two hd's on same IDE channel

  • Thread starter Thread starter Steve James
  • Start date Start date
What about sequential data transfer?

~misfit~ said:
I've tested it, althought with my relatively new nForce2 Ultra 400 board,
and having an old 120MB drive in PIO mode running alongside a modern drive
running UltraDMA mode 5 on the same channel made virtually no difference to
access times of the modern drive.
Click to expand...
 
The access time is determined by the physical attributes of the drive, i.e. rotational speed and head seek time, not by the I/O protocol. Sequential data transfer of a modern drive can be limited by I/O protocol; as slow as 2 MB/sec in PIO mode even if the drive is capable of 50 MB/sec.
 
Steve James's log on stardate 09 ožu 2004
I have two HD's on the same IDE channel, I think one is faster than
the other (ATA and RPM), will the fastest one be 'held back' by the
slowest one or not ?
Click to expand...

No, no and no again. Old urban legend ... ATA interface, until 3rd
volume of ATA 7 standard is paralel protocol. So, what happens with
_any_ two devices conected to one channal is very simple to understand.
Whatever tries to acces the device on one channal wil be able to to
that only to the _one_ device in the same cycle. That is the biggest
problem of ATA interface, and only then (communication between two
devices on same channal) will come to _latency_ since controller can
communicate with only one device per cycle, wich indirectly brings
slower transfer. But, devices will comunicate with controller
independently with the speed they decide, regardles on other device's
transfer protocol.
I have my O/S and program files on the main (fastest) drive and only
use the slower one for storage and the pagefile. (to reduce head
travel)
Click to expand...

Fine. You don't have to worry that your fast drive will work any
slower.
 
Mike said:
The access time is determined by the physical attributes of the
drive, i.e. rotational speed and head seek time, not by the I/O
protocol. Sequential data transfer of a modern drive can be limited
by I/O protocol; as slow as 2 MB/sec in PIO mode even if the drive is
capable of 50 MB/sec.
Click to expand...

Ok, thanks for clearing that up (I think). Maybe, instead of saying
"virtually no difference to access times of the modern drive" I should have
said that I benchmarked the system with a PIO drive and a UDMA drive on the
same channel and the benchmarks (which included read/write of various sizes)
was virtually unchanged for the UDMA drive even though it was sharing a
channel with a PIO drive. Of course. the PIO drive did top out at just under
2MB/sec but, as long as I wasn't accessing it at the same time as I was
accessing the UDMA drive it made no difference to the UDMA drive having a
PIO drive on the same ribbon compared to it being alone on the channel.
Better?
 
Back
Top