If you're the kind of person who enjoys watching popular science programs on TV, then you might have heard some of Dr. Michio Kaku's fascinating theories on what the future has in store for us. In this post, however, we'll take a quick trip back to the past, inspired by something that Kaku states in one of his recent books: "Today, your cell phone has more computer power than all of NASA back in 1969, when it placed two astronauts on the moon."
Seems hard to believe, we know, but it is actually true – a hand-held apparatus on which we fling birds at pigs has greater computational capabilities than the arsenal of machines used for guiding crafts through outer space some 45 years ago.
Apollo Guidance Computer, 0.043MHz clock speed
Multiple IBM System/360 Model 75 mainframe computers, costing up to $3.5 million apiece all the while occupying tremendous amounts of space, were operational at NASA at the time. Each could perform several hundred thousand addition operations per second, and their total memory capacity was in the megabyte range. As for the 70-pound Apollo Guidance Computer, which the Apollo 11 Command Module had on board, it was a machine that had 64 kilobytes of memory and operated at 0.043MHz. In comparison, an iPhone 5s, which you can easily fit into any pocket, has a CPU running at speeds of up to 1.3GHz – enough to enable the execution of millions of calculations each second. And the iPhone's 1GB of RAM should well suffice for storing the 6 megabytes of code that NASA developed to monitor the status of its spacecrafts and astronauts in 1969.
The Cray-1 supercomputer ran at 80MHz
Several years later, in 1975, a supercomputer named Cray-1 came around. It was a capable and awesome-looking piece of machinery that "flew" at a rate of 80MHz. While generally used for scientific projects, such as simulating the interaction of fluids, one of these bad boys helped render the CGI for the first Tron movie, released in 1982. But a Cray-1's raw computational power of 80 million floating-point operations per second (FLOPS) is laughable by today's standards; the graphics unit inside the iPhone 5s
produces about 76.8 GFLOPS – nearly a thousand times more. And surely, the iPhone can render 3D graphics looking better than Tron's Lightcycle scene. On a related note, the Cray-2 supercomputer was released 10 years after the Cray-1 and was the world's fastest supercomputer until 1990. But even with a performance of up to 1.9 GFLOPS, the liquid-cooled, 200-kilowatt machine still ranks behind the Apple iPhone, at least when it comes to GFLOPS ratings.
Deep Blue – still not quite as good as a smartphone
Deep Blue is another supercomputer that you might have heard of. It is the machine best known for winning against world chess champion Garry Kasparov with a score of 2:1 in a 6-game match. That happened on May 11, 1997, when Deep Blue was the 259th most powerful computer in the world. It boasted a performance figure of 11.38 GFLOPS and could evaluate 200 million positions on the chessboard each second (although still not good enough to run Crysis, we suppose). Today, some 17 years later, the ARM Mali-T628MP6 GPU inside the Exynos-based Samsung Galaxy S5
outputs 142 GFLOPS. And the 192-core GPU on the Tegra K1 SoC produces an even more impressive peak of 364 GFLOPS. Sure, these might not be superior to Deep Blue when it comes to playing chess, but in terms of brute, number-crunching power, these mobile graphics processors stand stronger.
So yeah, technology is definitely advancing, and it is doing so at a rapid pace. What takes a supercomputer to calculate today will most likely be a piece of cake for the smartphones (or whatever they evolve into) that we'll be using in 2020, just like today's smartphones have the processing potential of a vintage supercomputer. What we'll be investing this computational power into, however, is a whole different topic. Feel free to share your predictions down in the comments!