• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Is my E6600 a dud? - Runs Hot

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

spiderhole

Registered
Joined
Sep 27, 2006
Just put together my new system with following specs...

P5B dlx
E6600 (SL9S8, batch L627A906, 08/30/06)
Antec Trio 550 psu

What I'm finding is even with very conservative overclocking the temps seems really high. I started off with the Big Typhoon + AS5 and then switched over to Intel stock cooling just to compare. Both resulted in around same temps (~1-2 degrees). When I removed the BT, the thermal grease seemed to be perfectly spread (used recommended grain of rice method). Intel HS is now using stock thermal material.

Here are some results...

FSB @ 300 (2.7005 cpu), mem mult 2 (600), Vcore set to AUTO

Temps (Idle/Prime95 Load)

38/57c Asus PC Probe

45/63c CoreTemp

mb 35c


Stock Speeds (266/2.4ghz)

31/52c PC Probe

36/57c CoreTemp

mb 34c

After a couple of hours of heavy Prime95 testing at both speeds I didn't get any errors..but the temps just seem to hot to be safe. All of these readings are with my case cover off. They go up about 5 degrees with it closed. Am I crazy or do these seem really hot for such moderate overclocking? I'm tempted to try another aftermarket cooler, but after such poor results with the Big Typhoon I'm skeptical. Any ideas or suggestions appreciated.

Thanks,
Mitch
 
Well first.... Welcome :)
Second don't have the Vcore set to auto. Set it between 1.3-1.35V to be safe that the board is not jacking up the voltage when OCing. You should beable to hit 3Ghz easily within those voltage settings and even possibley 3.2Ghz and up.

The temps on the OC seem high, but the chips can take the heat, IMO though I'd lower them before going higher for long term clocks. For at stock speed my setup got 28C IDLE and 35C Load, I can hit 3.0Ghz with 1.325V (Orthos stable) and hit 32C Idle and 42C Load. As soon as you start ramping up your OC your tempature will climb quickly. Seems that clock frequencey drives temps higher then a voltage increase in my cases.
 
Ok, been running some more tests with P95 at higher speeds.

Went up to 320fsb (2.88ghz) then tried 333fsb (3ghz) and noticed my temps remained at about the same level at idle and load as 300fsb (+- 1 degree). However, at 333 P95 would give me the following error on both instances after a few minutes..

"FATAL ERROR: Rounding was 0.5, expected less than 0.4"

I have my Vcore locked to 1.325v right now. Should I try bumping up to 1.35 and give it another go at 333fsb?

Thanks,
Mitch
 
i would.your cpu is probably starved for voltage under load at that freq.
 
Possibly a very concaved IHS, just sounds like a bad mount to me though, not enough pressure.
 
Anyone who is not using Intel's new temperature thermal readouts is not getting realistic temperatures.

With Intel's new "Thermal Analysis Tool" (TAT) you will get a real output. This is now a digital thermal sensor and the new Core2 cpu's do not give proper readings with the old monitor methods of algorithmic ratio.

Download:

http://users.skynet.be/fb127723/TAT.exe

I have an E6400 @ 3.2GHz with 1.37V. With a Big Typhoon I get 43/42 Idle and 57/55 full load (Orthos).

When I use Everest and other methods to test, they show different results. Everest is 25C, Speedfan is 28C, etc.

The difference is due to the older softwares using a ratio algorithm to aproximate the thermals due to analogue sensors. The Core2 beasties have a Digital thermal sensor built into the core and which give real core temps. We have never had these readouts before so they might seem hotter or more extreme. They are not. Simply put, we were misinformed as to the true core temperatures before Conroe and are not anymore.
 
Barkiss said:
Anyone who is not using Intel's new temperature thermal readouts is not getting realistic temperatures.

With Intel's new "Thermal Analysis Tool" (TAT) you will get a real output. This is now a digital thermal sensor and the new Core2 cpu's do not give proper readings with the old monitor methods of algorithmic ratio.

Download:

http://users.skynet.be/fb127723/TAT.exe

I have an E6400 @ 3.2GHz with 1.37V. With a Big Typhoon I get 43/42 Idle and 57/55 full load (Orthos).

When I use Everest and other methods to test, they show different results. Everest is 25C, Speedfan is 28C, etc.

The difference is due to the older softwares using a ratio algorithm to aproximate the thermals due to analogue sensors. The Core2 beasties have a Digital thermal sensor built into the core and which give real core temps. We have never had these readouts before so they might seem hotter or more extreme. They are not. Simply put, we were misinformed as to the true core temperatures before Conroe and are not anymore.
i had no idea.thanks for the info.i was wondering why my 6600 was idling at 37 and my p4 idled at 32.mystery solved lol.
 
maxfly said:
i had no idea.thanks for the info.i was wondering why my 6600 was idling at 37 and my p4 idled at 32.mystery solved lol.

Your welcome. You have the "Core Temp" which also reads the true digital temp so I would advise you to follow that reading as a true temperature.

Here's my system idle and load temps.

