My Sandy doesn't like deviders

[PoL]

Hi guys!
Well, I got here next to me this beauty, and well, it does overclock pretty good, but doesn't really like to have his VX PC4000 under deviders.

The config I used to have was 305*10 and a 166 devider for 24/7, and well, one day I began to experience some problems.

For example, after being stable for a 32M superpi run, it would boot being at the desktop...
Then, for example, if I changed a timming of the memory in the BIOS, next time in windows would be a reboot.
Till I got to a point where every time I logged to windows the computer tool like 35 sec to reboot... every time.. no matter what vcore or vdimm...

Well, I decided to try a 150 devider... the same...
I tried 305*9 166 devider... pretty the same...
Until I try 255*11 1:1 and everything goes perfect...

OK... I set again 305*10 166 and the same prob... every time rebooting after 35 sec from the log in more or less...

I ran the other day 305*10 1:1 with some TCCD and no prob, so that's why I guess my sandy has some issues with deviders...


What do you think?

[brandinb]

mine too it really sucks because it makes the cpu pretty unflexable
________
Fix Ps3

[PoL]

mine too it really sucks because it makes the cpu pretty unflexable

Are you serious? That's pretty crazy... I thought I was the only one cuz I found nobody with this same problem... Now I really know what's going on.

What is your step?

Anyone else with this bug?

[uwackme]

Mine too. I wanted to use the 166 divider and have 275Mhz HTT 230Mhz memory using the Mushkin PoS 2x1G sticks I got, but although memtest works, Windows crashes during boot with any dividers. So my only way to work it was 230Mhz 1:1 x11 which is far from the 2750Mhz the 3700+ runs at with normal BH/CH UTT ram at 250Mhz tight.

Oh well, off to the iRAM boards go these 1G sticks.

[PoL]

Man... this sucks...
Damn... winnies have that cold bug...
Sandys this f*cking devider bug...
Is there something wrong with all chips?
Is this a well known problem or this only happens to some chips?

AMD bad boy

[Major]

I have simular issues with my SD, 10x multi is unstable and so are some of the dividers at various multi

[PoL]

That's pretty curious cuz mine seems to love 10x. I think it's the more stable multiplier from far, but it is one of those who have issues with deviders.

[babalouj]

my 3700+ hates the 10x multi also but loves the 11 and 9 multis.

[PoL]

my 3700+ hates the 10x multi also but loves the 11 and 9 multis.

But I think they hate 10x just because of the deviders. If you set anythin * 10, but with a 1:1 devider, everything works perfect.

[Deus Falsus]

use bios 510-2fix, and set "max async latency" and "read preamble" to AUTO. I've had 5 venice and one SD, while the venice allowed for a higher memory oc @ 1:1 the San Diego doesn't and is way more picky to memory manipulations, actually IMO it has a weaker memory controller than the venice. Nonetheless none of my venice cpu reached higher than 2750mhz core frequency, whereas the SD reaches 2850 on air prime stable, the tradeoff being its weak memory controller. BTW I have G.skill tccd and it has done ddr 630 1t stable in another cpu, but with the SD I'm forced to do either 260mhz 2.5/3/3/3 or 285mhz 3/4/3/6, I have to use CAS 3 when going higher than 270 or windows will reboot unexplicable, even if memtest passes an eternal loop. SD mem controller is a beotch to master.

[PoL]

use bios 510-2fix, and set "max async latency" and "read preamble" to AUTO.

what does that? It makes this devider bug to desapear?

[Deus Falsus]

It made my problems disappear.

[babalouj]

i dont use dividers and my 10x doesnt work.

[PoL]

i dont use dividers and my 10x doesnt work.

Definitely that is weird...