Re: udma66 ...
Hello!
> > though i'm yet to meet drive which could saturate even UDMA(33)
>
> I'll have to second this. If you try this drive on a 33M bus, and it's
> the only device on the bus, you're not likely to see any improvement
> just by sticking it on a 66M bus. The drive is the limiting factor
> here. I don't think there's a single-disk system that does much above
> 20M or so.
I'm using a Western Digital Expert (2MB Cache, 7200RPM), with a Promise
Ultra/66 board, and kernel 2.2.12 with the hedrick patches.
hdparm -t /dev/hde gives me 22.5 MB/s
If I don't turn on the UDMA/66 feature (through this line in lilo.conf:
append = "ide2=ata66") the drive runs noticeably slower, I think hdparm -t
gave me around 17MB/s.
I think that proves that at least in my case, there is some benefit of the
UDMA/66 bus.
Greetings,
Alexis Maldonado
Engineering College
University of Costa Rica
Reply to: