VIA chipeset & ata133 issue

[meaty]

What is everyones opinion on this so called problem.
I read this here about how the
via chipset mobo's have crap performence with ata133.
I really want to up my harddrive speed , doesnt everyone.

 

[Wind]

It has been discussed here at AT. The "flaw" can be solved by applying a patch .

Get it here.

 

[meaty]

thx wind
Now my only grip is when will IDE go 10,000 rpm lol

 

[Wind]

<< thx wind
Now my only grip is when will IDE go 10,000 rpm lol >>

Pleasure & Merry Xmas to u.

 

[oakfan]

Although this problen is really highlighted by the ata133 drives, I think it rears its ugly head with ata100 systems.

I'm running a Soyo SY-K7V Dragon+, which has the Via 226A chipset and an onboard promise ata100 raid chip.

Without the latency patch I get sandra scores of 30800. After applying the patch I get 31300, while sandra puts the goal for an ata100 raid0 7200rpm setup at 36000.

Anyone else have any sandra scores for their Via motherboards?

 

[mboy]

My score is in the 26000-27000 range after a fresh defrag on the 4gb OS partition of win2k on my 8KHA+ using a Maxtor 40gb 740dx ata133 at ata100. Applied the patch and didnt notice much difference if at all.

 

[DClark]

The overall score for Sandra's test won't change much, but look at the numerical results in the bottom scolling area after the test. My overall result didn't change much (both were 20,3XX), but even with an old 686A southbridge (ATA66) I noticed quite a difference:

Before (using the 4.34 4-in-1s):
Buffered read: 43MB/s
Buffered write: 30MB/s
Sequential write: 29MB/s

After (still with the 4.34 4-in-1s, put with the patch):
Buffered read: 51MB/s (+8MB/s)
Buffered write: 37MB/s (+7MB/s)
Sequential write: 28MB/s (-1MB/s)

The other bench results (sequential read, random read, and random write) weren't affected.

If you want to check out the difference, you can uninstall the patch and re-run Sandra, making note of the results.