My Grandpa Hein grew up with horses when automobiles were experimental and he lived to see men land on the moon. That was a tremendous change he went through. In the same way I have gone through another tremendous change from large computers occupying rooms to an even more powerful computer in my pocket, my smart phone.
The Cray 1 was one of the first supercomputers. Supercomputers were and are built for performance, for quickly calculating numbers. The Cray 1 blew away the competition in the 1970s. In 2013, Roy Longbottom in comparing a Raspberry Pi 1 to a Cray 1 wrote “In 1978, the Cray 1 supercomputer cost $7 Million, weighed 10,500 pounds and had a 115 kilowatt power supply. It was, by far, the fastest computer in the world. The Raspberry Pi [1] costs around $70 (CPU board, case, power supply, SD card), weighs a few ounces, uses a 5 watt power supply and is more than 4.5 times faster than the Cray 1.” I own a Raspberry Pi 400 (RPi 400) and the CPU is an ARM A72, which is a pretty normal cell phone CPU. He later ran a floating point benchmark on the RPi 400. My $100 RPi 400 is 78.5 times faster than a Cray 1.
What is the big difference between the Cray 1 and the RPi 400? It is the clock speed. The Cray 1 ran at only 80 MHz and my RPi 400 runs at 1800 MHz. The RPi 400 clock speed is 22.5 times faster than the Cray 1. Taking away the clock speed means the floating point improvements or gains are roughly 3.5 times. And it is miniaturization that puts the transistors closer together to allow for these extra performance gains, and miniaturization is also what allowed the clock speed to increase.
Moore’s Law is an observation that about every two years the number of transistors in a microchip doubles. This self-fulfilling observation has driven the industry to make smaller and smaller transistors and thus allowed the clock speed to increase without the chips overheating. Unfortunately Moore’s Law has ended or is ending because the transistors are now approaching the size of atoms. It is becoming physically impossible to continue to shrink the transistors.
Computer storage also has had an amazing evolution. We have gone from a computer disk the size of a washing machine to fingernail size disks. We have gone from 8 inch floppies that hold 80,000 bytes to disks today that hold 22,000,000,000,000 bytes (22 TB). I remember purchasing in 1994-95 for work a 2 GB (2,000,000,000 bytes) disk for $2000. 30 years later you can get a 22 TB disk that costs $420 and has 11,000 times more storage. My phone has more storage than that 2 GB disk from 30 years ago. Computer storage has also benefited from miniaturization. That fingernail size (11mm x 15mm) microSD disk can today hold up to 1.5 TB.
Computer technology has transformed the world. I have written about the amazing computer performance advances that have taken place, but computers have become parts of phones, automobiles, and even washing machines. Think of the influence of the internet. The internet is all about computers sharing information. Computer technology is everywhere. So what is next? Right now, there is a lot of talk about artificial intelligence (AI). Will computers eventually become intelligent beings? Or will AI assist us in our tasks and only be a tool for us to use? We will have to see.
Pretty amazing. David Charles and I have commented several times along these lines in years and decades past.
Just a few clarifications here: You wrote, “We have gone from 8 inch floppies that hold 80,000 bytes to disks today that hold 22,000,000,000,000 bytes (22 TB)”. You could add 80K for the floppies, although later floppies of the same 8″ class held 1.2 MB. But regardless, technically speaking you’re comparing floppy disks with hard disks — which isn’t actually an apples-to-apples comparison. There were hard disks in 1978 holding 80 MB or more, such as the one that was hooked up to the first microcomputer that was ever hooked up to such a ‘beast’. (That computer was an Ohio Scientific Challenger — which almost certainly was the Challenger III, released in 1977. I’ve included a paragraph below from Wikipedia on that model, which was unusual for its time in more ways than one.) Also, you referred to a microSD as a disk, but that’s actually a flash drive (i.e. SSD).
Regardless of nomenclature, yes, the rapid evolution of computing has been amazing, even if the last 12 years or so has seen a more-or-less plateauing of processor speeds. However, quantum computers, already under development with some prototypes already having been released in limited ways, will likely multiply computing speeds by a thousand times or more, which will enable processing of truly immense databases that will be a boon for weather and climate modeling, molecular engineering, and many other resource-intensive applications.
Meanwhile, another interesting aspect of the rapid increase of computing technology is the graphics end of it, with early computers like the one I started using in 1979, which was a Tandy/Radio Shack TRS-80 Model 1, had only 128 x 48 graphic blocks that it could display, and in 1-bit black & white only, of course. By comparison, I was just watching someone play a highly-detailed, immersive 3D video game in 4K resolution in at least 24-bit depth at 60 frames per second on a 40-inch monitor. That too represents a quantum leap comparable to that from the Wright Flyer in 1903 to an F22 Raptor in 2005.
Jeff Lemke
P.S.: From Wikipedia regarding the Challenger III — [Notes: I’ve removed footnote numbers for easier readability in this environment. Also, this author did not mention the size of the C-D74 hard drive unit. It may have been around 80 MB, but that would take some research. However, another source, “Stan Veit’s History of the Personal Computer” (1993), mentions that Mike Cheiky at OH Scientific was “not only one of the first to design a floppy disk system for computers, but he was the first to add a Winchester hard disk to a personal computer. The Okidata Company had a 14-inch 84 MB hard disk drive, which he interfaced to his Challenger personal computer at a time when most PC’s were using cassettes to store data!”
Now about the Challenger III, from https://en.wikipedia.org/wiki/Ohio_Scientific:
“The Challenger III was released alongside the Challenger II in 1977. It was a desktop computer featuring three microprocessors—a MOS 6502A, a Zilog Z80, and a Motorola 6800—on one board. This combination of processors allowed the computer to run virtually all software for microcomputers on the market at the time of its release. Only one processor can be active at a time, preventing it from computing in parallel, but software interrupts allow programs to switch from processor to processor on the fly.”
“Ohio Scientific oriented the Challenger III as a development kit for students of computer science wanting to learn how to program for all three processor; as a small business or industrial machine, for organizations wanting to consolidate mission-critical applications for multiple platforms onto one unit; and for the extreme hobbyist. An external, single-sided (later double-sided), dual 8-inch floppy drive unit was available for the Challenger III, as was the C-D74 hard drive unit. Ohio Scientific was keen to match the Challenger III with the C-D74, offering both in a 74-inch tall rack-mount case as a complete system christened the C3-B — the first microcomputer system to include a hard drive. A variant of the C3-B with a cheaper, lower-capacity 24 MB drive was released by 1979. The C3-B was particularly useful as a database manager serving multiple client terminals. To this end, Ohio Scientific provided a serial I/O board called the CA-10, allowing up to sixteen terminals to connect to the Challenger III: A version of the Challenger III with integrated CA-10 and dual 8-inch floppy drives (but without the integral keyboard) was introduced as the C3-OEM in late 1978.”