Dancer wrote:
>
> I may be talking out of my hat here, but I remember something similar in the news
> from last year. Some university-of-queensland fellow had found some way (so the
> story goes) to pack a terabyte or so of fast static ram into a single unit about
> the size of a pcmcia card, and put an IDE interface on one end. It was supposed
> to be about 20ns or so, IIRC. And it was supposed to be cheap. The papers said he
> got hired by a memory manufacturer who wanted to (a) remain nameless and (b)
> retail the unit.
>
> Hadn't heard a thing since.
>
I don't believe a single word you cite (i.e. I believe you, when you say
you read it somewhere, I just do not believe what the u-of-q boy said).
Look at my organization in the header. We have now 64MBit chips
available, maybe there exist 1Gbit chips that run at 80 degree Kelvin
somewhere in a japanese lab. But then such a chip is well
fingernail-sized and has a capacity of 'only' 128MByte. So you need 8
chips for one GByte of *dynamic* RAM, 8000 for one TByte. Would you be
able to fit 8000 fingernail x 0.5mm chips into a single pcmcia package
(or even in a 3 1/2 inch HD case)? Static RAM is by no means near that
degree of integration, not to speak of such new things as MRAMs
(magnetic RAMs) or or other RAMs using the XMR effects.
This is reminiscent of that guy that told everyone he had polymer near
the market a couple of years ago. He was hired by a scandinavian company
and no one has heard anything more of them (no, it was definitly *not*
Linus :-)...
Marc
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/