Shall I open the box?

[5t3v0]

Originally posted by: NewBlackDak

I'll try for more, but my motherboard only lets me have 1.4 vcore. That along with crap vdimm adjustment is making me look at these:
EPoX EP-9NPAJ
DFI nF4 SLI Infinity

I have the EPoX EP-9NPAJ-SLi & its a nice board. It is runing my Opteron quite happily on 280MHz HTT. My only criticism is that the vcore fluctuates a lot - it drop s about 0.025 when the cpu is under load & then over-volts by the same when idle. As a result I've had to use a higher vcore than I would have liked. I do use C&Q to lower this when not using it heavily. There's a new-ish BIOS out for it that adds 150 & 183 ram dividers. I'll let you know how it overclocks with the X2.

Originally posted by: NewBlackDak

Just talked to my wife, and she says "programs still open and both boxes are still red, but it's louder than usual". That's 14 hours and counting now!!

That's good news. Get the wife to stop it now - 14 hours is more than enough but if it fails before you get home it will just worry you. That's a typical wife comment to say that its louder than usual. Exactly the sort of thing my missus complains about.

Originally posted by: NewBlackDak

I put the HotFix on, but didn't install the driver. I thought the driver was just for CNQ?

There's an MS hotfix, an AMD driver (which I think you're right is for C&Q) and then there's the AMD optimizer. Personally, I would install all. I like C&Q - it saves power, reduces temperature which keeps my fan below 1000rpm, & it will increase cpu life which, although not relevant for me, might help the chap who buys it from me on Ebay.

 

[videopho]

Is the MS hotfix and the optimizer the same thing? Some here have said it is and you only need one or the other, whereas I've read elsewhere that you should have the AMD Driver in combination with both the AMD Optimizer and MS Hotfix installed. Anyone have any insight?


Depends who or where you talk/look to. I will try both tonight. Worst case you can always remove one or another.

hot-fix install

 

[5t3v0]

No they are all different. The driver and hotfix are for power management. If you dont use C&Q, you wont need them. That xtremesystems page you have posted a link to has seperate links for each which are the same as these below.

MS Dual Core Hotfix
AMD Dual Core Driver
AMD Dual Core Optimizer

Edit: Fixed broken links

Edit: Anyway, what's this got to do with the LDBFE? :frown:

 

[videopho]

Originally posted by: 5t3v0

Is it a LDBFE?

Mine is installed. Where would I see or tell what it is w/o removing hsf?

 

[5t3v0]

Originally posted by: videopho

Originally posted by: 5t3v0

Is it a LDBFE?

Mine is installed. Where would I see or tell what it is w/o removing hsf?

You cant.

 

[videopho]

Originally posted by: 5t3v0

Originally posted by: videopho

Originally posted by: 5t3v0

Is it a LDBFE?

Mine is installed. Where would I see or tell what it is w/o removing hsf?

You cant.

cpu-z reads

Family F
Model B
Stepping 1
Ext family F
Ext Model 2B
Revision BH-E4

Would that help?
BTW I lowered FSB to 1000 and o/c to 2.4ghz. Prime Torture test is passing first 2 rounds thus far. Keep my fingers crossed. Where would I change Vcore? Mine reads variably from 1.28v to 1.394v according to pc-wizard/cpu-z

 

[NewBlackDak]

Got home and stopped it after no errors for 18 hours and 4 minutes. Temp got up to 57C during the hottest part of the day, but apparently it was a non-issue.

I don't install CnQ or enable it in my BIOS. I use RMCLock. It gives you complete control over FIDs and voltage.

 

[myocardia]

I don't use it/can't stand it either. It's much better not to, if you're an overclocker. Manual control is much better, in my opinion. I use ClockGen, and it works extremely well for me.

 

[5t3v0]

Originally posted by: videopho
cpu-z reads

Family F
Model B
Stepping 1
Ext family F
Ext Model 2B
Revision BH-E4

Would that help?

Not really. There's a lot of variants of the E4 stepping depending on core revisions, wafer batch, where & when it was made etc. This is only stamped on the top of the IHS.

Originally posted by: videopho
BTW I lowered FSB to 1000 and o/c to 2.4ghz. Prime Torture test is passing first 2 rounds thus far. Keep my fingers crossed. Where would I change Vcore? Mine reads variably from 1.28v to 1.394v according to pc-wizard/cpu-z

In my bios, vcore adjustments are made in the "Power BIOS Features" page along with the frequency, mutliplier, divider settings. Check your manual for where yours is. If prime fails, raise the vcore by the smallest increment available and try again (mine is 0.025v). It is advised not to go above 1.45v on air cooling. Good luck.

PS. Keep an eye on temps & keep load temp under 60C.

 

[5t3v0]

I downloaded both RMClock and Clockgen way back but have been too lazy to try them out. CnQ has worked fine on my overclock so I've kept with it. It switches between 2 states: 2.5GHz/1.5v and 1.4Gz/1.2v & I dont even notice.

 

[videopho]

My new x2 still looks to be extremely unstable at 2.2g or higher. Prime95's torture froze sometime this morning. It had me restarted @ 6am but I had to leave for work. Will call my son later to verify. FSB is currently @ 1000mhz Vcore =1.4 Temp = 39c. (yeah I did raise vcore at bios start-up. May be I need to reduce voltage increment per 5t3vo's recommendation. )
The previous 3500+ (single) was dead stable at 2.4g on air eversince I'd used. What's up with this x2 this time around? PSU? Mobo?
Another thing I did last night was un-installed all the amd drivers/optimizer and did a fresh re-install prior to re-running Prime.

