The main reason why you have difficulties with GHz amplifiers made
from discrete parts is that every lead, every piece of copper trace,
every solder bump acts as an RCL component at those frequencies.

Within a processor you don't have these stringent restrictions, since
everything is so much smaller and hence the RCL time constants are so
much smaller.

This does not mean that it's easy and today's processor designers have
a lot of analog electronics to deal with...

Another problem is the length of the interconnects between different
parts of the chip. These are normally long enough to transport several
clock cycles worth of information at the same time, i.e. you have
several bits on the same line at the same time, following each other
at almost the speed of light...


Uwe.


> Hello!

> I ask this just, because I'm curious:

> When Intel is saying e.g., their processor is running at 3000MHz, what does
> this mean?
> Is there really any digital signal inside at that frequency, and a PLL that
> locks it to the in/out frequencies?

> I know, if they say, there front side bus operates at 800MHz, it's only
> 200MHz. Then they just multiply it by 4, because it's 32bit. But I cannot
> compare anything to vintage 8bit processors, so that doesn't make much
> sense.

> The second thing is: Processors are advanced thechnology. But could it be
> possible to have these speeds for a digital circuit, which means, a
> transition frequency of  >10GHz to remain any waveform other than a pure
> sine? If I try to make a circuit with a simple 1stage transistor amplifier,
> I'll use the most expensive III-V-HEMFET for 10GHz, and it's still a
> challenging task, and power below 10W; and they can do this with >100 000
> 000 silicon tansistors, 100W of power, on a single chip for $250?

> I ever tried to wind a wire loop around a running procesor and connect it to
> a spectrum analyzer. But you can see anything from nearly dc up to the max
> freq. of the analyzer. No strong peak at the declared frequency.

> Thanks for your answers!

> Steffen


-- 
Author: Uwe Zimmermann
  INET: [EMAIL PROTECTED]

Fat City Hosting, San Diego, California -- http://www.fatcity.com
---------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB CHIPDIR-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).

Reply via email to