X2 owners that overclock....

[chinkgai]

i just bought a pc&pc turbo cool 510 express (not the sli)

here are system specs

X2 4200+
OCZ PC4000VX
DFI Lanparty nF4 SLI-DR
X850 XT PE
WD Raptor 74gb
and three other hard drives (300gb, 250gb, 200gb)
PC&PC Turbo Cool Express 510
BIX III, MCP655, TDX

also got other misc stuff like fans and optical drive, etc


first question, what voltage do you run? overclocked of course....

second question, what does ur pwm ic (power supply sensor temp) say its at?

mine is over 50c during the day (mid 80s C outside, no AC) under load...

this is running 1.623v right now with 3.3v to the ram, all else stock

 

[n7]

I'm not familiar with non air OCs...is that water you're on?

I'd say the vcore you're using is rather high, but i guess for non-air it's okay.

50C under load is perfectly fine; you can get up to 65C before you should get too worried.

 

[chinkgai]

oh nice to hear, whats ur ocz load at?

yes im on water with only the cpu in the loop, so no chipset or gpu block to heat things up so i run nearly 20c cooler than if i were on air (had slk948u)

edit: and at what vcore

 

[chinkgai]

oh and i noticed that u didnt have an x2

with my venice at near the same voltage, it didnt run nearly as hot! in fact, nothing in my system did

thats why i switched to water

 

[n7]

I run my Venice at 1.55V.

I can't be much help much since i don't have an X2, but i have been happy with my Modstream.

TBH though, i don't pay much attention to it.
It runs everything beautifully; that's all that matters to me.

 

[rise]

pwmic isn't the psu. its the area to the right of the cpu retention bracket, inbetween that and the ide/power connectors. its good to 100c easily and most say even 110c is fine.

edit- i should add, mine is ~38 idle and ~45 load but i've seen it hit around 52c on hot days. i know some of the screenies i've seen at dfi street people run it up to 60c.

 

[chinkgai]

oh, lmao

 

[Furen]

It's the motherboard's power regulators (which feed the cpu, ram, etc). X2s literally use up close to 2x as much power as a single core venice/san diego at the same frequency, so the load on the motherboard is much higher (but since most mobos are made to be able to use the 130nm FX-55, they do just fine. As an aside, Intel didnt have anything that had higher power requirements, which is why they needed to go to new motherboards).

 

[GICodeWarrior]

Originally posted by: Furen
It's the motherboard's power regulators (which feed the cpu, ram, etc). X2s literally use up close to 2x as much power as a single core venice/san diego at the same frequency, so the load on the motherboard is much higher (but since most mobos are made to be able to use the 130nm FX-55, they do just fine.

2x the power? Not according to this article...

From: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2410&p=2

How much more power would a dual core Athlon 64 X2 consume? To answer this question, we looked at three CPUs on the same platform and measured system power consumption. We used a 130nm Athlon 64, a 90nm Athlon 64 and a 90nm dual core Athlon 64 X2 - all clocked at the same frequency and with the same per-core cache sizes. The results are below:
(Image)
At idle, the Athlon 64 X2 falls directly between the 90nm Athlon 64 and the 130nm Athlon 64, but does the same hold true when the system is fully loaded performing our WMV HD encode test?
(Image)
AMD Single Core vs. Dual Core Power Consumption
Surprisingly enough, it is. Clock for clock, the Athlon 64 X2 will consume less power than a 130nm Athlon 64, and less than 20% more power than a 90nm Athlon 64.

~Rusty

 

[Heckler 5th]

Originally posted by: Furen
X2s literally use up close to 2x as much power as a single core venice/san diego at the same frequency...

ummm, no. guessing did not serve you well here.
edit: sorry to be an a$$ but your statements are flat-out wrong.

 

[Duvie]

Originally posted by: Heckler 5th

Originally posted by: Furen
X2s literally use up close to 2x as much power as a single core venice/san diego at the same frequency...

ummm, no. guessing did not serve you well here.
edit: sorry to be an a$$ but your statements are flat-out wrong.

