[EMAIL PROTECTED] wrote: > > >Optical RAM: > > > >How small can you make a drop of <insert indicator> in > >solution (water)? The indicator might be pH based [mebbe], > >thermally based [not], electrically based [not], or some > >other very simple and scalable mechanism to change each "bit" > >of RAM [remember, color frequency allows many values for > >each transistor].
They can use multiple frequencies in electrical devies as well as in optical ones, we wouldn't have stereos if that weren't true. The reason why they don't use them in computer chips is that a device to filter/detect merely the presence or not of a signal is so much simpler, more reliable, smaller, and cheaper than one that can detect which frequency differences as well. When trying to make somthing on the scale of a computer chip, all of these factors add up very, very quickly. This problem doesn't just go away with optical circuits, either. (it's much easier to make a dectector to measure the presense/absence of light than one that can tell you what color(s) it is. By the same token, it's simpler to make a gate that either passes or stops all light rather than filters selectively. The reason why optics tend to be better for communication devices are related to two properties of light vs. electricity: 1. light in fiberoptics travels a bit faster, and with much less signal loss across a distance than electricity in wires. 2. fiberoptics support a much wider and higher-frequency bandwidth of light than wire does electricty.
