How Computers Work - Hardware

[Love Pulse]

Computer Hardware
The computer is an amazingly useful general-purpose technology, to the point that now cameras, phones, thermostats .. these are all now little computers. This section will cover a lot of ground, introducing major parts and themes of how computer hardware works. "Hardware" refers the physical parts of the computer, and "software" refers to the code that runs on the computer. These are deep and complicated topics, but many of the basic principles make perfect sense and will give you an insight about what's going on behind the scenes.

Chips and Transistors
Modern computers use tiny electronic components which can be etched onto the surface of a silicon chip. By far the most common electronic component is the "transistor" which works as a sort of valve for a flow of electrons. The transistor is a "solid state" device with no moving parts; it is a basic building block used to construct more complex components. In particular, a "bit" (below) can be built with an arrangement of 5 transistors. The transistor was invented in the early 1950's, replacing the vacuum tube. Since then, transistors have been made smaller and smaller, allowing more and more of them to be etched onto a silicon chip.

Moore's Law
Moore's law states that the density of transistors on a chip doubles about every 18 months. It is not a scientific law, just a broad prediction that seems to keep working. More broadly, it captures the idea that per dollar, computer technology (not just transistors) gets exponentially better as time goes along. This is quite clear if you look at the cost or capability of computers / cameras etc. you have owned. It might mean that each new generation of computer is faster. OR it could mean that keeping capability constant, computers get cheaper and cheaper .. showing up in more in more contexts as it becomes cost effective (e.g. in your car, your thermostat, ..).

CPU Central Processing Unit - inevitable referred to as the "brains" of the computers. The CPU does the active "running" of code, mani[CENSORED]ting data, while the other components have a more passive role, such as storing data.
RAM "Random access memory", or just "memory". RAM is the working scratchpad memory the computer uses to store code and data. RAM is effectively a large storage space of "bytes" under the control of the CPU. RAM is relatively fast, able to retrieve the value of any particular byte in a few nanoseconds (1 nanosecond = 1 billionth of a second). The other main feature of RAM is that it only keeps its state so long as it is supplied with power. It is not "persistent". Suppose you are working on your computer and it suddenly loses power and the screen goes blank. You understand that what you were working on is gone .. RAM has been wiped clean, leaving you only with what you last saved to disk (below).
Hard Disk or Flash Memory "persistent storage", typically today on a computer provided by a spinning hard disk (also known as "hard drive"), or by "flash memory". The hard disk reads and writes magnetic patterns on a spinning metal disk to store bits, while flash memory is "solid state" .. no moving parts, just silicon chips to store bits. In either case, the storage is persistent, in that it maintain its state even when the power is off. Flash memory (also known as a "flash drive", typically in the form of a usb key) is persistent like a hard disk, and is faster and uses less power than a hard disk. However, per byte, flash is significantly more expensive than hard disk storage. Flash memory has been getting cheaper, so it may take over niches at the expense of hard drives. Flash memory is much slower than RAM memory, so it is not a good replacement for RAM. (Note that Adobe "flash" is unrelated; it is a proprietary media format.)
File System The hard disk or flash memory provides persistent storage as a flat, undifferentiated collection of bytes without any organizational structure. Typically the hard disk or flash memory is formatted with a "file system" which organizes the bytes into the familiar pattern of files and directories, where each file and directory has a somewhat useful name like "resume-2010.txt". Essentially, each file refers to a block of bytes, so the "flowers.jpg" name refers to a block of 48KB of bytes which are the data of that image. The file system in effect gives the user a name (and probably an icon) for a block of data bytes, and allows them to do operations on that data, like move it or copy it or open it with a program. The file system also tracks "meta information" about the bytes: how many there are, the time they were last modified.
Microsoft uses the proprietary NTFS file system, and Mac OS X has its Apple proprietary HFS + equivalent. Many devices (cameras, MP3 players) use the very old Microsoft FAT32 file system on their flash cards. FAT32 is an old and primitive file system, but it is good where wide support is important (Microsoft claims to hold patents needed to read or write FAT32, but it is widely implemented anyway). Proprietary and patented formats make it hard to store and use your own data.

Pictures of Hardware
Below images of a low-end Shuttle computer with a 1.8ghz CPU, 512MB of memory and a 160GB hard drive. It cost about $ 200 retail in around 2008.

Here is the flat "motherboard", a little smaller than a 8.5 x 11 piece of paper, that the various components plug in to. At the center is the CPU. At the far is the RAM memory. Just to the right of the CPU are a couple support chips. Prominently, one of the chips is covered with a copper "heatsink" .. this presses tightly against the chip, dissipating the heat from the chip into the surrounding air. The CPU also had a very large heatsink, but it was removed to make the CPU visible.

Computer Hardware