Re: VIA MVP3 gives bad IDE UDMA33 performace ?!

From: Daniel J Blueman (daniel.j.blueman@stud.umist.ac.uk)
Date: Sun Mar 26 2000 - 18:35:26 EST


Hi,

Harald Koenig <koenig@tat.physik.uni-tuebingen.de> wrote:
> is there any chance to get better IDE performace for VIA MVP3 chipset ?
>
> I'm using a 40GB 7200rpm Maxtor disk which can deliver ~28 MB/sec,
> but with my main board (DFI K6BV3+) I only get ~13-16 MB/sec :-(
>
> here is some information from kernel 2.3.51 startup and hdparm:
[snip]

I get the exact same issue. When I benchmark my harddisk under linux, I get
~15MB/s under the Intel BX UDMA33 controller, and ~19MB/s under the HPT366
UDMA66 controller, tested with hdparm -t (hdparm -T gives me >128MB/s).

Now, under Windows 2000, I get 28.2MB/s. All measurements taken from the
outer recording zone.

Mark Hahn says this is hdparm being a not-so-good benchmark, however, it
seems it gets quite close to the disk, so it is a nice, pure test, and can
measure one very important metric. If it was the case that the hdparm was a
bad test, bottleneck and so on, then why the variation in performance
between the iBX and HPT366 controller? - certainly not the IDE bus bandwidth
(as shown by the consistent W2K results).

I'd love to be able to achieve platter speeds without the need for raw I/O
(well, rawer than what hdparm does, anyway).

Harald - what did you use to do your benchmarks?

- Dan
__________________________
Daniel J Blueman
Undergraduate - BSc Computing Science
UMIST university - Manchester
Direct line: 0161 933 3569
Mobile: 07775 583766
Email: daniel.j.blueman@stud.umist.ac.uk
SMS: daniel.j.blueman@sms.genie.co.uk
----- Original Message -----
From: "Harald Koenig" <koenig@tat.physik.uni-tuebingen.de>
To: "linux-kernel-list" <linux-kernel@vger.rutgers.edu>
Cc: "Andre Hedrick" <andre@linux-ide.org>
Sent: Friday, March 24, 2000 12:20 PM
Subject: VIA MVP3 gives bad IDE UDMA33 performace ?!

> Hi,
>
>
>
> # hdparm -I /dev/hda
>
> /dev/hda:
>
> Model=aMtxro5 04898U , FwRev=AD26C00Q,
SerialNo=8K30NECK
> Config={ Fixed }
> RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57
> BuffType=3(DualPortCache), BuffSize=2048kB, MaxMultSect=16, MultSect=off
> DblWordIO=no, OldPIO=2, DMA=yes, OldDMA=0
> CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=80041248
> tDMA={min:120,rec:120}, DMA modes: mword0 mword1 mword2
> IORDY=on/off, tPIO={min:120,w/IORDY:120}, PIO modes: mode3 mode4
> UDMA modes: mode0 mode1 *mode2 mode3 mode4
>
>
>
> I've tried 'hdparm -m8 /dev/hda' and similar to change `MultSect=off'
> but this doesn't make any difference.
>
> I'll attach a small PS file which shows transfer rates for the whole 40GB
> (position in MB vs. MB/sec) with both UDMA33 (my board) and UDMA66
> (K7 system from a friend).
>
> as you can see, my UDMA33 system gives exactly half the performace
> for the outermost cylinders and surprisingly I get the same transfer
> as UDMA66 only for the very innermost cylinders! there both PCs
> give ~16.5 MB/sec.
>
>
> any idea/hint/suggestion what I can try to get better transfer rate
> with my main board ?
>
>
> btw: transfer rate is the same for clean 2.2.14 and 2.3.47/2.3.51 kernels.
>
>
> thanks for any hint,
>
> Harald
> --
> All SCSI disks will from now on ___ _____
> be required to send an email notice 0--,| /OOOOOOO\
> 24 hours prior to complete hardware failure! <_/ / /OOOOOOOOOOO\
> \ \/OOOOOOOOOOOOOOO\
> \
OOOOOOOOOOOOOOOOO|//
> Harald Koenig, \/\/\/\/\/\/\/\/\/
> Inst.f.Theoret.Astrophysik // / \\ \
> koenig@tat.physik.uni-tuebingen.de ^^^^^ ^^^^^
>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:18 EST