Poor Overclocking Newer (ie:0503) Winchester Onwers - Check This Out!

**andyOCZ** · 03-12-2005, 08:39 PM

OK here,

I bought an 0503 CBBHD Winnie and put it in my DFI NF4 Ultra-D only to be dissapointed by a low overclock (as I expected). Seems it would run memtest86 up 2650mhz, but no way would Windows load. I checked my voltage in MBM and Sandra and they were not following what I had set. I rebooted and changed the basic CPU VID value to 1.55v and I still had 1.42 volts (probably 1.45 which is default).

Frustrated by now, I decided to try lower the voltage to see if that would track. I tried 1.35v and it read fine in Windows. Hmmm..... so I set the voltage to 1.40v and the CPU VID Special Control to 123%. This read fine in Windows and i was able to hit 2610mhz stable in Prime95.

To sum up: I had a craptacular <2400mhz top overclock using the standard voltage adjustment in BIOS. By lower the voltage to 1.4v and using the overide at 123% I achieved 2610 P95 stable. Nothing ground breaking, but it turned a dissapointing CPU into the best one I have bought so far.

I have a few conspiracy theories floating around in my head here...........

Let us know if this helps you.

**dnottis** · 03-12-2005, 08:46 PM

I really hate the way the MSI boards are presenting the voltage for overclocks with the percentages. I found issues with the Neo2 not applying any more voltage when moving from 1.45 to 1.50 to 1.55, they all reported about 1.48 in CPU-Z. I did like the Epox 9NDA3 that had increments not these stupid percentages.

**Supertim0r** · 03-12-2005, 08:49 PM

Originally Posted by dnottis

I really hate the way the MSI boards are presenting the voltage for overclocks with the percentages. I found issues with the Neo2 not applying any more voltage when moving from 1.45 to 1.50 to 1.55, they all reported about 1.48 in CPU-Z. I did like the Epox 9NDA3 that had increments not these stupid percentages.

same problem i got with an older 3200+ / neo2
no problem now since i've got a 0448SPAW @ 2.8 stock volt

**Ackbar** · 03-12-2005, 08:51 PM

Originally Posted by andyOCZ

OK here,

I bought an 0503 CBBHD Winnie and put it in my DFI NF4 Ultra-D only to be dissapointed by a low overclock (as I expected). Seems it would run memtest86 up 2650mhz, but no way would Windows load. I checked my voltage in MBM and Sandra and they were not following what I had set. I rebooted and changed the basic CPU VID value to 1.55v and I still had 1.42 volts (probably 1.45 which is default).

Frustrated by now, I decided to try lower the voltage to see if that would track. I tried 1.35v and it read fine in Windows. Hmmm..... so I set the voltage to 1.40v and the CPU VID Special Control to 123%. This read fine in Windows and i was able to hit 2610mhz stable in Prime95.

To sum up: I had a craptacular <2400mhz top overclock using the standard voltage adjustment in BIOS. By lower the voltage to 1.4v and using the overide at 123% I achieved 2610 P95 stable. Nothing ground breaking, but it turned a dissapointing CPU into the best one I have bought so far.

I have a few conspiracy theories floating around in my head here...........

Let us know if this helps you.

Nice! I have a CBBID 0451 Winchester 3200+ that can only do 2350 at 1.6V! It automatically fails Prime at 2400 within minutes. If you can get 2600 out of yours, I'd say that's pretty good.

Wait a sec... that sounds a lot like my VNF4 where there seems to be a known bug that voltages between 1.45 and 1.6 (exclusive) are unattainable (ie. in neither the bios nor windows will any voltages set between 1.45 and 1.6 (exclusive) show up as such). Once I hit 1.6V+ the voltages actually change to that.

**~~The Mofo~~** · 03-12-2005, 09:13 PM

Well a lot of us have that crappy chip. Please report in and see if Andy's findings prove postive for others.

**dimasdw** · 03-12-2005, 09:23 PM

i got CBBID 0503 3500+
for super PI bench that proc can not pas 2700 Mhz ,in 1.55 v only benchable at 268x10 although i raise the vcore to 1.6 v the proc cannot get higher then 2680 Mhz....

