What is causing my instability?

[imported_Iketh]

what is the purpose of the "RAIDexpert driver"?? i launch it and ie switches to intranet settings and im asked for login and password, nothing else.....

EDIT: nm, i used the full ati catalyst installer instead of what dfi has on their website... they are missing the "ITX" driver i think is what it's listed as

 

[Sylvanas]

You seem to be getting good clocks with the Phenom, 2ghz NB seems to be where it's at. If you are trying for 2.6ghz you might want to check this as well, interesting that it's plausible that Phenom offers 100%+ clock scaling above 2.6ghz based on the results we have. And yes Vdroop is there by design...read Anand's review here

 

[Markfw]

Originally posted by: Sylvanas
You seem to be getting good clocks with the Phenom, 2ghz NB seems to be where it's at. If you are trying for 2.6ghz you might want to check this as well, interesting that it's plausible that Phenom offers 100%+ clock scaling above 2.6ghz based on the results we have. And yes Vdroop is there by design...read Anand's review here

vdroop is not by design, or it would be on all motherboards, its only the cheap ones that suffer from it. I contested that review, and I dare him to prove me wrong. My DQ6's don;t have it. Many review sites use that motherboard due to it's high OC potential. Why ? no vdroop, and a good solid power delivery system.

Just because one reviewer said it was by design, does not make it so.

 

[Sylvanas]

Originally posted by: Markfw900

Originally posted by: Sylvanas
You seem to be getting good clocks with the Phenom, 2ghz NB seems to be where it's at. If you are trying for 2.6ghz you might want to check this as well, interesting that it's plausible that Phenom offers 100%+ clock scaling above 2.6ghz based on the results we have. And yes Vdroop is there by design...read Anand's review here

vdroop is not by design, or it would be on all motherboards, its only the cheap ones that suffer from it. I contested that review, and I dare him to prove me wrong. My DQ6's don;t have it. Many review sites use that motherboard due to it's high OC potential. Why ? no vdroop, and a good solid power delivery system.

Just because one reviewer said it was by design, does not make it so.

From AT review: To make matters worse, unjustified negative consumer perception surrounding this often misunderstood design feature eventually forced a few motherboard manufacturers to respond to enthusiasts' demands for action by adding an option in their BIOS that effectively disables this important function. The higher end mobos like the DQ6 lack alot Vdroop because the consumer expected such a thing to be evident in a costly motherboard. The cheaper ones suffer from it because the enthusiasts are more likely to go out and buy a DQ6 so there no need to integrate this feature, and in most cases Vdroop can be nearly eliminated on 'cheap' motherboards by adding a resistor as a Vmod- if Gigabyte wanted little Vdroop on lower end motherboards it takes a $2 resistor- but they'd rather have you go out and buy the top end $300.

 

[Markfw]

I do not see how giving a cpu variable voltage, which inherantly causes instability, can be called a "Feature". I don't buy that explanation, as the facts are that the more the cpu draws, the worse the vdroop on the same motherboard, but on the DQ6, it has none, due to its 12 stage power delivery system. It sounds to me like somebody rationalized thing to justify the poor motherboard performance when the high power hog cpu's (quads) started to arrive. Maybe if Anand himself said it, I would believe it, but right now I don't.

My explanation fits all the facts. Their story is just that, a story that they would have you believe. Tell me any electronic device that WANTS variable voltage ! PSU are widely accepted as better, the more consistant the voltage delivery, and now they are trying to sell me on this ? No way.

 

[Sylvanas]

Originally posted by: Markfw900
I do not see how giving a cpu variable voltage, which inherantly causes instability, can be called a "Feature". I don't buy that explanation, as the facts are that the more the cpu draws, the worse the vdroop on the same motherboard, but on the DQ6, it has none, due to its 12 stage power delivery system. It sounds to me like somebody rationalized thing to justify the poor motherboard performance when the high power hog cpu's (quads) started to arrive. Maybe if Anand himself said it, I would believe it, but right now I don't.

My explanation fits all the facts. Their story is just that, a story that they would have you believe. Tell me any electronic device that WANTS variable voltage ! PSU are widely accepted as better, the more consistant the voltage delivery, and now they are trying to sell me on this ? No way.

