Forum Home
Press F1
 
Thread ID: 102071 2009-08-06 03:25:00 What's the real story with the ATi SB600/SB700 and AHCI/USB? Agent_24 (57) Press F1
Post ID Timestamp Content User
798721 2009-08-06 03:25:00 Can someone please clarify this or give me a link to an explanation?

After trying to search Google for information it seems to be that these series of southbridges have issues with their USB controllers and that AHCI support is bad.

But most information is either conflicting or otherwise seems untrustworthy.

Is this because of a bad design or because installing Windows XP on them in AHCI mode is hard?

If there is a problem with USB, what is it?

Are these problems avoided if using the nForce 700a series chipsets instead? Or is this just more of the same
Agent_24 (57)
798722 2009-08-06 03:35:00 Is this what youre talking about with AHCI ? (support.microsoft.com) Thats if you use Vista

Its probably hard with XP unless you slipstream the SATA drivers. Or have a floppy. Vista shouldnt be a prob, since it deals with the drivers (If you enable AHCI in the BIOS). It installs them for you
Speedy Gonzales (78)
798723 2009-08-06 03:45:00 I was not aware of that problem but I don't know what problem I am talking about. All I know is that searching on google there is inconclusive evidence that there may be some kind of problem with SB600/700 series, eg: that the only way to actually use them is with AHCI disabled

Maybe I am completely wrong, I do not know.
Agent_24 (57)
798724 2009-08-06 04:08:00 Looks like SB600 did have probs, (or still does), SB700 fixed them and SB750 (is the same as 700) but RAID 5 and ACC were added for overclocking. Info here (forums.amd.com)

The major downsides of SB600 according to user consensus are:

a) RAID errors and erratic functioning in RAID 0/1, specifically on Vista 64-bit.
b) AHCI not working properly so SATA drives nearly always are running in IDE UDMA 6 mode.
c) HDD and ODD SATA/IDE combinations on many motherboards may not function correctly or not function at all due to SB600 problems which are not remedied in the BIOS releases although could be.
d) Lower USB and HDD performances in general.
e) A clock generator bug which means it will induce erratic overclocking behaviour every 9/10 attempts and limit the processor overclock somewhat substantially with the known "poor overclockers". The cause is very likely on-die PLL's falling out of the required phase synchronization window parameters.
f) For the same reasons, poor and/or limited HT overclocking.

The latter two are critical problems for overclockers in that, you can fall 200 - 500 MHz short of the actual overclocking silicon frequency limit due to synchronization errors within the allocated time.
Speedy Gonzales (78)
1