Inside Your Dive Computer: the Algorithms That Keep You Safe

Right now, if there is a dive computer on your wrist, it is quietly running a calculation the diving world took a century to work out. On every dive it models how nitrogen seeps into sixteen or more theoretical tissue compartments, tracks how fast each one loads and clears, and keeps answering a single question: is it safe to go up, or do you owe the water a stop first?

That answer comes from the algorithm, the rulebook that decides what 'safe' actually means. Change the algorithm and you change the whole dive: your no-decompression limits move, your ascent ceilings shift, the penalty you carry into your next dive grows or shrinks.

Here is the part that catches new divers off guard. Put two people in the water on an identical profile, same depth, same time, but on different computers, and one can read fifteen more minutes of no-deco time than the other. Nothing is broken. You are watching two algorithms read the same physics and reach two different verdicts.

It begins in 1908, with a Scottish physiologist named John Scott Haldane and a problem that was crippling tunnel workers. Men digging beneath rivers kept surfacing bent double in agony, and nobody could say why. Haldane's leap was to stop treating the body as one thing and start treating it as many: a set of compartments, each absorbing and releasing nitrogen at its own speed. Keep the ratio of absorbed pressure to ambient pressure under a safe limit in every compartment, he argued, and the pain never arrives.

The U.S. Navy spent decades refining this through chamber tests and open-water trials. By the 1960s, their tables were standard issue for military and recreational diving. PADI's original Recreational Dive Planner was built on research commissioned from that same lineage.

Then, in the 1980s, Albert Bühlmann at the University Hospital of Zurich took the model further than anyone had. He published ZH-L16: sixteen tissue compartments, mathematically defined M-values (the maximum nitrogen pressure each compartment can safely hold), and years of clinical testing behind it. It went on to become the most studied decompression model in history.

Meanwhile, other researchers were asking a different question. What if the real danger is not dissolved gas in tissues, but bubbles: tiny pre-existing microbubbles that grow under pressure changes? That question led to two bubble-based models, RGBM and VPM.

Almost every dive computer you can buy runs one of four core approaches. Each one places a different bet on where the real danger hides, and each produces different no-deco times and deco schedules because of it. Here is what separates them, without the jargon.

Bühlmann ZHL-16, and its variant with Gradient Factors, is the dominant family. It tracks dissolved nitrogen across sixteen compartments and uses M-values to define the safe ceiling in each one. Gradient Factors (GF) are a user-adjustable layer added later: they let the diver dial conservatism up or down. A low GF setting adds a safety cushion above the raw Bühlmann limit; a high GF follows the limit more closely. Most technical divers use something like 30/85 or 40/85 as a starting point.

RGBM (Reduced Gradient Bubble Model) was developed by physicist Bruce Wienke in the 1990s. It models not just dissolved gas in tissues but also the growth of microbubbles already present in blood and tissue. It penalises fast ascents, short surface intervals, and repetitive diving more aggressively than Bühlmann does. In theory, tracking bubble mechanics gets closer to the real physiology of decompression sickness.

VPM (Varying Permeability Model) was developed by David Yount and Don Hoffman. Like RGBM, it focuses on bubble behaviour, but uses a model of bubble nuclei: microscopic gas seeds already present in the body. VPM tends to recommend shallower, longer decompression stops rather than deep ones. It is mostly found in dive planning software (V-Planner, GAP) rather than wrist computers.

DSAT is the algorithm behind PADI's original Recreational Dive Planner, developed through research commissioned in the 1980s. It is calibrated for no-decompression recreational profiles and tends to be more permissive than Bühlmann on single-day dives. Some Oceanic computers let users switch between DSAT and a more conservative alternative.

Shearwater Research (Perdix, Teric, Peregrine, Tern) uses Bühlmann ZHL-16C with user-adjustable Gradient Factors. Critically, it shows you the math: tissue bar graphs, current GF settings, and ceiling projections are visible on the dive screen. Technical divers and instructors gravitate here specifically because there is nothing hidden.

Suunto (D5, D6i, EON Core, EON Steel) runs a proprietary 'Fused RGBM', a hybrid that layers RGBM bubble penalties onto a Bühlmann compartment base. It has a long track record and is well-validated, but the internal weighting is not fully exposed to the user. On a long liveaboard, clients with Suuntos will sometimes surface with more deco obligation than clients with a pure Bühlmann computer on the same profile.

Garmin (Descent Mk3, Descent G1) uses Bühlmann ZHL-16C with Gradient Factors, close in principle to Shearwater. The display is less focused on diving metrics, but the underlying algorithm is solid and increasingly popular among divers who already use Garmin for other sports.

Mares (Smart, Quad, Puck) runs RGBM across most of its range. Mares was an early commercial adopter and has stayed loyal to it. The result is typically conservative behaviour on repetitive dive days: fine for a resort guest, occasionally frustrating for a guide on their fourth dive of the day.

Oceanic (Geo 4.0, Atom) is unusual in offering a dual-algorithm mode where the user can switch between DSAT (more permissive, closer to the RDP) and Pelagic Z+, a modified Bühlmann derivative. This flexibility suits divemasters and instructors who work in different training contexts.

Scubapro (G2, Aladin Sport) runs the UWATEC algorithm, a Bühlmann derivative developed internally and tested extensively over the brand's long history. It is reliable, less tunable than a full GF system, and well suited to recreational and technical sport diving without complex decompression planning.

No algorithm is universally best. They each make different bets, and the right answer depends on what kind of diver you are and what kind of diving you do.

Spend enough time around technical divers, cave divers and rebreather instructors and the same setup keeps turning up on wrist after wrist: Bühlmann ZHL-16C with Gradient Factors, running on a Shearwater.

The reason is not that it is provably safer than everything else. The reason is that it is auditable. You can read the original Bühlmann publications. You can understand exactly what GF 40/85 means. You can see your tissue loadings on screen during the dive. You can run the same algorithm in planning software (Subsurface, MultiDeco, Deco Planner) and verify that your computer and your plan agree before you get in the water.

That kind of transparency creates trust, and in technical diving, trust in your equipment is not optional.

RGBM and VPM have genuine merits. The bubble model logic is scientifically interesting, and RGBM has a solid track record in field use. But the proprietary weighting in commercial implementations means accepting a level of opacity that technical divers are increasingly reluctant to tolerate.

For recreational divers, none of this needs to matter. A modern computer running any of the four families will keep a no-deco diver safe on a 30-metre dive. The algorithm debate belongs to technical diving, where people are doing real decompression, mixed gas, and repetitive deep exposures where the differences actually compound.

If you are a recreational diver, the honest answer is: trust your computer. A modern device from any reputable brand will keep you well inside safe limits on a no-deco dive. Your ascent rate and your safety stop matter far more than which algorithm is doing the math.

If you are a divemaster, instructor, or exploring technical diving: understanding what is inside your computer is part of the craft. Bühlmann with Gradient Factors gives you the most visibility into what your computer is actually doing and the best integration with dive planning software.

If you run a dive center and choose computers for rental fleets or training: remember that the algorithm shapes how conservative clients experience their dives. An RGBM computer on a repetitive afternoon will sometimes surface clients earlier than the Bühlmann computer on their buddy's wrist. Knowing why that happens makes the conversation easier.

And if the gear and certifications are the part your team handles well, but the admin side still feels like controlled chaos: bookings, waivers, payments, and student records in one place is exactly what ScubaHub was built for.