Hi Ferenc,
We spoke today a bit about the uLED matrix option. The price of the matter is $4000-$5000 for a prototype from their first tape out, which does not compare favourably with an option of a second hand DLP projector with mercury lamp.
I delivered a brief about the uLED thing in question. Key points:
1. JBD uses a bonding process to glue together a GaN layer from sapphire carrier wafer, on top of silicon wafer with BEOL, and FEOL. In other words, only the LEDs themselves are GaN. 2. That is very different from uLEDs with where GaN is grown on Si, or where BEOL, and FEOL are patterned on the same sapphire substrate. That results in vastly superior LED quality, and blowing nearest competitors by millions of nits. Serious wattage from this thing is possible. 3. The guy in question, Li Qiming, may well be the only man in the world with a machine which can glue III-V to Si in serious commercial volumes today. This is also the biggest showstopper.
Hagen gave a good analogy with litho equipment on the market to cars: if second hand litho machines which used to be on the market were cars, then the cheapest one would've been a second hand Ferrari, while we are trying to build a bicycle.
I then said that even second hand bicycles would be freaking expensive. Second hand bicycles in question would be other single shot exposure machines.
While I keep optimism about uLED matrix, I agree that a DLP projector with a UV lamp would probably be the best for this "bicycle" stage.
Ferenc, what I wanted to ask you as a man with the best optics knowledge here:
1. Do you think a uLED matrix with micro-reflectors the like one JBD has be suitable for a proximity lithography? 2. Do you think it is possible for mortals to make this uLED matrix to work in a scanning mode (like in a scanner vs a stepper) if it can be driven fast enough?
Later we spoke about new funding options.
A new humongous fund called NGEU is now being created https://en.wikipedia.org/wiki/Next_Generation_EU which will bring $5B additional funding to Horizon Europe over 2 years. This will also be accompanied by a yet to materialise EU semiconductor fund, which will be many times bigger.
In light of this, I noted that with so much money flying around, it may well be harder to secure funding for a "bicycle" project than a "Ferrari" project, meaning that funding for something more substantial may be easier to get under current circumstances, despite it not being the current aim of the project.
I myself would've exploited every possible opportunity out of the fact that EU politicians are now peeing molten lead over chip shortages.