3 ms·
That's a whole other very interesting industry. You're right, in a way. I too find the effectiveness of the semiconductor industry exceptionally fascinating. B
by tonyarkles 10y ago
That's a whole other very interesting industry. You're right, in a way. I too find the effectiveness of the semiconductor industry exceptionally fascinating.
But it should be noted that most IC designs don't have 100% yield, especially at the cutting edge process nodes. Depending on where the defects are, dies are either thrown out or "binned" (e.g. if a high-end CPU has bad transistors in a section of the cache, that section of the cache is disabled and the CPU can carry on its life as a Celeron).
If you really want to melt your brain a little bit with scale, think about RAM. Every bit in a stick of RAM has n transistors (1 <= n <= 4, depending on the design), plus a capacitor. For an 8GB stick of RAM, you're looking at possibly 32 billion transistors and 8 billion transistors! Even just the fact that we squeeze that many of anything into the space we do is incredible! And then for good measure, we change the signals going to it 1-2 billion times/sec!
- fixermark 10y ago> e.g. if a high-end CPU has bad transistors in a section of the cache, that section of the cache is disabled and the CPU can carry on its life as a Celeron For me, that's the part that is really fascinating about the fab industry. We have a statistically-reliable method of fabbing, but we build in the necessary fallbacks in the chip design itself that if a chip partially fails quality control, we can scale it back and sell it as lower-end product. It's intriguingly analogous to the quality process in gemstones, except that chips aren't found---they're made via mechanisms that we can't perfectly control.