SerpApi.Org
At the end of the 1990s, IBM took the same basic machinery found inside regular HDDs and squeezed it into a package barely larger than a quarter. The Microdrive contained a tiny spinning platter, a motor, and a moving read/write head.
IBM appeared to have solved portable storage, but unfortunately for IBM, flash memory improved much faster than any mechanical drive ever could.
IBM squeezed an entire hard drive into a CompactFlash card
The Microdrive used familiar HDD technology at an almost ridiculous scale
IBM's first-gen Microdrive arrived in 1999 with 170MB and 340MB capacities. Each drive measured just 1.69 x 1.43 x 0.20 inches and weighed 0.56 ounces, yet it contained a single 1.08-inch magnetic disk spinning at 4,500 RPM. It also had the same fundamental components as a larger hard drive. Honestly, if you ask me, it almost looked cute with its spindle motor, moving actuator, read/write head, and controller electronics, all in mini size. It was small enough to fit between two fingers, after all.
The casing followed the CompactFlash Type II standard that kept the same footprint as the already-established Type I card but increased its thickness, creating enough room for the mechanical drive. The Microdrive used a CompactFlash-compatible ATA interface and could also work in a PCMCIA Type II slot with an adapter, which meant it could plug into the removable storage ecosystem that was already in use by cameras and laptops.
IBM didn't just scale every component down by the same amount, though (how would that even work?). The project initially looked into using microelectromechanical systems, or MEMS (not to be confused with memes), for parts such as the motor and actuator, but IBM concluded that carefully miniaturized conventional HDD designs were cheaper and worked better.
The Microdrive's electronics board was a whopping six times smaller than the board in a 2.5-inch HDD, all the while performing the same basic functions, and IBM added a load/unload mechanism because the tiny spindle motor didn't have enough torque for a traditional contact start-stop design.
The Microdrive solved a problem flash memory couldn't solve yet
Early digital cameras needed more storage than most people could afford
The Microdrive found its little niche in digital cameras, and that's where it really started to make sense. As per the pdf linked above, by 2000, IBM described digital cameras as the drive's most popular use, and you can see why.
The same source revealed that a 1GB Microdrive could hold 710 compressed photos taken at 2,048 x 1,530. Switch to uncompressed images, which ate up around 10MB each, and that number dropped to 100. A small flash card could fill up after just a handful of those, which wasn't ideal if you weren't just taking photos of your computer all day long.
Flash cards had the advantage of being solid-state, but getting a useful amount of capacity was pricey. News from 1999 reveals that IBM's original 340MB Microdrive offered more than five times the capacity of the solid-state cards available in the same general form factor, while IBM positioned the Microdrive inside a widening cost-per-megabyte gap between flash and larger HDDs.
Flash caught the Microdrive within roughly three years
And it was going so well
As time went by, the Microdrive's capacity ramped up, with IBM unveiling 512MB and 1GB models in 2000. Now we're talking.
The 1GB version was priced at an eye-watering $499, though, but it was still a feat: IBM nearly tripled the original drive's capacity in just over a year. Unfortunately, flash was moving even faster.
SanDisk announced a 1GB CompactFlash Type I card in late 2001, and it hit the shelves in 2002. Less than three years after the original Microdrive launched, flash had matched its headline capacity without needing a platter and all that other jazz (which was, like I said, admittedly adorable in a way).
IBM still had price on its side at first, as SanDisk's breakthrough cost just under $800, making even today's expensive SSDs look borderline cheap. But those prices dropped quickly, and even though the Microdrive was still faster than basic flash cards and higher-capacity versions were on the way, flash now offered the same storage for less money by late 2003. Cheaper, with equal capacity, using less power, and no moving parts ... yeah, the Microdrive was definitely in trouble.
The Microdrive kept growing even after flash caught up
It got a nice second act
When IBM and Hitachi combined their storage businesses in 2003, the Microdrive stopped being IBM's problem and became Hitachi's, and Hitachi wasn't about to call it quits. The company announced a 4GB version that quadrupled the previous capacity. It wasn't just the same drive with more data crammed in, either; Hitachi made the read/write head half the size of its predecessor, brought it 40% closer to the platter, and pushed the drive beyond 60 billion bits per square inch. Its transfer rate also improved by 50%, so flash may have caught up, but this tiny HDD wasn't giving up. Yet.
The Microdrive then landed one of its most recognizable jobs inside Apple's original iPod mini. It had room for around 1,000 songs and a $249 price tag, and the second-gen bumped up that capacity to 6GB. But that victory was short-lived.
Apple followed up with the 4GB iPod nano, much lighter and better, for the same $249, equipped with flash memory.
The Microdrive was brilliant, but flash is what we have today
The Microdrive deserves to be remembered as a brilliant solution to a problem that only existed for a few years. It didn't fail to do its job; it just filled a gap when it existed, and then faded into obscurity. Flash simply became too good, too cheap, and too convenient for miniature mechanical storage to keep up.