**zA.Gosu** · 03-12-2005, 11:39 PM

Not bad

**bachus_anonym** · 03-13-2005, 12:13 AM

Well, someone fill me in here because I seem fail to understand something here

What is the deal with people not giving Winnies more than 1.6v? What about 1.7v? Is it high temperature you guys are reporting or what?

I have a CBBFD 0441 Winchester that is Prime95 stable on my DFI @ 2670MHz and it needs 1.7v to do that. At low ambient temps (13C) it does 2760Mhz fully benchable. Voltage in BIOS is set to 1.425/1.55/110%. As we speak, at normal ambient temps (21C), mobo outside the case between two other mobos - that account for some higher ambient temp, XP-90 plus 92mm quiet Sanyo Denki fan - I don't ever see more than 45C under LOAD with D2OL crunching @ 2670MHz...

Are those "crappy CBBID Winnies" running exeptionally hot or something that you guys can;t just give it 1.7v and see what they can do then? Or people just hope that every Winnie should be doing 2700MHz @ 1.55v and any other is crap ???

Fill me in here, please... No seriously...

**coop** · 03-13-2005, 12:23 AM

I have a cbbid 0451 3500, it also has a range in the voltage settings that read the same, but with so many variable ways to get almost the same voltage it is worked around. Mine don't do to bad though, in fact it seems to get better. Tonight I switched to 310p and got about 100pts increase in 3d01 to 24205.

**Ackbar** · 03-13-2005, 12:24 AM

Originally Posted by bachus_anonym

Well, someone fill me in here because I seem fail to understand something here

What is the deal with people not giving Winnies more than 1.6v? What about 1.7v? Is it high temperature you guys are reporting or what?

I have a CBBFD 0441 Winchester that is Prime95 stable on my DFI @ 2670MHz and it needs 1.7v to do that. At low ambient temps (13C) it does 2760Mhz fully benchable. Voltage in BIOS is set to 1.425/1.55/110%. As we speak, at normal ambient temps (21C), mobo outside the case between two other mobos - that account for some higher ambient temp, XP-90 plus 92mm quiet Sanyo Denki fan - I don't ever see more than 45C under LOAD with D2OL crunching @ 2670MHz...

Are those "crappy CBBID Winnies" running exeptionally hot or something that you guys can;t just give it 1.7v and see what they can do then? Or people just hope that every Winnie should be doing 2700MHz @ 1.55v and any other is crap ???

Fill me in here, please... No seriously...

I can't speak for other people, but I have tried up to 1.7V, but I have stock retail HSF so I dare not push it much more at least for extended periods. I did not gain any stability at 1.7V. Using 1.6V I could push to 2350, I went to 1.65 without being able to hit 2400, pushed a little further and noticed nothing as well. From what I can tell, my CPU is maxed out at 2350 without much superior cooling and even then who knows how much more I could get. Even if I could get more at 1.75V, I doubt it would be enough for most people to consider my CPU to be a decent overclocker.

**Geforce4ti4200** · 03-13-2005, 12:29 AM

2.68GHz is a VERY good clock for a cbbid but I guess the 3500+ are speed binned that way, most of the bad ones are the 3000+ and 3200+ speed binned. I see alot of people buying 3500+ cause there sick and tired of not hitting even 2.4GHz. I dont know how venice will be binned, but amd would be smart to do the same and force people to buy the 3500+ and 3800+ for the 3GHz overclocks while the 3000+ and 3200+ do 2.6 to 2.8

**Geforce4ti4200** · 03-13-2005, 12:33 AM

"What is the deal with people not giving Winnies more than 1.6v? What about 1.7v? Is it high temperature you guys are reporting or what?"

cause it makes no difference for the majority of winnies. they stop scaling past 1.55, sometimes 1.6v take mine for example, I can do 2.5GHz(or more) at 1.45v but I use 1.58v for 2.6GHz and more volts doesnt help, sometimes even prevents it from posting. I guess the new 90nm die just doesnt like volts, it does great at stock volts then rising the volts gains you just a bit more then thats it

**MaxxxRacer** · 03-13-2005, 12:35 AM

interesing andy.. i just hope and pray to all the false gods of overclocking out there that I get a good 3000winnie...

