Computers are insane. Microprocessor lithography is magic.
We use lasers to hit microscopic falling droplets of molten metal to produce a magic light that reacts with a photosensitive layer that then gets etched by acid and transforms a silicon wafer into a rock that can think.
It’s wild.
Much of that complexity comes from just the generation of the EUV light… but aren’t there readily available resources that produce RF emissions with an even shorter wavelength?
We know how to generate X-rays and gamma rays with a certain degree of reliability. I have only the most basic understanding of these systems, which is why I can only approach it with major oversimplifications.
To me, it seems like the general methodology of processor production has remained very similar over the decades. Obviously, techniques and precision have been refined to near perfection… but when was the last real disruption to the process?
Why not stack layers of components etched by the laser itself? Why not include layers of copper to act as ground and power planes that behave as a capacitor to even out voltage spikes from a psu, while also wicking away heat to the packaging?
With the extreme precision available, and the insane capital backing an AI arms race, I would love to see some novel changes that revolutionize computing again.
I’m tired of incremental progress. We just got solid state battery production, humanity should match with more efficient, revolutionary computing as well.
