Unofficial Opteron 939 Thread

[Voltes-5]

I have a 165 CCBWE 0550 UPMW, and I agree on the memorycontroller.

On my 146 CABYE 0540 FPBW I could run the same TCCD memory with

the same timings at 310MHz, on the 165 I can only get them stable at

250-260MHz. I now run at 300MHz with the 166 divider.

I can get the 165 stable at around 2700MHz but it takes a lot of volt, 1.62 V the temp at full load is about 40 degrees C.

I will try different memory timings to try to increase the memory frequency.

 

It also proves that no cpu is the same, especially when overclocking. My 165 is at 315 with a 180 divider which means my Ballistix are running at 283. CPU vcore at 1.664 and ram at 2.8. I have been able to boot up windows at 322 x 9 with a bit more voltage at 1.68 and play Prince of Persia, but alas, cant prime for more than ten minutes. Strange thing is that one of the cores can go on for more than 5 hrs.

[mingyeung]

just got myself a 144 CABNE 0545

Anyone tell me what i can expect when i learn to master overclocking lol.

Oh and what ram would you recommend and 1 gig or 2

thanxs

 

get 2 gigs! way better for multitasking and is a better deal, since you will always run a divider anyways for the opterons unless you are getting PC4800, which i don't believe come in the 2 gig flavor.

 

i would recommend the OCZ PC3200 Platinum 2 * 1 GIG. only 279CDN with rebate. pretty cheap i would say

[RhumJM]

Had changed my Venice to an Opteron CABYE 0540 FPBW and do some experencies the last days - at the moment i have a rock solid rig on 2.9 GHz, but every more FSB its not Prime95 stable. Increasing Vdimm to 3.3 V helps to pass MemTest 1.65 (Test 5 and 8) for hours, windows start up an OK, but priming not possible. Also have losen the RAM timmings (specially MAL & RP) and increase the Vcore to 1.55 V (DMM measured), but nothing helps. Need same advice what to do now, cause some other people with the same steppings reach 3.0 GHz without problems. should change to 6.23-2 and use 136% Vidd special for more stable Vcore ??

 

 

EDIT: MBM5 are yet not calibrated - the shown Vcore are NOT the real measured by DMM (1.493 V)

[rahtid]

that would be the cpu it self cpu z read rev e4

[RhumJM]

that would be the cpu it self cpu z read rev e4

 

:confused: is there any newer revision (E6) available that reach 3 gig ???

IMO this is the same as all other using to oc so there must be another fact to change like ram timmings or something else....

[Death909]

Had changed my Venice to an Opteron CABYE 0540 FPBW and do some experencies the last days - at the moment i have a rock solid rig on 2.9 GHz, but every more FSB its not Prime95 stable. Increasing Vdimm to 3.3 V helps to pass MemTest 1.65 (Test 5 and 8) for hours, windows start up an OK, but priming not possible. Also have losen the RAM timmings (specially MAL & RP) and increase the Vcore to 1.55 V (DMM measured), but nothing helps. Need same advice what to do now, cause some other people with the same steppings reach 3.0 GHz without problems. should change to 6.23-2 and use 136% Vidd special for more stable Vcore ??

 

I have the exact same problem. My 146 is stable at 2.9ghtz, but using the 160/200 divider results in my VX being unstable at or above 240mhz (it can do 245mhz in memtest). I haven't been able to test this myself, but I think that by using a memory capable of higher HTT (TCCD) and dividers closer to 200/200 you could get closer to 3ghtz.

 

My VX is quirky (not in a good way, one stick doesn't get any errors at 246mhz but it won't boot into memtest at 247mhz) and the Vdimm controller is screwed, so I have to run 3.5v though my VX just to get it stable at 245mhz. Using the 160/200 divider, it is stable at 238mhz, but anything above that gives random BSODS.

[loc.o]

Try loosen drive strength to 3, data drive to 1, mal 8 and rp 5.0. I think it isnt the VX but the high HTT that's causing probs, timings need to be adjusted in order to run higher HTT.

[RhumJM]

I have the exact same problem. My 146 is stable at 2.9ghtz, but using the 160/200 divider results in my VX being unstable at or above 240mhz (it can do 245mhz in memtest). I haven't been able to test this myself, but I think that by using a memory capable of higher HTT (TCCD) and dividers closer to 200/200 you could get closer to 3ghtz.

 

My VX is quirky (not in a good way, one stick doesn't get any errors at 246mhz but it won't boot into memtest at 247mhz) and the Vdimm controller is screwed, so I have to run 3.5v though my VX just to get it stable at 245mhz. Using the 160/200 divider, it is stable at 238mhz, but anything above that gives random BSODS.

 

 

Using the 166/200 divider with no prob at 3.1Vdimm on my BH5 at 241MHz - with more Vdimm (3.3V in order to switch to the 5V Rail that causes massiv heat !) and same timmings my RAM goes rock stable at memtest for hours at 250MHz (3.0GHz).

So i think its not the divider at all or the ram - will try tweaking with loc.o settings...will also try to lower LDT and Chipset voltage.

 

 

Vcore (Load): 1.490 V (measured DIRECT over C290 near CPU with Metrix DMM)

Vcore (Idle) : 1.472 V (same as above)

Vdimm (Load): 3.104 V (measured DIRECT over EC79 near DIMM1 socket with Metrix DMM)

Vdimm (Idle): 3.110 V (same as above)

[Spring2188]

get 2 gigs! way better for multitasking and is a better deal, since you will always run a divider anyways for the opterons unless you are getting PC4800, which i don't believe come in the 2 gig flavor.

 

i would recommend the OCZ PC3200 Platinum 2 * 1 GIG. only 279CDN with rebate. pretty cheap i would say

 

 

Just curious I have an Opty 148 that I OC'd to 2860 mhz using 2x512 OCZ el plat rev 2's. 1:1 ratio. Why do you say you'd always use a divider??? Just curious cause I am wanting 2 gigs of Ram but don't know to go 4X512 or 2x1. I have seen though that a lot of people are using dividers on these Opty's, just curious why, I'm fairly new to OC'ing

[Guest SaLadin_SS]

Is any one aware of were to get the windows fix for the opteron. also if the 065-170-and 175 are the same why are some people complaining about there mem controler in the 170?

 

thx

[My Ugly PC]

Some people have noted Opterons mem controller to be different. Either to just need different timings or to need mem above a certain divider or MHz. If you listen to the preachings around here, mem MHz will have a very small performance difference. IMO who gives a .. If your gettin' better cpu MHz on lower voltage thats all you need. Its so well worth it IMO.

 

To be fair, many chips have bad mem controllers. I've seen some X2's that suck, and many a crappy venice. If your not willing to concede to luck of the draw, then your not of the risk/reward ratio that opty's are all about. You could be just as bad off mem wise with a crappy venice IMO, I think opty mem issues are an exaggurated deal.

 

2x1gb will yeild less headaches generally, than 4x512. You wanna talk about mem controller issues, 4 modules will stress a mem controller pleny. I got two kits of this, I think its a great value..

 

http://www.dfi-street.com/forum/showthread.php?t=33702

 

Ugly

[Guest SaLadin_SS]

i see so its not a specific run of optys that has a specif problem with its mem controler, for example a specific problem with the mem controler in the opty 170 CCBWE 0546 XPMW? Its more a a gerneral problem that may or may not be and issue in a much braoder goup of 64s?