found a new love

[songokussm]

just thought i would let everyone know about my experiances with my new opteron 165.

ordered the 14th from the egg. shipped out on the 14th. arrived the 20th (i hate ups).

default-
1800mhz
1.3v
30-idle
35-load

current overclock-
2800mhz
1.425v
34-idle
58 load (after dual priming for 40m)

max overclock-
1.5
3000mhz
idle 50
idle to high for load test.
no stress test could or could not be stable. temps to high.

i need better cooling water or i think phase is the new kid on the block. But ill prob wait until k8l, am3, and r600

i have the heatpipe version of amd's cooler and my stepping is ccbbe 0610dpmw.

--
edit: ok after some further testing ive decied to drop down to 2700mhz. so i can get some lower temps.
----

edit2: dropped votage down to 1.35 and stable at 2800....with 52 load temp. weird optys dont like votage
----

 

[IEC]

Nice OC. I wish I had an Opty 165... still chugging along on my 3200+ Winnie @ 2.4GHz.

 

[Skott]

If I remember correctly once you get to 2.7-2.8 there is a wall on aircooling with that chip. To get higher at a stable setting requires watercooling. 2.8 is a good OC for the 165 though. Dont be unhappy.

 

[songokussm]

update: 2800 failed to high temps. back to 2700 @ 1.375

Just a question. what should i beilve, bios or windows app? in bios i am setting votage to 1.375 and cpuz and speedfan are telling me 1.34.
after an hour of priming 2700 @ 1.375 (bios) my temps are 55-62. ochies

and i just remembered i used amd's heatsink goop, maybe as5 will have some say!

edit: just applied as5, no drop in temps. so that means the amd goop is pretty good.

 

[Tanclearas]

AS5 can take a while to have a decent effect. You might notice slightly better temps over the next few days/weeks. Not a huge decrease, but a couple of degrees.

Been running my 165 at 2600 for over a year. Couldn't quite reach 2700 with it. 2655 was the max, and that's just with the stock cooler. Still loving it, and it'll hold me over just fine for a while yet. I think it's probably the best CPU investment I've ever made.

 

[BlueWeasel]

Based on some of the user reviews on Newegg (which sometimes can't be trusted), it seems like the newer 165's don't seem to overclock as good as the older Opterons.

My X2 3800 is currently running at 2600Mhz but it takes 1.45v. Anything higher really isn't stable regardless of voltage.

 

[songokussm]

just switched it to a dfi ultra-d mobo and im having great results on one hand and wierd results on the other.

the good-
2700 (300x9) @ stock votage 1.350 (bios) very stable. high temps ~55 load 35idle

the weird when overclocked the system doesnt always boot. it just restarts. it detects everything, i can acess everything when booting from a diag cd. just sometimes it takes a few restarts.

and its very stable. 3 hours cpu torcher, 3 hours blend, two intances of superpi mod 32m. all passed.

 

[OcHungry]

Originally posted by: songokussm
just switched it to a dfi ultra-d mobo and im having great results on one hand and wierd results on the other.

the good-
2700 (300x9) @ stock votage 1.350 (bios) very stable. high temps ~55 load 35idle

the weird when overclocked the system doesnt always boot. it just restarts. it detects everything, i can acess everything when booting from a diag cd. just sometimes it takes a few restarts.

and its very stable. 3 hours cpu torcher, 3 hours blend, two intances of superpi mod 32m. all passed.

you probably have a weak PSU, Memory timings wrong, or low vdimm, chipset and hypertransport voltages. increase those voltages a bit and play w/ the clock Skew a little.
below is a link to clock skew guide for Asus and DFI board
http://www.overclock.net/faqs/119265-how-clock-skew-can-save-your.html

 

[songokussm]

psu 700watt
memory timings are from the spd values. and default memory voltage as my memory is currently not overclocked.
chipset and hypertransport have all be bumped up .1v. which according to dfi is unneeded, but me makes me feel better.

 

[tomt4535]

My opty 165 ccbbe 0610dpmw is at 2.8 stock volts right now. Runs stable and cool as can be. It does 3.0ghz but with alot of voltage and the temps/power consumption go up too much for me.

 

[KingofFah]

I've got an opty 165 CCBWE 0604TPMW (about as poor as they come). I can get 2.4 on stock volts, but have to go to 1.4-1.425 for 2.5, 1.475-1.5 for 2.6, and 1.55 for 2.7 I won't even test the 2.7, since at 2.6@1.5v the coretemp temps go to 57/65c during load. One core is significantly cooler than the other, and the hotter core is the one which errors and forces me to increase vcore to get it stable.

I keep it at 2.5ghz 1.425v which is plenty faster than my dying XP1600 AGOIA that had recently started crashing at even stock speeds, so I had to underclock to 1ghz, lol.

 

[songokussm]

wow king thats a big jump for you. for xmas this year i build about 9ish systems for family and all of them were just blown away at the proformance difference going from socketa to socket 939/am2 (mother in law 'upgraded to socketA...wheres teh evil smily?)

well my system is rock solid @ 2.7 @ stock voltage. 2.8 required 1.475v. not worth it. and anything above that requires water or better cooling. its kinda nice that i now how a chip that will seriouly last me until am3

 

[smokeyjoe]

I've got a 165 CCBBE that I can run on ORTHOS for an hour with no problems at 315*9 and 1.35v on my ASUS premium. This is with my RAM set at 200 mhz 3-5-5-8 2T in BIOS - basically the largest/slowest divider and really loose timings.

I have tried changing the divider to 266 mhz which would put me at 2835/14 = 202.50 Mhz with the same loose timings, but I can't even get into windows. Kind of bummed about that, I thought this RAM would be ok with a 2.5 Mhz increase?

I am still new at overclocking, anyone have any suggestions on what to set my RAM?

 

[songokussm]

i had a similar problem. on my dfi board i changed all of the advanced ddr settings to auto and normal. system is rock soild now. So that may help you as well.


windows boots at 3006mhz but temps are very high so no testing

 

[zanejohnson]

lets see some benches.... i wanna see how it compares to my 930D @ 3.7

 

[songokussm]

sure! let me know what benches you like and ill run em

 

[aigomorla]

Originally posted by: BlueWeasel
Based on some of the user reviews on Newegg (which sometimes can't be trusted), it seems like the newer 165's don't seem to overclock as good as the older Opterons.

My X2 3800 is currently running at 2600Mhz but it takes 1.45v. Anything higher really isn't stable regardless of voltage.

from my experience with my X2, its a toledo however with the CCBWE steppings, it OC's better on the 9x then the 10x. It always fails at 280x10 vs 311x9 Why i have NO CLUE. Also on the x9 settings it requires less voltage which i HAVE NO CLUE also. But you might want to try x9 settings.

Originally posted by: zanejohnson
lets see some benches.... i wanna see how it compares to my 930D @ 3.7

My brother P4D which is right now at 4.3 will pwn my AMD X2 at 2.8ghz if ur currious on performance. When i was tweeking it, at 3.8ghz its roughly on par with a 2.6ghz AMD. and once u break that 4ghz barrier, hyper transport KICKS MAJOR BUTT. If intel could push the old P4D's past the 5ghz barrier im sure they would of raped the C2D's right now


EDIT: Im talking about prestler P4D's and not smithfields

 

[songokussm]

i went looking at water cooling kits today. im totally lost. can someone point me in a good direction?
i want-
killer cooling
quiet
vga/cpu able
complete kit
very reliable
uv would be nice but not required.

doesnt have to be water, ive read some things about pelter cooling, but i have no clue. ill wait for some direction.

