Another Opteron 165 Thread 0547 tpmw

[twoCenter]

new to this forum.
I received an opteron 165 PIB today. CCBWE 0550 UPMW.
Haven't experimented yet. I should have some results by the end of the weekend.

 

[lopri]

http://img382.imageshack.us/my.php?image=2799smallfft8hrs16gw.jpg
http://img382.imageshack.us/my.php?image=2799smallfft8hrs20wm.jpg

 

[mrkun]

Originally posted by: lopri
http://img382.imageshack.us/my.php?image=2799smallfft8hrs16gw.jpg
http://img382.imageshack.us/my.php?image=2799smallfft8hrs20wm.jpg

Nice desktop wallpaper.

 

[Chumster]

Originally posted by: lopri
http://img382.imageshack.us/my.php?image=2799smallfft8hrs16gw.jpg
http://img382.imageshack.us/my.php?image=2799smallfft8hrs20wm.jpg

I can't remember if you've mentioned this before - but what are you using for cooling? And in what case?

TIA!

Chum

 

[lopri]

I'd like to thank KiaMan, the OP of this thread, again. I had this chip for 5 days before opening the box - I was expecting a 0546XPMW and was disappointed to see a 0547TPMW. KiaMan sent me emails with his results and that convinced me to try this stepping. The result is quite satisfying.

Oh, and I don't have a case. It's my 2nd rig and an open testbed. I was going to get a case but couldn't decide on which, so tentatively had the setup on top of my file cabinet, and it ended up being left that way. (Too lazy to actually *build*) It doesn't look too bad, though.
http://img297.imageshack.us/my.php?image=desk4qm.jpg

For cooling, I use TT Big Typhoon, and have an additional 120mm fan to cool PWM area. I noticed the PWM temp rising through the roof with this board while testing, so got myself a Nexus fan. The whole setup is no louder than my main rig (in the sig).
http://img297.imageshack.us/my.php?image=setup1ft.jpg

Like I said before in this thread, I tend to believe my successful OC'ing is partly due to the active cooling in the PWM area.

 

[nealh]

Like I said before in this thread, I tend to believe my successful OC'ing is partly due to the active cooling in the PWM area

what kinda of temps are you getting in PWM

I have P180 with 120mm fans...top fan blowing in on my ram, top rear fan pulling air out,nexus fan on my SI 120, 120mm YL intake fan and 120mm YL blowing on the chipset

I have an Artic silencer on my x800XL supposedly remove heat as well

I turn the fan to blow on my SI 120 and oddly my PWM temp at load went from 62 to 57C

very strange to see such a big drop given it is a nexus fan which moves minimal air..also my chip set temp went up 1C and my HDD temps went up 2C...

Also I noted the PWM temps are higher at same vcore on my 170 vs 165..very strange...cpu temps are very close ....

I should point out my PWM temp was as high as 64C for nealry 13 hrs of dual Prime95 small fft with my 170@2700 with no errors(I stopped prime95)


any thoughts as to whether a lower PWM is more important vs lower HDD/chipset temps..granted the PWM is like 5C lower vs 2C on HDD and 1C on chipset

 

[ablatt]

Is the week the processor was made (0546 or 0547) the most important consideration or is the day of the week (X vs. S) also as important?

For example can we assume 0546 SPMW Optis are almost the same as 0546 TPMW or XPMW since they were made the same week or is every day totally different?

 

[designit]

Originally posted by: ablatt
Is the week the processor was made (0546 or 0547) the most important consideration or is the day of the week (X vs. S) also as important?

For example can we assume 0546 SPMW Optis are almost the same as 0546 TPMW or XPMW since they were made the same week or is every day totally different?

No, the day of the week matter a lot. the FPMW's looks to be the best week days and batch. Dont know much about week 0547, but CAB2E 0546 FPMW of any opty(144,146,etc) are turning out to be great overclockers.
Mine will do 2.9 w/ 1.5v and 3 ghz w/ 1.55. Temp hasnt gotten higher than 42c w/ stock cooling. there are some w/ 144's CAB2E 0546 FPMW's doing 3ghz.

 

[Chumster]

lopri: Thanks for the follow up. I'm currently waiting on my SI-120 (and to a lesser extent my P180) to get mine up and running. Glad to see the 0547's doing so well.

Chum

 

[nealh]

lopri chips is an exceptional one...if they all did this everyone who got it would be posting...2800 is not high % oc on any 165/170..if you expect this you will be disappointed

the 165 look in most to reach 2.5-2.6 ....2.7 is not common and 2.8 is rare(I think I have seen 3 around)

170 hit 2.6-2.7 in most cases...there are few more at 2800 but many need 1.55 v or greater.....

I have been burned twice with stupid expectations..I hope to get 2700 on 165(retail 534 mpmw)..got 2610
170 546 xpmw...hit 2700(1.5v 15 hr stable small fft) but anything more needs over 1.5v(not sure 1.51v will get 2750, pretty sure 1.55v will but my DFI SLI is noted to overvolt underload which would make my vcore closer to 1.58v...no thanks for 50mhz)

 

[The Sauce]

Just for the ignorant amongst us, could someone please define "PWMIC area?"

