep coming back to. Reviewing a connected wearable in 2014 the way you would review a wristwatch, as a self-contained object, missed the part of the product that mattered most. The Peak’s value lived on Intel’s servers. When those servers went away, the device lost its meaning. This is the dynamic we now treat as default. See our wearable sensor accuracy explainer for the broader framework.

How the Peak compares to what came next

It is tempting to draw a straight line from the Peak to the Fitbit Charge HR to the Apple Watch to the Oura Ring, and most of that line holds. The Peak’s sensor stack was richer than the Charge HR’s and shipped earlier. Apple’s first watch had a more polished UI and a worse battery. Oura turned the multi-sensor sleep idea into a category.

The Peak’s lessons for the modern market are not flattering for the modern market. A 2026 mid-range fitness tracker is more accurate during HIIT, has a faster touchscreen, and syncs to a phone in seconds. It is not necessarily smarter about your health. The Habits system, with all its faults, was a real attempt at translation, at turning numbers into a behavior the user could change. Most of the Apple, Fitbit, and Garmin apps in 2026 still mostly show you numbers. If you are picking between a Fitbit and a Garmin today, you are picking between two devices that, in interpretive terms, do less than the Peak tried to.

What we would not bring back is the cloud dependency, the proprietary charger, the cold-weather touchscreen, and obviously the battery. The history of the Peak is the history of a product that knew where wearables were going and got there with the wrong batteries and the wrong corporate parent. For why that parent matters, see why the Basis Peak failed. For modern devices that try to do what the Peak did, see Basis Peak alternatives.