Someone who bought a Ring Pro to see their sleep data will, by the end of September, own a game controller as well, on Ultrahuman's own account [8]. The shipped portion of the computer pitch therefore runs on the Nordic Semiconductor chips already inside the ring, because the Qualcomm-based device the company describes is still being built [4].
Price the rest against what the business sells now. $365m against a $140m annual run rate is about 2.6 times revenue [1], and the $200m target for January 2027 asks for roughly 43% more than the current run rate [2]. Units back the direction of travel: about 800,000 rings sold to date against roughly 700,000 in February, on the order of 100,000 since [17][3].
Both of those are sales figures rather than use figures [16][17]. Retention on the finger, depth of use and time to a first useful reading are not among the numbers Ultrahuman gives, and for anyone deciding whether to build something ring-native, rings sold is the wrong denominator. The number that would settle it is how many of the 800,000 are still worn daily.
The uses Mohit Kumar floats are mostly not health ones: a pointer or mouse, a game controller, a car key, an interface for talking to AI [7]. The other name on the cap table points the other way. Labcorp is an investor because of Blood Vision, the preventative blood-testing service Ultrahuman spent two years building [18], and the two are exploring whether blood-flow signals from the ring, read against blood-test results, can flag cardiovascular and fertility risk [19]. Qualcomm's interest is easier to read: it launched a white-label toolkit and a chip line for AI glasses in June [11], and Qualcomm Ventures has backed Ring and Fitbit before [21].
Geography sits underneath all of it. The US was about 45% of revenue this quarter and India about 11% [14], so on the current run rate roughly $63m a year comes from a single market [6], while some of the new money goes into physical stores in India and the UAE, where Ultrahuman says offline space sells rings that online listings do not [15].
Two tests sort the four floated uses. The first is whether the use needs a signal only skin contact gives. The second is whether the work has to happen on the finger rather than on the phone in the pocket. A car key needs neither, so the ring would compete there on convenience alone. Gesture input for glasses needs the on-device half and none of the physiology. A game that responds to heart rate and skin temperature needs both, and Kumar's line for it is the only item on his list a handset cannot answer: "A game controller never reads your heart rate and your temperature, but this one does" [9]. For a team weighing whether to write code for a ring, the useful thing to track is whether anything at all lands in that both-tests corner. The other three corners are competing with hardware the user already carries.
