[FoRK] 4-6G dual-core machines... recommended desktops?

Eugen Leitl eugen at leitl.org
Wed May 5 02:25:01 PDT 2004


On Tue, May 04, 2004 at 07:35:06PM -0700, Joseph S. Barrera III wrote:
> > a minimum of 2 gigs of RAM
> 
> Sounds like we're asking for a 64 bit machine,
> given that the max addressible by 32 bits is 4 GB.

There's this pesky 2 GByte/process limit, and NT/Win2k/XP seems to have a 512
MByte limit above which performance suffers. 

Speaking of 64 bit, given that Fedora Core 2 test3 works very well, I'm
building me a budget Athlon64 (2 GHz, 2x512 DDR400, ASUS K8V SE Deluxe VIA
KT800, WD360GD Raptor, Antec Sonata black TruePower 380 Watt). You can get a
DVD image for AMD64 from the torrent: http://torrent.dulug.duke.edu/
Final's due by May 17th, but it's working good already.

Apart from general everyday use I intend to play around with
http://leitl.org/docs/AMD_block_prefetch_paper.pdf
 
> But it doesn't really matter, as Longhorn won't ship until 2008.
> Moore's Law saves Microsoft's ass yet again.

Current R&D is extremely risk-averse, so all these gigatransistors/die and
teratransistors/wafer potential's going to continue to get wasted. I'm watching polymer
and organic electronics with a beady eye, though, as there's some intense
rustling in the woodwork going on. It scales into molecular electronics range, and is
trivially cheap to prototype multilayer, with the equivalent of an inkjet on
steroids. Don't get me started on what creative people could do with 100 MHz switches
in 100 layer technology, with prototyping costs in low 10-100 buck range, on
a <100 k$ rig. Yowza!

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
Url : http://lair.xent.com/pipermail/fork/attachments/20040505/7711d290/attachment.pgp


More information about the FoRK mailing list