 

[mrkun]

Originally posted by: Snatchface
Just for the ignorant amongst us, could someone please define "PWMIC area?"

You can kind of infer that from the picture. I'm not entirely sure myself though.

 

[Lark888]

PWMIC area on the DFI Lanparty NF4 is the 4 small heat sinks between the CPU and the IDE slots on the board. If you look closely, you can see the temp. sensor adjacent to the "top" one near the memory slots on the CPU side. I have a small 40mm fan attached to these heat sinks and have good solid temps.

 

[Lark888]

Strange, I must have had some setting goofed up last week when I first attempted to overclock. My Opt 165 0547 TPMW chip wouldn't successfully boot above about 2450 and was stable at only 2350 or so. It would trip out of Prime95 in just a minute or two. Didn't improve with increased voltages either (up to 1.55v).

I decided today to find out the limiting component and followed the sticky on "Quick and Dirty Overclocking" taking notes. I reset all the BIOS to follow guidelines and now am running Prime95 on both cores at 9x300fsb while typing. I've had no troubles today starting at 250fsb and going up in increments of 5. I've only run dual prime95 for 15-45 minutes at each step so final stablity is not known but this is still way different than my first time. Voltage is set at 1.45 and reads out at 1.408 or so from two different monitoring programs.

9x305fsb passed the short test. 9x310 (2800 Mhz) failed on Core 0 after a short time; Core 1 ran over an hour before I shut it down to change voltages. Currently running 9x310 with about 1.46v indicated (1.5 selected in BIOS). Passed the short test on both cores...we'll see how it does on the longer run.

Nope, runs about 20 minutes on Prime95 and then one or the other core errors out. Was improved when I increased my CPU fan speed to 2500rpm. Chance that better cooling would maintain stable. For now, I'm going down to 2700.

I'll get some screenshots up later... (well, later is now)
2519 Mhz
2567 Mhz
2608 Mhz
2657 Mhz
2700 Mhz
2800 Mhz
2800a Mhz
2800b Mhz
2800c Mhz

 

[lopri]

On the DFI board, PWMIC is where those 3 aluminum heatsinks sticking out between CPU socket and IDE connectors. It basically controlls power regulation to the CPU. Without the active cooling, I saw the temp of PWMIC rising over 50C and generally ALWAYS hotter than the CPU itself. I didn't like it at all. Well, indeed I had that cooling for single-core CPUs, too. But dual-core seems to be a totally different ball game. Also, remember that my temps are strongly affected by the ambient temp since it has no coverage. When I turn the heat on to keep the house warm, I saw both the CPU and the PWMIC temps go over 50C. The temps in my OC screenshots are achieved @60F ambient temperature (That's 15C). So for better OC'ing, basically I'm enduring this winter with minimal house heat.

I also have similar experiences as Lark888's. This chip behaves very strangely. For instance, it passes Prime95 @1.48V but fails @1.5V. WTF? Also It does 2.7GHz @1.36V, so figured I'd find out the best 1:1 OC @2.5-3-3-8. Tried 2.4GHz (9x266) @1.36V and it fails Prime95. Again, WTF?
I don't know if it's the chip or the board, but definitely weird. Also I noticed there are a few *magic* combinations in the BIOS with this board. CPU Multi, VCore, VCore Special, etc. On my board the following seem to work well.

Auto x Auto
1.350 x 104% (1.360V CPU-Z)
1.400 x 104% (1.392V CPU-Z)
1.250 x 123% (1.488V CPU-Z)
1.350 x 113% (1.488V CPU-Z)
1.425 x 110% (1.520V CPU-Z)

VID Special 104% works best in general, and 113% seems better than 110%. I have no idea if this is a general case though. I've had this board for only over a month.

 

[mrkun]

It's been my experience that these DFI NF4 boards behave strangely.

 

[nealh]

Originally posted by: lopri
On the DFI board, PWMIC is where those 3 aluminum heatsinks sticking out between CPU socket and IDE connectors. It basically controlls power regulation to the CPU. Without the active cooling, I saw the temp of PWMIC rising over 50C and generally ALWAYS hotter than the CPU itself. I didn't like it at all. Well, indeed I had that cooling for single-core CPUs, too. But dual-core seems to be a totally different ball game. Also, remember that my temps are strongly affected by the ambient temp since it has no coverage. When I turn the heat on to keep the house warm, I saw both the CPU and the PWMIC temps go over 50C. The temps in my OC screenshots are achieved @60F ambient temperature (That's 15C). So for better OC'ing, basically I'm enduring this winter with minimal house heat.