 

[songokussm]

phase cooling sounds to complicated

 

[Markfw]

Originally posted by: aigomorla

Originally posted by: BlueWeasel
Based on some of the user reviews on Newegg (which sometimes can't be trusted), it seems like the newer 165's don't seem to overclock as good as the older Opterons.

My X2 3800 is currently running at 2600Mhz but it takes 1.45v. Anything higher really isn't stable regardless of voltage.

from my experience with my X2, its a toledo however with the CCBWE steppings, it OC's better on the 9x then the 10x. It always fails at 280x10 vs 311x9 Why i have NO CLUE. Also on the x9 settings it requires less voltage which i HAVE NO CLUE also. But you might want to try x9 settings.

Originally posted by: zanejohnson
lets see some benches.... i wanna see how it compares to my 930D @ 3.7

My brother P4D which is right now at 4.3 will pwn my AMD X2 at 2.8ghz if ur currious on performance. When i was tweeking it, at 3.8ghz its roughly on par with a 2.6ghz AMD. and once u break that 4ghz barrier, hyper transport KICKS MAJOR BUTT. If intel could push the old P4D's past the 5ghz barrier im sure they would of raped the C2D's right now


EDIT: Im talking about prestler P4D's and not smithfields

Hyper transport is an AMD feature. And it will take a 4.3 ghz P4D to equal a 2.5 X2, 3.8 runs more like a 2.2. Got benches ? I do, and not superpi, but real benchmarks. And a Conroe @ 3.5 = twice the speed of an X2 @ 2.5, so a P4D would need to be 8.6 ghz to keep up with my Conroe.

 

[Markfw]

Originally posted by: zanejohnson
lets see some benches.... i wanna see how it compares to my 930D @ 3.7

This has been hashed a lot of times.... 930@ 3.6 ~ X2 @ 2.1-2.2

 

[Duvie]

Originally posted by: Markfw900

Originally posted by: zanejohnson
lets see some benches.... i wanna see how it compares to my 930D @ 3.7

This has been hashed a lot of times.... 930@ 3.6 ~ X2 @ 2.1-2.2

http://www23.tomshardware.com/cpu.html?modelx=33&model1=434&model2=481&chart=181

this shows that as well...2.2ghz X2 just better the a 3.6ghz P-D....

as well as this one

http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=194
http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=188
http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=182
http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=169


http://www23.tomshardware.com/cpu.html?modelx=33&model1=434&model2=464&chart=177

Here the 2.4ghz X2 beats the 3.6ghz PD

http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=193
http://www23.tomshardware.com/cpu.html?modelx=33&model1=434&model2=464&chart=177


Here the 2.0ghz 512kb cache model X2 is faster then the 3.6ghz P-D

http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=192


Here is the rare occasion the 3.6ghz only gets beat by the 2.8ghz X2

http://www23.tomshardware.com/cpu.html?modelx=33&model1=436&model2=464&chart=165


Bottom line is that 2.2ghz is about a 3.6ghz P-D....EE versions are a bit of an anamoly due to the HT cores they have...

If we assume AMD and INtel scale similarly as they go up in speed then commonsense says....

A) When i was tweeking it, at 3.8ghz its roughly on par with a 2.6ghz AMD. and once u break that 4ghz barrier, hyper transport KICKS MAJOR BUTT. If intel could push the old P4D's past the 5ghz barrier im sure they would of raped the C2D's right now

THIS IS FALSE...perhaps was using an EE version which make it not so ridiculous

According to results above and this person wasn't just running 1 test the ratio says 3.8ghz P-D would only be as fast as 2.3ghz...


B) My brother P4D which is right now at 4.3 will pwn my AMD X2 at 2.8ghz.....

THIS IS FALSE...

If you take same ratio as above a 2.8ghz X2 will be equal to a 4.5ghz P-D or a 4.3ghz P-D would be equal to a 2.6ghz X2. There would be no owning....1-2 test out of the 14-16 at Toms would even show anything like that.

 

[songokussm]