How are these temperatures?

[Mogh]

I see people reporting lower temperatures so not sure if these look right.

43C idle (after PC has been on 10 minutes)
54C-58C in game

I am now using Shin Etsu on retail HSF. I tried Arctic Silver 5 and temps were even higher. Not sure what more I can do. I have 2 Antec 120mm fans (1 front, 1 back) in an SLK-3700-BQE. It is a bit loud especially when playing games.

Room temp 68F
CPU FAN 3700 idle, 4300 in game

Specs in sig.

 

[imported_waldo]

They aren't bad, but could be a little lower. You won't run into any problems, but I run Xeons at 39C Idle, and 45C load, and I thought AMD's were supposed to run cooler no?

 

[Mogh]

Originally posted by: waldo
They aren't bad, but could be a little lower. You won't run into any problems, but I run Xeons at 39C Idle, and 45C load, and I thought AMD's were supposed to run cooler no?

I am beginning to wonder if that is actually true.

Your case is identical to mine except for the color. Are you using two 120mm fans to achieve those temperatures?

 

[Amaroque]

That seems hot. My Winchesher @ 2.6 GHz doesn't get hotter then 50c. Full load, Prime95.

 

[gwarbot]

My p4 doesnt get hotter than 24c even under heavy load and its on stock cooling. It really weirds me out.

 

[Mogh]

Originally posted by: Amaroque
That seems hot. My Winchesher @ 2.6 GHz doesn't get hotter then 50c. Full load, Prime95.

From what I read Newcastle .130mm runs hotter than Winchester .90mm core. But not sure what the difference is.

I contemplated on going with a 3500+ because of this but did not think the temperature difference would be that much different.

 

[JBT]

A tad high IMO but certainly isn't going to hurt you. It really depends on the cooler you are using. IF you are using retail then yeah that seems about right.

 

[Mogh]

Originally posted by: JBT
A tad high IMO but certainly isn't going to hurt you. It really depends on the cooler you are using. IF you are using retail then yeah that seems about right.

Yes, I am using the retail cooler. This was my first dive into AMD and my loudest system I have owned because of the temperatures. Never had these issues with P4's although I have yet to try a Prescott.

 

[Amaroque]

Originally posted by: Mogh

Originally posted by: Amaroque
That seems hot. My Winchesher @ 2.6 GHz doesn't get hotter then 50c. Full load, Prime95.

From what I read Newcastle .130mm runs hotter than Winchester .90mm core. But not sure what the difference is.

I contemplated on going with a 3500+ because of this but did not think the temperature difference would be that much different.

I know there is a difference in heat, but I didn't think it was that much. I thought the difference was only about 2c. That just seems a bit warm to me.

Try running the Prime95 "torture test" That will really heat the CPU. If you don't get any errors after an hour or two, I wouldn't worry too much.

 

[daveshel]

Too hot. I set thresholds at 60.

 

[Spike]

Originally posted by: daveshel
Too hot. I set thresholds at 60.

Personally I set my thresholds at 52c just to be safe (don't usually go over 46c full load during the summer and 42c full load in winter)

Those temps are fine for your CPU, especially using the stock heatsink. I used to run an AMD 1800+ and I got frequently into the 60's when gaming (64 was my highest I remember) and after over a year I never had a problem with it. I still used that CPU in my second computer (3 years of usage now) and it works fine, though my temps are much lower due to a better case/HSF

-spike

 

[Cawchy87]

Load doesn't go over 41 for me with an AMD xp-m 2400 @ 2.2 ghz.

I have an antec sonata which is very similar. But I have an ALX 800 U hsf. Mabye invest in a better hsf.

 

[Amaroque]

Originally posted by: Cawchy87
Load doesn't go over 41 for me with an AMD xp-m 2400 @ 2.2 ghz.

I have an antec sonata which is very similar. But I have an ALX 800 U hsf. Mabye invest in a better hsf.

Most ppl that quote load temps, are comparing apples and oranges. Unless you're water cooled, that AXP (esp if it's a Barton) will be hotter then 41c running Prime (closed case) in a 22c room.

 

[Mogh]

Anyone with a 3400+ 130nm process like to post your temperatures?

