From: Alfio Puglisi (puglisi@arcetri.astro.it)
Date: Wed Aug 14 2002 - 09:28:12 MDT
On Tue, 13 Aug 2002, Samantha Tennison wrote:
>I've been following many of the bleeding edge trends
>the last few months, and from all appearances there
>appears to be a quantum leap in computer performance
>and capacity on its way. Here are the following leaps
>I am currently aware of:
>
>MRAM - Magnetic RAM:
>Holographic Storage:
>PCI Express (previously 3GIO):
>Graphics Acceleration - post NV30, and IBM's "The
>Cell":
All the new technologies that you mention are not big enough to be
called "quantum leap". Most of them fit nicely into Moore's Law series of
doublings. Terabytes of disks are nothing special (there already are 200GB
drives shipping), PCI-X++ bandwidth levels are much needed today, and I
wouldn't be surprised if 10Ghz chips will be out in 2005.
The 1 teraflop IBM processor should not be compared with today's CPU: it
will probably be a very problem-specific machine (i.e. graphics), and not
a general-purpose processing unit. For example, nVidia graphics processors
are already much faster than the fastest Pentium 4, if you just look at
the flop numbers.
Alfio
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:16:05 MST