electro-Physics: 3.3 volts

From: Dwight K. Elvey <dwight.elvey_at_amd.com>
Date: Fri Dec 10 12:25:57 2004

>From: "Steve Thatcher" <melamy_at_earthlink.net>
>
>there is a secondary aspect to voltage reduction and that is reduced power
dissipation... which also translates to shoving more stuff into smaller
spaces...
>
>-----Original Message-----
>From: "Joe R." <rigdonj_at_cfl.rr.com>
>Sent: Dec 9, 2004 9:51 PM
>To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk_at_classiccmp.org>
>Subject: Re: electro-Physics: 3.3 volts
>
>At 09:30 PM 12/9/04 -0500, you wrote:
>>Today I started mothballing the VAX 6000 and noticed that a large part
>>of its power supply is on 3.3 volts, now common, but probably not in 1988.
>>
>>Seems like something was invented, probably in chip design that made
>>3.3 so useful. I wonder what that was, when that landmark was reached.
>
> It wasn't a new invention. In was the fact that ICs were getting so
>dense that they couldn't squeeze in any more circuits unless they could
>reduce the space that the insualting layers took up. The layers were
>already so thin that they'd break down at about 5 1/2 volt so they had to
>reduce the voltage to 3.3 volts. Since then they've further reduced it to
>1.8 volts and (I think) now to 1.1 volts. And there are plans to reduce it
>even more. All so that they can reduce the bulk of the insulating layers
>and add more gates.
>
> Joe
>

Hi
 I don't think I'm giving away any trade secrets by mentioning
that the primary reason for going to lower voltages is to
reduce power consumption. The thinner oxides are to make them
work at lower voltages( needs more gain ). Power is a squared
factor of the voltage. Even a small reduction in voltage is
a big payoff in power. If we could make the transistors work
at 0.1 volts, we'd be doing it.
 Remember, we make these things.
Dwight
from AMD.
Received on Fri Dec 10 2004 - 12:25:57 GMT