SystemLoad.jpg

SystemIdle.jpg
 

Attachments

  • System Idle.jpg
    System Idle.jpg
    139.1 KB · Views: 71
  • System Load.jpg
    System Load.jpg
    145.7 KB · Views: 63
Last edited:
spiderhole said:
Ok, been running some more tests with P95 at higher speeds.

Went up to 320fsb (2.88ghz) then tried 333fsb (3ghz) and noticed my temps remained at about the same level at idle and load as 300fsb (+- 1 degree). However, at 333 P95 would give me the following error on both instances after a few minutes..

"FATAL ERROR: Rounding was 0.5, expected less than 0.4"

I have my Vcore locked to 1.325v right now. Should I try bumping up to 1.35 and give it another go at 333fsb?
What's your current setting for the Static Read Control setting in the North Bridge Configuration screen? I had the same problem with P95 once I started hitting the 330+ MHz FSB range, after disabling Static Read Control the P95 issue disappeared without having to increase vcore.
 
Thanks for all the great info and advice, guys. So, any recommendations for a replacement cooler for the Big Typhoon that isnt as tall? Even something that is an inch or two shorter would work much better in my case.

-M
 
the zalman hsf are pretty good but expensive.there may be a sticky somewhere to help you choose.
 
Ok, I ordered an Arctic Cooling Freezer 7 Pro for $16 (what a deal) and will see how that works. I chose it over the Zalman because of the price and the fact that I won't have to take my mobo out and remove the BT backing pad to install. Will post the results.

-M
 
I'm afraid what I'm looking at is a grossly concave IHS on my e6600 and will check and see when I go to install the Freezer 7 Pro. Lapping doesn't sound like my idea of fun, so hopefully I can get acceptable temps with the new cooler at 2.7ghz. Kinda makes me mad since most seem to be getting WAY better overclocks with the 6600 and even the 6400.
 
spiderhole said:
I'm afraid what I'm looking at is a grossly concave IHS on my e6600 and will check and see when I go to install the Freezer 7 Pro. Lapping doesn't sound like my idea of fun, so hopefully I can get acceptable temps with the new cooler at 2.7ghz. Kinda makes me mad since most seem to be getting WAY better overclocks with the 6600 and even the 6400.

I'm in the same quandry as you. No matter how hard I try to clamp down the typhoons mounting plates (They're actually BENDING) I can only get it to contact the outer 1/3rd of my IHS on my 6600. I believe I either need to RMA this chip or void warranty by lapping this chip. I've owned a northwood and a prescott and both of those prior chips had FLAT IHS plates. Buying a better HSF or a waterblock won't solve the problem of only 33% contact of the IHS with the heatsink. I'm thinking about lapping this guy since it's able to run at 3ghz @ only 1.25Volts dual prime stable. Imagine what it can do once it's adequately cooled.
 
spiderhole said:
I'm afraid what I'm looking at is a grossly concave IHS on my e6600 and will check and see when I go to install the Freezer 7 Pro. Lapping doesn't sound like my idea of fun, so hopefully I can get acceptable temps with the new cooler at 2.7ghz. Kinda makes me mad since most seem to be getting WAY better overclocks with the 6600 and even the 6400.

I've never lapped a cpu, but I don't think it's that big of a deal unless you go overboard with it. I was thinking of lapping mine because it's slightly concave, but my temps aren't too bad so I'm not worrying about it right now. If I do lap the cpu I'm going to use a whetstone that's used for sharpening knives. I'll start with the medium and finish with the hard. I doubt that'll give you the mirror finish that people go for, but the stock finish isn't mirrored either. That'll at least get it flat, and the rest can be made up with AS5. I mention the stones because I think that'd be easier to deal with than the typical sandpaper.

Btw, I'd only use natural stones for that. The synthetic (particularly cheap ones) could give unpredictable results. Start with the Arkansas medium, and finish with the Wa****a hard.<--LMAO wth? W*a*s*h*i*t*a


Addendum: It's a little late now, but I'm very happy with my Thermalright XP-90c. It works especially well with my specific setup due to my side case fan. I'm getting 55c at full load using prime95
 
Lap the cpu if you're willing to lose the warranty. After getting tired of seeing idle temps of 45C and load temps of 60C @ 3ghz @ 1.2V on my concave E6600 I decided to lap her. I was initially scared at first since I started seeing copper. :eek: but perservered. Only took about 10 minutes since I started with an agressive 400 grit moved on to 600 then 800, 1000, 1500. Wasn't mirror finish, but the improvement is remarkable. In the exam same settings Coretemp is now 33C idle and 49C load after 30 minutes of dual prime. If you're willing to risk it, lapping is recommended.
 
hUMANbEATbOX said:
if you care about being moral, the warranty is void the moment you OC it. :rolleyes:

True True, but even before I started o/c the chip it was already idling at mid 40's and hitting 60C while loaded on stock voltage. This is all due to my chip being concave since most don't run that hot to begin with.
 
Lapping a chip either way if its even alittle concave or even convex can help with temps the question is just how much.
 
I think your temperature are consider to be "okay". unlike mine, idle 50/ load 65-70 with 1.4875 (set in bios).
 
Back