I also have similar experiences as Lark888's. This chip behaves very strangely. For instance, it passes Prime95 @1.48V but fails @1.5V. WTF? Also It does 2.7GHz @1.36V, so figured I'd find out the best 1:1 OC @2.5-3-3-8. Tried 2.4GHz (9x266) @1.36V and it fails Prime95. Again, WTF?
I don't know if it's the chip or the board, but definitely weird. Also I noticed there are a few *magic* combinations in the BIOS with this board. CPU Multi, VCore, VCore Special, etc. On my board the following seem to work well.

Auto x Auto
1.350 x 104% (1.360V CPU-Z)
1.400 x 104% (1.392V CPU-Z)
1.250 x 123% (1.488V CPU-Z)
1.350 x 113% (1.488V CPU-Z)
1.425 x 110% (1.520V CPU-Z)

VID Special 104% works best in general, and 113% seems better than 110%. I have no idea if this is a general case though. I've had this board for only over a month.

Just remember what you set the bios vocre is the real vcore..not the bios reading or cpu-z

so 1.325*113%=1.5v
1.425*110%=1.57v...also it has been shown at idle you get vcore setting but under load the DFI Ultra-D/SLI DR overvolt 0.025

FWIW..it seems my PWMIC temps are higher with my 170 vs 165 and I never stay below 55C at load and was 15 hrs stable with PWMIC at 64C

I owuld love to give my 170 more vcore but I am afraid of the voltage and heat...1.55v seems to be the cutoff(15% voltage increase) and with DFI NF4 overvolting..1.5v is as high as I will go...

I have found something interesting..I was told at DFI to set chipset to 1.6v, LDT to 1.3v...but I am stable at least at 270 fsb with 1.2v and 1.5v....but temps did not drop as as expected

 

[lopri]

I've heard about it. But I have one question: If the board overvolts by 0.025V under load, why doesn't CPU-Z catch it?

(Actually I've seen it once with my 144. When set Auto x Auto, it was 1.36V on idle and 1.37V on load, according to CPU-Z)

Edit: This is what happens without active cooling on PWMIC. Ambient temp 20C/70F.
http://img367.imageshack.us/my.php?image=2799largefft8bh.jpg

 

[lopri]

BTW, can anyone run memory over 270MHz with ANY opterons? My memory passes memtest up to 310MHz, but it never passes Super PI over 270MHz. This is with at least 3 different single/dual-core Opterons.

 

[nealh]

Originally posted by: lopri
I've heard about it. But I have one question: If the board overvolts by 0.025V under load, why doesn't CPU-Z catch it?

(Actually I've seen it once with my 144. When set Auto x Auto, it was 1.36V on idle and 1.37V on load, according to CPU-Z)

Edit: This is what happens without active cooling on PWMIC. Ambient temp 20C/70F.
http://img367.imageshack.us/my.php?image=2799largefft8bh.jpg

I wish I knew...I asked this over at xtremesystems and they were emphatic about this issue..
http://www.xtremesystems.org/forums/showthread.php?t=77857&page=57
check posts 1422/1423

 

[KIAman]

I read somewhere at DFI-STREET that it was confirmed that the CPU voltages are EXACT what you set in your bios regardless of what windows or CPU-Z says. It was tested with a voltmeter.

They claimed there was no overvolt.

When I find the post, I will post links.

Thanks for the info on PWMIC. I was assuming that was my PSU's temps, LOL and was wondering why my PSU's fan wasn't going whacko when temps crept up past 50c.

 

[lopri]

I think I've found another magic combination.

1.425 x 104% (CPU-Z 1.424V)

2 hrs. Prime Large stable before I stopped it.

 

[nealh]

Originally posted by: KIAman
I read somewhere at DFI-STREET that it was confirmed that the CPU voltages are EXACT what you set in your bios regardless of what windows or CPU-Z says. It was tested with a voltmeter.

They claimed there was no overvolt.

When I find the post, I will post links.

Thanks for the info on PWMIC. I was assuming that was my PSU's temps, LOL and was wondering why my PSU's fan wasn't going whacko when temps crept up past 50c.

there have been numerous posts around including DFI street that the board is overvolting

what you set in the bios is what you get at idle but underload the vcore is higher

read post #23

http://www.dfi-street.com/forum/showthread.php?t=29656&page=2&pp=15

this is now several different people at xtremesystems and DFI street

 

[Crescent13]

Congrats man!! I get mine on Wednesday. I also got a Zalman CNPS 9500 (best air cooler out there IMO), and the Jetway Sli Mobo (only second in overclocking to the DFI, but with better overall performance). Can't wait to try it out .

 

[Lark888]

I'm declaring victory on the 2700 Mhz speed 9x300 for the Opteron 165. Played BF2 for hours last night and then ran 12 hrs 21 minutes of dual Prime95 without a glich. Even had the low speed rpm on CPU cooler XP-90.

By the way, my low temperatures on the DFI Chipset are from a modified Alpha Pal CPU cooler that I chopped up to fit around an ArcticCooling Silencer video fan. Overall sound is very low on this setup and can only be heard in a quiet room.

Thanks to the OP for posting - I was holding off on getting an Opteron and am very pleased with performance/value of this one.

Info on Stability fun