Exactly... In his own flawed logic about why Intels needed new mobos he should have realized the reason we can run current X2's on current sckt 939 mobos is that the X2's run about as much as 130nm FX55 chips....I think above images confirm that...

NOw ocing can be quite a different beast as we raise vcore...so that is where mobos will really vary and the load will really be put on the power regulators....

 

[Furen]

Wow, I didnt expect everyone to jump out and flame me for that...

First I'll admit that I exagerated quite a bit (seems like you guys have never heard of hyperbole, note that i didnt have any malicious intent for doing it), but not as much as some of you guys think.

In Anandtech's review the measure the system power consumption so the rise in cpu power usage is diluted by the system's total power. I am not talking about how much your total system consumption increases but how much more load there is on the power management ICs on the motherboard.

Now, let's go back to my "close to 2x as much power" argument. First off, it is my understanding that Dual-core X2s use slightly lower voltages than single core a64s (90nm) because they're either Spec-ed to use less voltage or most motherboards just undervolt them.

Now here's a link that shows close to a 60% power consumption difference between a 3800+ (single core) and an X2 4800+ (yes, I know that the x2 has more cache per core but this is the only CPU power consumption chart that I can find at this moment):

Xbit Labs X2 3800+ review.

When you overclock and raise the voltage (as he IS doing) the slightly lower voltage advantage is negated and the total power gets closer to the "2x as much" mark. I think everyone has to think of their X2s as two cpus on the same die in order to understand that it's power consumption scaling compared to single core cpus is MUCH higher as frequency (and, most importantly, voltage) is increased.

Duvie: I admit that my argument about Intel needing new motherboards to use its dual core CPUs is somewhat fallacious, not because it is wrong, but because it sounds as though it is the only reason why they needed new motherboards. Remember THG's dual core "stress" test? Why do you think the 840 EE killed the epox NF4 motherboard? It's because the pwm ICs were able to cope with the CPU's power load. Also, I dont remember saying that X2s use more power than an FX-55, in fact I said "but since most mobos are made to be able to use the 130nm FX-55, they do just fine," so dont flame me for things I did not say.

Well, flame away... (and Chinkgai sorry for abducting your thread ^^)

 

[imported_wyrmrider]

I do not see a flame ware here just a discussion among gentelmen
check out the Lost Circuits tests referenced in this thread
http://www.amdzone.com/index.php?name=P...1&sid=232903e6bff51fa12f80b6ab17d5f40e
for another slant

chinkgai
try reducing your voltage a little bit at a time and see if it makes any difference
your individual transistor junciton temps can go up without showing on your monitor
it's the junction temp that can shorten your cpu life
the monitor shows an average whereas parts of your cpu may be hot, another cooler depending on voltage and load
do not use any moe voltage than you have to
there is a point where more voltage only adds heat and not performance- stay well away from that point on the curve

wyrmrider

 

[chinkgai]

hey guys, thanks for all the responses.

i've posted this up on xtremesystems and quite a few people pointed out to me along with someone else in this thread that the pwm ic temp i see isnt actually the psu temp reading. people on dfi street confirmed this and said its nothing to worry about at the current levels. some people report this temperature much more extreme than mine while priming.

currently when watercooled my cpu runs between 42-44 on dual prime load and i've been 8 hours stable at 2751mhz 1.623v with the above stated temps. i think i should be fine.

one weird thing is that usually on my previous antec psu's....they had a psu line that ran to a fan sensor line on the mobo but the pc&pc doesnt. maybe it does this internally based on load? interesting..

 

[Velk]

Mine runs around 60 at load using 1.45V and about 50 at load using stock (1.3V). Both it and the northbridge temperature increase much faster than the CPU does with overclocking, and are my main barrier to going above 2650, as I find the noise made by the chipset fan at full speed to be intolerable.

I have an antec p180 case, and I currently have the vga duct removed as I am lacking a fan that actually fits in the damn thing. Once I find one that fits I'll see how that changes things - I presume currently that one of the problems with the chipset temp is that the 7800GTX is pumping hot air at it and obstructing airflow, which is presumably why the case has a vga duct.