Re: Absolutely horrid IDE performance...

Aaron Tiensivu (tiensivu@pilot.msu.edu)
Thu, 26 Nov 1998 11:24:10 -0500


> The Celeron is definitely not the problem... shoot from what I've seen the
> lack of cache actually helps performance (very slightly) on large bulk
> memory transfers.

Hmmm..
I've got my C300a roasting at 464mhz @ 2.0V and it really is Intel's dark
secret. :) 1.3Mkey/sec with RC5 smokes MIPS chips and a lot of others.

> Ok... I'll try after Thanksgiving (don't want to toast my server right
> before I leave:). Which kernel are you using? Do you have UDMA enable in the
> BIOS? The board is a SuperMicro P6SBA (BX chipset) and it has a PIIX4 on it.

That's super odd. The kernel for some reason is thinking you've got the
original PIIX chipset. That, I think, is where a majority of the problem lies.

I have used, with this harddrive, 2.0.35 and 2.1.12x.. I really didn't do much
of anything under 2.0.35 but it did correctly ID the PIIX4.

> Which the ShuttleHOT 569A w/ P75 here at work gets even worse performance on
> the IBM 8.4gig 5400rpm drive.

That's super odd once again. :)
It might be worthwhile to check out developer.intel.com and check out the
errata on the PIIX4.. I briefly scanned it but didn't notice anything that
would really explain this.

I'll have to do a bonnie suite test when I get a chance under 2.0 and 2.1
This has got my curiosity peaked.

--
Which is worse, ignorance or apathy? Who knows? Who cares?
IBM: It may be slow, but at least it's expensive.
Linux: Get it up, keep it up. Viagra for PCs.
AOL doctor to new mother: You've got male!

- 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/