Looks like these are two completely different approaches:
* The first one replicates the technology of the 1960s and indeed produces his own analog IC. As marvelous as it is, it is like making your own computer from 74xx logic ICs: possible, but technically far inferior to commercially available parts. And the chemistry involves all sorts of nasty substances: Hexamethyldisilazane, NH4F (Ammonium Fluoride), HF; he needs a 50W Ytterbium fiber laser and a DLP - all possible, but only for the sake of actually doing it.
* The second one is essentially an IC pooling company. 100$ for an own (digital) IC does not sound much, but 350$ sound more realistic if you look closer. And that is for volume one, the price however seems to be linear as this is a sample process.
There is little I can think of where I would like to have my own IC. In the digital domain, a CPLD/FPGA will usually make more sense for low volumes. It is great for prototypes, sure, but what do you do for actual volumes? Approach TSMC for 1k/year? Likely, they will not even answer.
And Itsy Chip does not seem to support analog ICs (whose design, by the way, is not trivial and very much different to analog design on a PCB).
Sorry, I am not convinced. Even if the technology is fascinating.
We have custom chips made through MOSIS. The process we have been using was AMI semiconductor, now taken over by ON Semi (Spun off from Motorola). It is a 0.3 - 0.5 um process with CMOS plus high resistance resistor layers and a capacitance layer, so good for mixed-signal and analog stuff.
Last project we had done was 160 pieces of a 5 x 7 mm chip for about $40K. In this business, that is CHEAP!
Jon
There is little I can think of where I would like to have my own IC. In the digital domain, a CPLD/FPGA will usually make more sense for low volumes. It is great for prototypes, sure, but what do you do for actual volumes? Approach TSMC for 1k/year? Likely, they will not even answer.
Here you go for low volumes:
http://www.europractice-ic.com/sv_quote_request.php
Here's the guy's write up:
http://sam.zeloof.xyz/first-ic/Once he gets his process going well, I believe there will be plenty of derived works for all of us to enjoy. Maybe electronics clubs will soon all have their own fabs like this
That looks interesting. If I just had more time... I remember back in the university we tried to make something much simpler like just some test structures on metallization layer. But then that mask alignment/exposure machine (an ancient Brother or was it Canon?) just refused to load any wafers, so that ended bit prematurely.