I'm not going to quote 2 pages of the article but you must read all of it to understand all that is going on. In short, Vdroop ensures that the VRM's don't need to work as hard in order to maintain a supply voltage that does not exceed the maximum allowable CPU voltage limit (CPU VID). Now when you take away CPU Vdroop as Asus has done via the 'Load Line Calibration' setting in the BIOS the VRMs must work much harder at maintaining a constant voltage and you get spikes way above the CPU limit. The tests done in AT's review we're at 1.25v now if you were to be running a higher Vcore the Spikes would be much higher and the results are best summed up here:

The first is that setting lower voltages with this option enabled actually results in a condition in which the CPU voltage under load is higher than the idle voltage. Imagine our confusion as we desperately struggle to understand why our system is Prime95 stable for days yet continues to crash under absolutely no load. What's more, in spite of the absence of droop and for reasons unknown, enabling this feature artificially raises our CPU's minimum stable core voltage at 4.0GHz from 1.28V to about 1.33V. As a result, our system uses more power under load than is otherwise necessary. Our efforts to reduce our processor's supply voltage backfired - instead of lowering the system's total power consumption we managed to affect a 20W increase.

Originally posted by: Markfw900
I do not see how giving a cpu variable voltage, which inherantly causes instability, can be called a "Feature".

Bottom Line its a safeguard if you want your CPU to last and be MORE stable.

 

[imported_Iketh]

"Imagine our confusion as we desperately struggle to understand why our system is Prime95 stable for days yet continues to crash under absolutely no load."

lol that was my problem exactly.. it's gone now that i got the sm bus controller driver installed... my new stable high is 246 ref but i cant run 4-4-4-12 at that speed, so im testing if 238-240 @ 4-4-4-12 is faster than 246 @ 5-5-5-14 in f@h

 

[Markfw]

Its an ASUS, my 2 ASUS boards are crap. I still don't believe it. When they test a DQ6, then I will believe. I have personal experience thats proves (at least to me) that the boards are more stable. My P5K (without and vdroop fix enabled) couldn't stay stable at 3 ghz with my B3 Q6600. Changing nothing but the motherboard (to a DQ6) now results in a rock solid 3320 with all the same hardware and bios settings, including vcore !

Again, I will believe it when they prove it to me. Thats article is crap in that regards IMO. And I have proof to support my contention. And Duvie has seen it also. Continuing to quote from the article won;t convince me without something other than the motherboard they are using, and possibly a link to an Intel spec that states this is a "feature". Right now, its just the authors opinion, and you know what thats worth.

 

[Duvie]

Bottomline it may be a feature INtel suggest, but I can tell you from ocing expertise enthusiast do not like, and the mobo manfacturers know this. They know we are ocing...or they wouldn't put the controls in...

I got an idea for the morons at INtel..instead of implementing a program where they undervolt under load...how about undervolt when at idle and the goddamn thing isn't working (wasting energy and generating needless heat) and boost it up to what the fvck I set it at when I want to run full load....I know it is speed stepping but I am suggesting dont change the speed...If you can drop it .06v under load, start .06 lower and boost it once I need the power...

That being said...

Congrats on the find so far Iketh. If you drop the ram timings can you go higher? Or is the performance hit to big. I know you are a folder so are you balancing your performance needs off of the frame times?

 

[Duvie]

You keep saying "by adding an option in their BIOS that effectively disables this important function". What? I dont see an options in my bios for that? I didn't see one in marks....What did he turn off? I dont think they actually gave us the choice. The only thing I disabled was virtualization and spread spectrum...everything else is stock. Under folding at home load I see maybe .02v max droop. I had more then that on my last amd neo2 board...

 

[Sylvanas]

Originally posted by: Duvie
Bottomline it may be a feature INtel suggest, but I can tell you from ocing expertise enthusiast do not like, and the mobo manfacturers know this. They know we are ocing...or they wouldn't put the controls in...

