Thanks
I was actually thinking about the production line ASIC. Say you have a volume of 1million devices per year, the complexity of 20+ opamos, would we be talking sub dollar prices? (I know there's a lot of variables, but just a ball park number)
Thanks
Klaus
u
if
at
ke
w would a 3mm transistor perform?
-do-it/
esting increase in the rate of CPU performance growth.
I very very much doubt it
-Lasse
I doubt I will ever get to install an ion implanter in my basement.
What might be more interesting is if the mainstream IC fabs ever decide to go to electron beam lithography for mass production of consumer ICs, when the UV feature sizes stop shrinking.
If there were factories doing mass production from e-beam lithography, then there is no particular reason why there would have to be large setup costs for a new design. This would mean that many more applications could justify the design of an ASIC, as it would only be necessary to justify the engineering cost, without the extra half-million dollars or more for a state of the art mask set.
I expect the engineering cost would come down also: at present nobody is going to try using a team of entirely inexperienced chip designers, or untested (and inexpensive) cad software to produce their million dollar mask set - the cost of a failure is too expensive. Therefore you pay more for experienced engineers, and software that costs more than a sportscar, even if you probably could have got away with something cheaper, with a moderate probability.
Once the cost of trying out the design comes down to maybe just a little more than the cost of the wafer (similar to the cost of a multilayer PCB), many more people without much prior experience would be able to have a go at chip design, so the design process would gradually get much cheaper.
Chris
u
if
at
ke
ES2 - European Silicon Structures in Aix-en-Provence used to do that, with a bunch of shaped beam electron beam microfabricators. They got taken over by Atmel in 1995.
Actually, electron beam microfabrication is just too slow for mass producti on.
Semiconductor simulation got good enough around 1990 that ASICs worked firs t time from then on. My electron beam tester didn't make it to production a round then essentially because people stopped having to use an electron bea m tester to find out where first silicon wasn't working so that they could get the second pass right.
An electron beam tester saved about six months development on the 68000 bac k in the mid-1980s. Simulation got good enough to kill that market by 1990.
-- Bill Sloman, Sydney
Presuming we're talking reasonable voltages, I'd guess definitely under a dollar, maybe as low as 20-25 cents.
"Complexity of 20+ OpAmps" is no longer considered complex, see...
for moderately complex... around 6mm x 6mm... circuit design NRE higher than layout and foundry costs ;-)
Contact me directly if you need assistance. ...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
MOSIS prices on small chips using the oldest/cheapest technology are not that much more than a fancy multilayer PCB- maybe a factor of 3 to
-- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.