Rack Mount Servers
Ken Ambrose
kena at well.com
Sun Aug 11 22:56:32 EDT 2002
On Sun, 11 Aug 2002 kurth at usaexpress.net wrote:
> I'm just tired of the limits of the x86 line....640k,8gb.....why don't manf
> take into account that just because a 10Tb drive dosen't exist today, one
> will exist with 6 or 8 months...and plan.....no, that's too simple....
While I am in -no- way defending some of the dain-bramaged decisions that
brought about today's PCs:
1) SCSI, regardless of platform, just wouldn't care.
2) With the ATA/133 spec (48-bit address space), you can go up to 131072
TERABYTES before you run into problems (2^48*512 block). I imagine
we will hit it someday. I ain't holdin' my breath on that one.
Something that can hold every movie I've ever seen, uncompressed,
in a -fraction- of its total capacity is when I actually think
demand for bigger drives will begin to slacken. ;-)
3) 640 K is an 8088 limitation that DOS/Windows kept alive until
(more or less) Windows 3.0. After the '286, though, "real" OSen
stopped having that problem.
Does this mean I think X-86 is the be-all end-all? No. Do I think its
price:performance ratio wins out over most other platforms? Yes.
Especially if AMD's Hammer makes it big in 64-bit land. Are there
advantages that Sparc and Alpha systems have (eg. their console modes)
that I wish X-86 had? Sure thing. My servers will nevertheless remain
X-86, and inexpensive, and not tied to a single CPU vendor's whims.
$.02,
-Ken
More information about the gnhlug-discuss
mailing list