I got an idea for the morons at INtel..instead of implementing a program where they undervolt under load...how about undervolt when at idle and the goddamn thing isn't working (wasting energy and generating needless heat) and boost it up to what the fvck I set it at when I want to run full load....I know it is speed stepping but I am suggesting dont change the speed...If you can drop it .06v under load, start .06 lower and boost it once I need the power...

That being said...

Congrats on the find so far Iketh. If you drop the ram timings can you go higher? Or is the performance hit to big. I know you are a folder so are you balancing your performance needs off of the frame times?

The question is not if it is good for overclocking, I was responding to whether it is an important feature on behalf of Intel- you yourself have said that you would rather Intel not use this *feature* which can be called a feature due to there being a way to turn 'it' off in a recent Asus update- once again the Asus bios option is just an example- it may be different for Gigabyte boards. It cannot be argued that Vdroop is not a feature, the extent of it may be lessened on high end motherboards however that doesn't prove that cheaper motherboards have an inherently poor power delivery system by design- just that small things have been added to enthusiast boards to reduce Vdroop, I used the example of Vmodding with a $2 resistor earlier which produces the same results as a '12 phase $300' board.

 

[Tempered81]

duvie, mark, iketh & sylvanas: what is the board to have for ocing a 8400 wolfdale? Brand, chipset, model...

I take it you guys could concur on a perfect recommendation.

edit: i've heard x38 45 and x48 and that having 12phase power is important

 

[Sylvanas]

Originally posted by: jaredpace
duvie, mark, iketh & sylvanas: what is the board to have for ocing a 8400 wolfdale? Brand, chipset, model...

I take it you guys could concur on a perfect recommendation.

edit: i've heard x38 45 and x48 and that having 12phase power is important

Depends on what you want to use it for, if you want Crossfire then go with the X-38 or DDR3 then go with the X48 however IMO can't go past a P35 for value for money and it'll be fully compatible with Intel's 45nm CPUs. As for manufacturers the big 3 Asus, Gigabyte and MSI all have great boards its up to you which one they'll all be around the same give or take a few SATA ports- personally though I've had good experiences with MSI boards so that would be my pick.

Back OT, Iketh I take it you are running the memory @ 1:2 if so can you run 1066mhz?

 

[imported_Iketh]

not atm on this board, i need 266 ref and the 1/3/08 beta bios raised it from 250 to about 259-260 max, and i cant get it stable at 255 with 10.5x cpu and 8x-7x nb/htt... if they get 280+ ref possible on this board with phenom, ill be happy

i cant get these corsair ballistix pc2-8500 above 940 @ 4-4-4-12, i've tried up to 2.33v, but they have posted @ 5-5-5-15 1200mhz...

 

[imported_Iketh]

Originally posted by: jaredpace
duvie, mark, iketh & sylvanas: what is the board to have for ocing a 8400 wolfdale? Brand, chipset, model...

I take it you guys could concur on a perfect recommendation.

edit: i've heard x38 45 and x48 and that having 12phase power is important

if i could purchase intel atm, i'd get a mature p35 board, spend an hour overclocking and be done with it... i wont buy brand new tech hot off the assembly again like i did here...

 

[Markfw]

Originally posted by: Iketh

Originally posted by: jaredpace
duvie, mark, iketh & sylvanas: what is the board to have for ocing a 8400 wolfdale? Brand, chipset, model...

I take it you guys could concur on a perfect recommendation.

edit: i've heard x38 45 and x48 and that having 12phase power is important

if i could purchase intel atm, i'd get a mature p35 board, spend an hour overclocking and be done with it... i wont buy brand new tech hot off the assembly again like i did here...

Well, if you want the best(and mature), Gigabyte P35-DQ6. Its less than the DFI 790fx-m2r you bought ! I don't know why you didn't go this way the first time. The only difference, is that the Q6600 is $50 more than your 9500. And as you have acknowledged, one hour and (with a good HSF) you would be at 3500.

 

[imported_Iketh]

i bought the phenom to replace my 4200+ on an m2n-e... the 2 extra cores were desperately needed

month later i bought the dfi...

that's basically how im in my position... my hopes are this dfi board will be the heart of my main rig for 3+ years, but nehalem is knocking