 

[Dethfrumbelo]

I have a 3200+ 90nm and at 2.55 GHz it's 32-33 degrees idle and 43-44 degrees under Prime95 in a 76F room. I have pretty much the same case as you.

I'm using an XP-120 though, and I lapped the hell out of it so that there's a copper mirror on the bottom. I used ceramique because of the short burn-in time.

 

[Mogh]

Originally posted by: Dethfrumbelo
I have a 3200+ 90nm and at 2.55 GHz it's 32-33 degrees idle and 43-44 degrees under Prime95 in a 76F room. I have pretty much the same case as you.

I'm using an XP-120 though, and I lapped the hell out of it so that there's a copper mirror on the bottom. I used ceramique because of the short burn-in time.

Just comparing mine to yours it sounds like 90nm is a -10c difference. Anyone here with the a 130nm Athlon 64 so I can know for sure if these temps are correct?

 

[Amaroque]

Just FYI... Deth is using a very high end HSF, and he has it lapped.

You could try cleaning the off the CPU, and HSF. Then reaply your TIM.

Also, If there is a large difference between the case side off, and on. I would look into better cable mgmt, and better case air flow.

 

[Insomniak]

Wee bit high, those temps.

I have an Athlon 64 3000+ .09 clocked at 2.2Ghz (400Mhz OC) and it idles at 31, and hits 44 under load.

You're fine as long as you're below 60C, but as always, the cooler the better.

 

[Cawchy87]

Originally posted by: Amaroque

Originally posted by: Cawchy87
Load doesn't go over 41 for me with an AMD xp-m 2400 @ 2.2 ghz.

I have an antec sonata which is very similar. But I have an ALX 800 U hsf. Mabye invest in a better hsf.

Most ppl that quote load temps, are comparing apples and oranges. Unless you're water cooled, that AXP (esp if it's a Barton) will be hotter then 41c running Prime (closed case) in a 22c room.

My house is not 22c for sure. Plus, my computer is in my basement which helps. This is taken after long gaming sessions. But I am running prime torture test as I type this so I guess i'll see.

 

[ver00]

I get 38~40 c idle and about 50~52 c under load.

Whats this prime95 thing your talking about? Where can I get it, I did a forum search but couldn't find anything.

 

[drpootums]

I have the same CPU as you (3400+ nc at stock) and with my Zalman i get 44-48 C idle and 50-54 C load. I know mine is high too, but i have a combination of 2 thermal pastes (long story), and a lot of them on my HS/F and i think that's the reason. I wouldnt worry too much (i think my BIOS reading is a bit high though, so urs might be too).

 

[Mogh]

Originally posted by: drpootums
I have the same CPU as you (3400+ nc at stock) and with my Zalman i get 44-48 C idle and 50-54 C load. I know mine is high too, but i have a combination of 2 thermal pastes (long story), and a lot of them on my HS/F and i think that's the reason. I wouldnt worry too much (i think my BIOS reading is a bit high though, so urs might be too).

I tried Arctic Silver 5 and it was actually higher. Now I am on Shin Etsu and I am hoping given a few more days temps will drop. I spent alot of time with %90 alcohol to be sure all previous grease was off.

My next step was to get rid of the stock HSF and do something like you have tried. But maybe considering your results it may not make a difference.

 

[Amaroque]

Originally posted by: Cawchy87

Originally posted by: Amaroque

Originally posted by: Cawchy87
Load doesn't go over 41 for me with an AMD xp-m 2400 @ 2.2 ghz.

I have an antec sonata which is very similar. But I have an ALX 800 U hsf. Mabye invest in a better hsf.

Most ppl that quote load temps, are comparing apples and oranges. Unless you're water cooled, that AXP (esp if it's a Barton) will be hotter then 41c running Prime (closed case) in a 22c room.

My house is not 22c for sure. Plus, my computer is in my basement which helps. This is taken after long gaming sessions. But I am running prime torture test as I type this so I guess i'll see.

My point was that ppl use different programs to put a load on the CPU. I can run rthdribl which pegs the CPU at 100%, but my CPU load temps are only 42c from that (same Winchester).

It's difficult to compare CPU load temps because of so many variables... ambient temperature, CPU loading program used, Hardware monitoring program used, case open or closed, ect...

and that's assuming the same CPU type.