**coop** · 03-13-2005, 12:36 AM

I have had mine up to 1.796 (1.425x126) for testing. It did not do any better than it does where it is atm, 1.734 (1.275x136). It actually was not quite as stable. I don't know why. I use water and the temps are 32c - 34c only going to 35c when I had the voltage higher. One thing that keeps me from going past 1.8v even though the temps are low is the cost, $275.00 for the cpu. I know that going over 1.8v is not some magical number that if it is crossed it will ruin it, but if I can't see any diff. between 1.67v and 1.79v, I have not thus far seen a reason to go on. I've thought about the possibility of some sort of threshold that if passed will make a diff. but somethings can't be undone. (wussy) I'm still trying to get the most from here for now.

**andyOCZ** · 03-13-2005, 12:42 AM

bachus anonym,

The point was that many have these NEW Winnies that are reportedly not responding to voltage. I suspect that something is causing the BIOS to "stick" at default voltages, or possibly revert back to default voltages when loading Windows.

My theory is that by lowering the voltages below default, and overriding it using a percentage (ie: 123%) I was able to get around this issue. It seemed to work in my case. Where I had been adding voltage before using the 1.55 max VID, this was getting me no where, so I gave up, thinking I had hit the limit for my CPU.

Many are reluctant to put 1.7v into thier CPU, or may have cooling that is not the best, causing heat issues. These CPU's are sensitive to heat. I don't mind the voltage, but I just tried 2650 at 1.7v and failed P95, so it seems my CPU like about 1.65v at the most. I guess 2610 may be my limit. I can live with that.

This is not rocket science, it's just something that might help anyone that gave up like I nearly did.

**bachus_anonym** · 03-13-2005, 12:51 AM

andyOCZ,

Pardon, for sort of hijacking your thread... I just couldn't help asking this here...

And personally, I think that the information u gave here is very valuable :thumbsup:

**dimasdw** · 03-13-2005, 01:05 AM

I found heat problem with my CBBID its realy hot 250x 10 in 1.4 Vcore in 24 C room temperature ,ini load cpu temp is 52 C ..i use HT 101 with 56 cfm fan....use XP-120 with 100 cfm hit 48 C...i tried any kind method to put AS5 and the result is same...

i tried to OC my CBBID again with window opened temp room 8 c

for max clock hit 2739

i tried use 1.7 Vcore but while bench with sandra cpu aritmethic the cpu temp hit 71 so...the PC auto restart....

**xavicl** · 03-13-2005, 01:50 AM

9.5x multi sux, use 9 or 10 instead. I do 305*9

**one-silence** · 03-13-2005, 01:59 AM

i have the same problem with the vcore issue, no matter what voltage i change it to, it doesnt hit higher then 1.55. right now its set to 1.55 x 9.9% and its still reading 1.53-1.55. im pretty much stuck at 2.6ghz, i want more vcore! too scared to do a vcore mod.

week 0501 CBBID
http://silentproject.com/stable.jpg

**dimasdw** · 03-13-2005, 02:08 AM

Originally Posted by xavicl

9.5x multi sux, use 9 or 10 instead. I do 305*9

thats CBBID ? and what week ?

**xavicl** · 03-13-2005, 05:07 AM

its CBBID 0450

not so good, but no so bad than 0451. Stable is 300*9=2.7GHz

**Rudzer** · 03-13-2005, 05:55 AM

Will try that tomorrow in my piss-poor Winchester 3000+ that doesn't oc

... :/

**zA.Gosu** · 03-13-2005, 06:01 AM

Actually, 2.7 is in the top 1% of all winnies.

**IvanAndreevich** · 03-13-2005, 10:25 AM

andyOCZ
I noticed a similar thing on my Neo2, I think. It has the VID up to 1.55V and voltage adjustment up to 10% workable on most bioses. I was running 1.45 VID + 10% = 1.60V stable. Then I recently tried 1.5V + 5% = 1.575 (actual more like 1.55V) and it was stable for 24hr prime95! 1.45 VID + 8.3% = 1.57 was NOT stable when I tested it earlier (it was colder too in the winter). So, I think we're onto something here?

**grenadier** · 03-13-2005, 11:41 AM

So wait... how are you adding different %'s to your voltages? All I have is a list of numbers to choose from as far as I know...

Thread: Poor Overclocking Newer (ie:0503) Winchester Onwers - Check This Out!

Thread Tools

Search Thread

Rate This Thread

Display

Poor Overclocking Newer (ie:0503) Winchester Owners - Check This Out!

Bookmarks

Bookmarks

Posting Permissions