Making a chip is a long and tedious process. We, programmers, often abstract the hardware running our codes. And that’s a good thing. Few of us, thrive at the junction point between the hardware and the software. At this level, you have to think about how a line of code is translated by the compiler, and how those instructions are executed by the hardware, so you can squeeze the last drop of performance and efficiency. At this scale, you think branch predictor, pipeline, cache line or execution engines. And every new program is a fascinating challenge! But there are more layers below this line that a programmer cannot reach. It is the realm of the transistors and the energy particles used to manufacture them. Today, I decided to share a few pics to show the complexity and the beauty of the chips we need to run our modern societies. On the wafer – the gold-ish pictures –, you may notice a tinny black dot on some of the chips. These are the failing ones. To engrave these circuits, when photolithography is used, you need several masks to define the patterns required to build the silicon circuitry (and a lot of chemistry as well). The mask shown here is made of quartz. Although these chips and the mask are old-tech compared to the state of the Art, it was impossible to take pictures that tributes their finesse. At least I tried ;-). Have a great WE!
If you want to learn more about the process of making a chip, you can read here: http://www.intel.com/content/www/us/en/history/museum-making-silicon.html.