🚀 go-pugleaf

On Sun, 05 Jan 2003 18:03:20 +1000, Floyd Davidson wrote:

> Matt van de Werken <m_vdw73@optusnet.com.au> wrote:
>>On Sun, 05 Jan 2003 02:22:06 +1000, Floyd Davidson wrote:
>>> I'll grant that if you are not running X and work only on one virtual
>>> console at a time, one cpu will be just as fast as two.
>>> 
>>> 
>>Even then, not necessarily. Try make -j 2 to compile your kernel. I get
>>a clean kernel compile time of less than 5 minutes for my system (dual
>>Athlon MP1800+, 512MB DDR RAM).
> 
>     make -j2 MAKE='make -j2' bzImage
> 
> In my case that takes less than a minute and a half.
> 

Well, I've never actually 'time'd it, but I do know it's much faster than
my old dual celeron 366 system, probably about the same time as making a
cup of coffee (which I'm not in the habit of timing, either...).


> ...
> 
>>Since going to dual CPUs about 3 years ago, I'd never go back to a
>>single processor system. I'd *much* rather spend the money and get two
>>not quite state of the art CPUs for around the same price as one
>>latest-n-greatest CPU.
> 
> I'm on my second SMP system.  The first one was an Asus P2B-DS with a
> pair of PII's at 350Mhz.  A year ago I finally "replaced" it, with a new
> Tyan S2462UNG board with a pair of Athlon MP 1600+ cpu's.  Then I put a
> couple 533Mhz Celeron's into the P2B-DS.  They both have SCSI disks,
> though the Tyan board gets several times the i/o speed that the Asus
> board does.  The Tyan uses a RAID5 array with 4 disks spread across two
> SCSI controllers, which accounts for most of the disk speed difference.
 
Nice systems. I can't quite stretch the budget to SCSI disks, though, so
I'm stuck with a couple of (non-RAID) IDE disks. I did buy the fastest
ones I could afford, though.

Cheers,
MvdW