[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: odd behaviour with harddisk



On Tue, 19 Nov 2002, Alvin Oga wrote:

> > > -- any system ... no matter how fast/slow, is working well
> > >    if the ( buffered disk reads ) timing is at least 50% of the rated disk
> > >    speed  of 66MB or 100MB or 133MB  or rated sata 160MB  or rated scsi
> > >    speeds
> > > 	- have not see any system achive full-rated-disk-tranfers
> > Isn't this strongly dependent on the disk drive subsystem, not just the
> > controller?

>
> its more dependent on the motherboard ( onboard ide controller chips )
>
> - same disk will under-perform by 1/2 or more if a slow/bad ide driver on
>   the "bad" mb
> 	- all else staying the same/unchanged
>
> 	- don't get cheap mb... :-)
>

I think I see the confusion.  You're getting disk media speed, and
interface speed confused.

I could have a udma-4 drive with low density platters, at 5400 rpm,
compared to a 7200 udma-4 with 80 gig platters.

The interface is the same, at ata-100.

But the buffered disk speed is going to be a LOT more, and that can't be
explained by ide controller chips.  It's going to be the hard drive
itself.

Yes, slow cpu's and chips can cripple performance, I have many p120's here
with 100+ gig drives in them, that are limited to about 10 meg a second.
In an uncrippled machine, such as the P4 2.4ghz sitting here, it does
40ish, which is close to the actual media speed, not the ultra-133
interface.  We're probably a year away from seeing ultra-133 saturated :)

Mike



Reply to: