Often regarded as the world’s first analog computer, the Antikythera Mechanism is the most technologically advanced instrument known from antiquity. Built roughly 2,100 years ago, its complex system of interlocking bronze precision gears charted the movements of the planets, sun and moon. This Greek, hand-powered device also predicted eclipses, tracked the moon’s phases and displayed the dates of the ancient Olympics.

“It’s almost unbelievable in its brilliance,” says Tony Freeth, a member of the University College London Antikythera Research Team, who has been studying the Antikythera Mechanism for two and a half decades. He adds that it encompasses most “everything that was known about astronomy at the time.”

Prior to its discovery in 1901, the ancient Greeks were believed to have crafted only simply made fixed gears for use in such objects as windmills and watermills. But the Antikythera Mechanism proved that they had likewise mastered epicyclic gears—or gears mounted on other gears. The mechanism further showcased their intricate metalworking skills, including finely carved gear teeth about a millimeter long.

“This is very advanced gearing,” says Freeth, who has a background in filmmaking and mathematics. “We wouldn’t expect to see it before the Middle Ages. And to see it in this instrument was a complete shock, honestly.”

Though likely not one-of-a-kind, these devices were undoubtedly rare, says Alexander Jones, a historian of ancient science at New York University, and author of A Portable Cosmos: Revealing the Antikythera Mechanism, Scientific Wonder of the Ancient World. He speculates that “maybe there was just one high-end workshop around 100 B.C. that had the skills and knowledge to produce them.”

“It must have taken months to complete one, and the clientele was probably small too,” Jones says. He notes that when they stopped working, their metal parts would have been recycled—a common practice in antiquity.

Due to a shipwreck off the Greek island of Antikythera, near Crete, around 60 B.C., at least one of these devices escaped being turned into scrap metal. It sat unnoticed on the seafloor until the year 1900, when some sponge divers on their way to North Africa from the island of Symi stopped at Antikythera to shelter from a storm.

Looking for sponges there, they instead discovered the shipwreck, starting with what one diver apparently described as a “heap of dead naked people”—actually marble statues. Taking a bronze arm with them, the sponge divers presented their find to the Greek government, which then contracted with them to conduct an underwater excavation of the site. With a gunboat standing by to deter looters, the operation took place in late 1900 and 1901.

The divers, wearing primitive helmets and canvas dry suits, could only stay at the bottom for about five minutes before surfacing. Yet they managed to uncover one artifact after another, from bronze and marble statues to glassware, furniture, ceramics, coins and two pairs of golden earrings.

Though the large bronze statues were all in fragments and parts of the marble statues had been badly corroded by seawater, they still generated much excitement in the press. So too did several bronze statuettes and other intact smaller items. By contrast, a corroded lump about the size of a shoebox garnered little attention at the time.

In 1902, however, a Greek politician visiting the National Archaeological Museum in Athens—where the artifacts were being stored—noticed coin-sized gearwheels within the lump.

“It was realized immediately … that this was ‘something’ special,” Jones says. Because bits of visible text referred to the sun and Venus, it seemed to have something to do with astronomy. “But no one had a clear idea then of what kind of thing it was,” Jones says.

Though early researchers offered up various opinions, Freeth says they were nearly all wrong. Some, for example, thought it was a navigation instrument. During visits to Athens in 1905 and 1906, German language expert Albert Rehm got closer than most. “He was the first to really identify the Antikythera Mechanism as being a sophisticated astronomical calculating machine,” Freeth says.

The next major breakthrough came when British physicist Derek de Solla Price studied the Antikythera Mechanism from the 1950s until the 1970s. Price, who commissioned the first X-rays of the device, figured out that its basic layout consisted of a box with displays of dials on the front and back faces. “His attempt to reconstruct the mechanism was faulty in many ways," Jones says, "but all the subsequent research has been basically an extrapolation of the approaches he pioneered.”

Other scientists conducted a second X-ray study in 1990, which generated 3-D images of the mechanism and its dozens of corroded bronze gearwheels.

Freeth became interested in the subject around the year 2000. After years of lobbying the Greek authorities, he and his colleagues received permission to collect more advanced 3-D images, using a machine they installed in the basement of the National Archaeological Museum in Athens. “It’s just like a CT scan you might have in a hospital, but at much, much higher resolution,” Freeth says. “You can’t put a human being into one of these machines because it would kill them instantly.”

Complicating matters is that less than half of the apparatus was pulled from the sea. Though more of it may remain out there somewhere, Jacques Cousteau and others have explored the Antikythera shipwreck and come up empty. In addition, the original lump split into 82 separate fragments long ago.

Freeth calls it a very difficult 3-D jigsaw puzzle. “It would have saved us quite a lot of time and effort and struggle if it had been left in one piece,” Freeth says. Nonetheless, he and his colleagues published multiple papers shedding new light on the mechanism.

Disagreements remain over the mechanism’s exact design and functions. But researchers now have a pretty good idea of how it worked. Turning a crank or knob to set the gears in motion, users could view the positions of the sun, moon and five known planets at any given date in the past or future, including their positions relative to the zodiac constellations.

As seen from Earth, these celestial bodies don’t just travel in steady circles. Rather, they go forward and backward through the sky and change speeds, illusions caused by two things the ancient Greeks didn’t comprehend: elliptical orbits and heliocentrism. Their gearwork needed to be advanced enough to account for these herky-jerky motions.

In addition to the planetarium-like display on the front of the Antikythera Mechanism, two large spiral dials decorated the back. One predicted eclipses using a Babylonian-based 223-month lunar cycle, and the second tracked a 19-year-cycle of the moon’s phases. Other smaller dials showcased the cycles of the ancient Olympics and other athletic competitions.

“These games were a really central part of ancient Greek culture,” Freeth says. “They used to suspend wars while the Olympic games were on. Then, when the games finished, they went back to fighting each other.”

In one of Price’s papers, he referred to quotes from the Roman orator Cicero, who lived around the time of the Antikythera shipwreck. Cicero purportedly credited Archimedes, the Greek mathematician and inventor, with developing a machine in the third century B.C. “on which were delineated the motions of the sun and moon” and planets and which accounted for their “various and divergent movements with their different rates of speed.”

Ever since, researchers have wondered whether Archimedes invented the Antikythera Mechanism, or at least influenced its later design. “My own view is that Archimedes probably had the idea [for this] wonderful, complex, extraordinary system,” Freeth says.

Jones, on the other hand, downplays the Archimedes connection. “I don’t consider the sources that attribute something of the kind to Cicero … as historically accurate narratives,” he says.