News New Compound Gives Chips Speed Boost

http://www.bit-tech.net/news/2005/12/08/Speedy_CPU_Compound/

 

w00t for hillsboro oregon!!!!!!!!

 

Qinetic pwn, they do some wicked research.

 

2015 is hardly anything to rejoice about, by then who knows what may lay in store (other than silicon).

 

If the semiconductor is so good then what is to stop them from using it right now? What are the dissadvantages to using it? The article really didnt go into it.

 

There are likely problems in
a. reliably producing pure large quantities of the stuff. (Very pure and very large)
and
b. building circuits on the stuff reliably.

Have you ever read what the process is to build a chip like a modern cpu? It's pretty intense and there are many factors.

 

By 2015 who says we wont be onto 0 nanometre processes? To be honest, I look at something like that as vapourware. It may as well not exist, theres no point in saying "We're gonna do this" if "this" is so far in the future that many many many other things are going to happen before "this" hits the shelves.

 

Ah, yes, I read about this earlier today over at Tom's Hardware. Apparently, the big problem is that currently, the material can only be used in creating gates 85nm in size, while the current 90nm process uses gates that are only 50nm wide. In other words, I guess it's only currently usable for something like the 150nm interrim process. (the one that came between the .017 and .013um process) However, I do think that this is also a step in the right direction; with some more work, I think it could be sped up to be on the same level as the traditional conductors. Otherwise, it'll only be for lower-end products only, meaning no dual cores to start with, for instance.

No, we won't be at a 0 nanometre process then. Rather, it's halved every two "full" processes; the 65nm process is coming up next, while the 130nm process was the outdated one; 45nm and 32nm come after this. In other words, each process halves the total size of the circuitry, and because each process comes every 18-24 months, Moore's Law is thus upheld. The big problem, of course, is that with recent advancements pushing transistor counts into insane numbers, there's been a bit of a problem with heat and power consumption keeping up...

 

Not to sound grumpy but... indium and antimony are both quite rare metalloids in comparison to silicon and copper, obviously this is bound to bump up the cost of a processor which isnt that great! But hey its good they're looking into alternative substrates long before the materials we're using now have ran there course.

 

Wow, I never thought processors would get better....

sar-ca-sm

 

thats the point, this may! its research. I'm sure there are many other materials they are looking into as well but it's prob made news because they have got somewhere with this.

 

I thought they were looking at the diamond-boron combo for processors. Synthetic diamonds are like the holy grail of heatsinking materials and I think they have already figured out ways to fab gates out of it. still a majillion technical issues to pound out (just as any new semiconductor) I hear that around 2010 or so they are predicting silicone to not be able to keep up with moore's law, not to mention the insane heat and wattage some of these chips are starting to put out