Over the last couple of months I have been considering again Extreme Ultraviolet Lithography (EUV). This has been a bit odd. I should add that I was (am) something of an expert on the light source parameters for this. I once designed and built first and second generation EUV – Laser Produced Plasma (LPP) source prototypes . I had hands on experience of building and operating these; exploring parameter space. I messed around with lasers and sources. I know lasers, I know molecular beams.
EUV is a so called Next Generation Lithography (NGL) and can be thought of as a fancy photocopier which writes high resolution circuits onto Silicon chips. As a photocopier it has a light bulb (source), a copier (scanner or stepper) and paper (Silicon). The semiconductor industry is driven by Moore’s law, allegedly handed to him on stone tablets upon the top of Mt Sinai. This says that chip dimensions should get smaller and chips more powerful, quickly. It has been gospel for decades. To do this a shorter wavelength of light is now needed. This is EUV at around 13 nanometers. The problem is that before the litho machines can be “desktop” or widely available and used, the source needs to be a lot brighter. Current versions cost tens of millions of dollars each and are very big indeed! They use ~ 0.5 Megawatt of electricity. They may be used for niche, high end, fancy parts of the circuits, sort of a calligraphy or manuscript illumination as opposed to main text.
To date the industry has invested multi-billion dollars in this technology, more than the cost of a Nimitz Class nuclear powered aircraft carrier.
BUT, the light source is still not bright enough for EUV to be inserted into high volume manufacture (HVM). It is late!! And the timeline for its introduction keeps on slipping. Big business has invested very heavily and the semiconductor community is getting itchy….
It is really weird for me to sit here knowing that I have thought a great deal about EUV and EUV-LPP in particular. Yet here we are over a decade later and with all that investment, all those highly paid engineers, shiny steel and polished optics; HVM EUV is still something of a chimera.
I never wrote down my ideas and only put a few preliminary ones into the prototypes I built. I can see where the problems with current source design are. No-one is interested so I keep these to myself.
The way I see it the industry can (may) solve its problems and everyone will breathe a sigh of relief. Or, things can slip again and again until such time as the brown stuff starts to hit the fan. If the whole semiconductor industry goes belly up, there would be huge knock on effects in the global economy.There is little I can do as a hermit who has been alienated and ostracized.
It is very weird for me. I have a question “Do I as yet have some physical plane Karma in respect of EUV?”
I don’t have the answer to this yet.
What if I had the answers people were looking for? How would they find me here in my cave in the forest? What would they pay me? Maybe I know nothing…who can say?
The thing about LPP is that it is a very non linear process. It requires a non linear and scientific approach, as opposed to an incremental engineering mind set.
It is so weird that this should raise its head again…Hmmmnnnnn….