Semiconductor industry... was Re: [opensuse] Micro pauses, or CPU spikes?
I worked in the semiconductor industry for a long time. I know the subject well. Funny though, you never seem to see SUSE in the industry... loads of Solaris based software though (running the wafer steppers and wafer scanners).
Wafer scanners?
I've never worked with the technology myself, but as far as I can understand, it is a laser that is blowing bubbles on a silicon dye? The micro or nano technology referring to the size of the dye used? Where do you use the scanners?
Whoops... waaay offtopic now. Lets move this to the Offtopic list Wafer stepper and scanner refers to the way in which the reticle is imaged on the wafer. In simple terms, a wafer stepper steps to a location on the wafer, the shutter is opened, the entire reticle image (the circuit pattern.. think photographic negative) is exposed on the wafer (using laser light) in an area usually 22mm x 22mm, the shutter is closed, and the machine steps to the next location and repeats. There is a lot of other things going on, like minute adjustments to compensate for the topology of the wafer, aberrations in the lens and so on, but this is the basics for how it works. A wafer scanner works similarly, but instead of exposing the entire image from the reticle in one "flash" of laser light, the image is scanned onto the wafer over an area usually 26mm x 33mm, in much the same way/principal as a photocopier works. OK, not exactly, but it's a good analogy. There isn't really dye involved in the process... it's called "resist". Again, think of it more in terms of photography. A photo-sensitive liquid (the resist) is applied to the wafer (making the wafer the rough equivalent to the film in your camera). It is sensitive to a certain light wavelength... like 193nM (current and experimental technology has successfully got it down to 30nM). When it is exposed to the laser light, the pattern from the reticle is left behind on the silicon wafer. The resist is washed off, and the process is repeated over and over building up the various layers. There are more processes involved depending on what you are making... but this should give you and idea of what is happening. There are all sorts of interesting problems and solutions working at this level. Like the resist for example... the polymer chains in the resist are long enough, and the lines being imaged are small enough, that you end up with "fuzzy" lines... the polymer chains in the resist fall out of the lines. There are solutions to this of course... different resists, different imaging techniques etc. An interesting industry if you like using massive machines to make incredibly tiny lines on silicon or ceramic. :-) The bigger the machine, the smaller the lines it can make. C. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (1)
-
Clayton