 

[5t3v0]

What's the temp & vcore while prime95 is running? Some boards dont deliver a stable voltage which might cause the crashing. Used to happen with my current setup until I raised vcore further to compensate. What memory divider are you using?

For testing the cpu, set the divider really low like 100 (1:2) so the memory is taken out of the equation. Also set HT multiplier to x3 to take that out of the frame too. Starting from your maximum known stable frequency, raise the cpu frequency e.g. 10MHz & run some prime95 small fft tests. If there's no errors, raise the frequency again and retest. When you start getting errors in prime95, raise the vcore by a notch (e.g. 0.025v) & retry. If it passes a few runs, start raising the frequency again. Keep on doing this cycle until you reach your max desired vcore and do a more throughout prime95 test. If it fails, drop back 5MHz & try again. Keep doing this until you can get at least 8 hours stability on prime95. Its time consuming but the only sure way. You are running 2 instances of prime95 with one on each core I presume?

I also found it more reliable to set memory timings manually rather than relying on SPD when using dividers so you might want to consider setting each option to the same as the default for 200MHz. When uderclocked, some memory timings may change - my previous GeIL did this and caused instability.

Other things to check for are that you have the latest BIOS installed and CnQ disabled while testing.

Your PSU seems it should be up to the job and the mobo has already be proven to 240MHz. Perhaps you just have a duff cpu?

Edit: There's a guy in this forum with the same PSU as you and a similar problem. Coincidence perhaps but worth keeping an eye on.

 

[5t3v0]

Wish I'd kept the Opty in. This 4200+ is only hitting 2.4 at anything resembling stability and taking 1.45v under load (1.5v idle) to get it. Currently priming at 58C. I could POST at 2.6 and run memtest but not boot Windows. Windows will boot at 2.5 but locks up under prime95. I'm sure it's this shitey Epox board letting me down again. Vcore regulation is shocking. This upgrade has been a bit of a regressive step. Should have gone with the 3800+ and saved myself £20.

Edit: Now up to 60C

 

[BlueWeasel]

My X2 4200+ results were relatively mediocre as well. I got a 0613 VPAW from ZZF on Friday and also had a friend's 0612 UPCW that we benchmarked.

Neither chip was stable past 2.6Ghz, even with 1.55v. The 0612 runs 2.5Ghz at 1.4v and 2.6Ghz at 1.52v. Apparently, the production weeks 0611/0612/0613 aren't that great.

While I'm happy to get 2.6Ghz, I'm kinda disappointed after seeing others pushing their 3800+ to 2.7Ghz+ on stock voltage. Like you 5t3v0, I wish I would have picked up a 3800+ instead.

 

[5t3v0]

Yeah m8, it sucks when the luck of the draw doesn't go in your favour. I'm getting instability in stress testing even at stock speeds which seems to be down to the 1T command rate. Dropping to 2T and raising vcore to 1.55v is getting me to 2.5GHz with some success (priming for 20 mins now which is 15mins longer than any other setting tried). This doesnt seem right though. I never had to use 2T with the Opty 144. Might just send this cpu back. I'm not superstitious, but I had a feeling this was going to be a bad one - hence the thread.

 

[TraumaRN]

Originally posted by: 5t3v0
Yeah m8, it sucks when the luck of the draw doesn't go in your favour. I'm getting instability in stress testing even at stock speeds which seems to be down to the 1T command rate. Dropping to 2T and raising vcore to 1.55v is getting me to 2.5GHz with some success (priming for 20 mins now which is 15mins longer than any other setting tried). This doesnt seem right though. I never had to use 2T with the Opty 144. Might just send this cpu back. I'm not superstitious, but I had a feeling this was going to be a bad one - hence the thread.

Mine is finally stable at 2.55 Ghz which I'm happy with at 1.5 V

I was hoping for 2.6 or better but oh well it's loads better than what i had before.

 

[5t3v0]

Originally posted by: DeathBUA

Mine is finally stable at 2.55 Ghz which I'm happy with at 1.5 V

I was hoping for 2.6 or better but oh well it's loads better than what i had before.

Considering mine needs 1.6v and 2T CPC to get 2.5GHz, that's a good result.

It's been suggested by Mucker that the Epox board is not reading the voltage correctly from the sensor and needs more to get the same result compared to other boards. My vcore certainly fluctuates excessively (mostly downwards under load) which is why I need a higher starting vcore. I guess you dont have this problem?

 

[Philippart]

i ordered an x2 in the current rush too, but got a 44th week 2005, i haven't overclocked yet though.

 

[NewBlackDak]

Got a DFI Lanparty UT nf4 Ultra-D Tuesday. Modded it to SLI, and popped in all my stuff. Don't know why I did the mod except for "I can".
Preliminary testing:
First night I took the Ram to 234 2-2-2-5-1T@3.1v and it ran memtest for ~10 hours without errors.
Last night I took the 4200 to 2700(245x11)@1.45 and it ran dual-prime small FFTs for ~12 hours without errors. That's as high as I'm comfortable with this cooling.

Will try the Ram at 245@3.2 tonight. If that works I'll be happy with a 2700 1:1.

Also, I put my Opty 148 in it before the X2, and it boots at 3.0 with 1.5v. I didn't try any stress testing with it, because I wanted to beat on the X2. I'm going to do the vmod to the 939Dual now, and see if I can't get a stable 300x10@1.5 with the Ram at the 133 divider.