• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

What do you consider "safe temps"?

R

rayman2k2

Supreme [H]ardness
Joined
Aug 6, 2002
Messages
8,069
Some of you might remember I had to take down my Celeron 2.66ghz a few weeks ago due to temps (it was averaging 61 degrees Celsius per Speedfan).

The CPU itself is in a eMachines, so modification isn't an option (I'm gonna try using some dryer hose and do some duct work however).

That said, what do you guys consider "safe" in terms of temperature when folding? The PC itself is used by my dad and my younger brother (so folding until it breaks isn't a route I want to go on).

Also, anyone fold on a Dell before? I have a Dimension 4400 series, and the thing only has one fan that is ducted to pull air over a fanless heatsink. Any options here?
 
Nothing directly related to your post, just something I found out thermal load testing a new case design I did.

Intel MB 945 GTP
2.8 Dual Core chip or a Celeron 2.66 CPU

CPU fan, front and rear case fans

Temps by speedfan

Fans controlled by the bios.

Under max load, blocking intakes of fans (simulate failure/dust build up ect) the front/rear fans would not leave lowest speed (~800 RPM) until the CPU reached 61+ degrees C. I thought this was pretty dang warm, but if the bios was not speeding up the fans then I guess the box itself did not think things were pretty warm. The fans would speed up to ~2200 RPM once > 65 degrees or so. CPU one spins faster, then front then rear case fans.

Under max fail condition, with dual core... 74 degrees C I think it was the highest. Still Primeing away under two instances 100% stable. I'll have to check my notes, but it was warm.... edit: CPU fan still working

Now, I hope the gear I'm using is better than what's in an eMachine/Dell (don't count on it... :) ) temps on MB/PSU have to be bad. CPU's are all the same, the supporting gear is what I'd be worried about.

You can always reduce the % FAH uses, I knock this down to 75% on production machines and thermal load test @ 100% just to try and be semi safe
 
Try dusting it out, if possible or needed.

I like to keep my main rig under 60°C and 70°C for processors in general.

Pentium 4 130nm process Pages 68-69.

Pentium 4 90nm process Pages 64-65

Pentium 9xx series Pages 76-77

s475 Celeron 130nm process Pages 82-83

s478 Celeron D Pages 63-64

LGA775 Celeron D Pages 72-73

Not all of that pertains to you, I think yours is in the s478 Celeron D pages, I just wanted to post this all somewhere so I could reference it. It's just to give you an idea of what Intel thinks. I think we can ignore the minimum temperatures though, Vapochill anyone? :cool:
 
My dualie box runs 2 1900+ MP's at 80 degrees under full load with air cooling :( I've got a watercooler in the works tho
 
sobe said:
My dualie box runs 2 1900+ MP's at 80 degrees under full load with air cooling :( I've got a watercooler in the works tho
Click to expand...



Celsius? :eek:
 
sobe said:
Yes

Click to expand...
HOLY CRAP MAN....add some fans, cut some holes and get that box out of the hole its in....i know that those run hot but you can do better than that.



 
Normally, I say not to worry about temps, because most people get worried in the low 60s when I don't think they need to worry. But damn, 80 is hot! I doubt my old Tbird@1.4something ran that hot and that was the hottest processor I've ever used. AXPs (when I paid attention) were usually low 60s processors when overclocked and I don't even pay attention to A64s since they pretty much don't get hot.
 
I actually had to stop 5 of my enterprise desktops from folding as they would overheat and shutdown. (when running at 50% only!!!)

Needless to say the cases had inefficient airflow, manufacturer cheaped out on fans etc... but in the end they are what corporations out there use so if anyone is thinking about rolling it out at work/school I'd recommend you do some testing before you do it.
 
I consider anything under 50C to be safe. It's only when I see temps in the mid-50s that I start to get crashes. I do have to reduce my overclock on my primary system in the summer; the AC barely keeps up, and I do start to see occasional crashes. (from 2.53ghz in cooler months, reduced to 2.4ghz in summer).
 
Bill Clo said:
I consider anything under 50C to be safe. It's only when I see temps in the mid-50s that I start to get crashes. I do have to reduce my overclock on my primary system in the summer; the AC barely keeps up, and I do start to see occasional crashes. (from 2.53ghz in cooler months, reduced to 2.4ghz in summer).
Click to expand...


QFT.

Above 50, and you're really smacking your CPU down. Every ten C is a halving of processor life.
 
mwarps said:
QFT.

Above 50, and you're really smacking your CPU down. Every ten C is a halving of processor life.
Click to expand...

Please site facts.
 
marty9876 said:
Please site facts.
Click to expand...

You're kidding, right? It's a known fact that the hotter the chip, the shorter is lasts.
This is real electromigration, not the fake kind people used to talk about in here. Besides that, you have diffusion. It's enough for me. Not here to start a war or preach to the masses.
 
mwarps said:
You're kidding, right? It's a known fact that the hotter the chip, the shorter is lasts.
This is real electromigration, not the fake kind people used to talk about in here. Besides that, you have diffusion. It's enough for me. Not here to start a war or preach to the masses.
Click to expand...

No, I'm serious. Intel lists thermal specs on many chips > 65 C. Many go up to 85 C

For me, 60+ C is getting hot in my experience. Different chips handle things differently. AMD is a whole nothing ball of wax I assume.

Again, from the testing I was doing the motherboard does not even in start spinning up the fans until over 60 C.

Saying a chip @ 52 C will last 1/2 it's life is a little drastic in my opinion. Please, shed some light here if I'm way off.
 
marty9876 said:
No, I'm serious. Intel lists thermal specs on many chips > 65 C. Many go up to 85 C

For me, 60+ C is getting hot in my experience. Different chips handle things differently. AMD is a whole nothing ball of wax I assume.

Again, from the testing I was doing the motherboard does not even in start spinning up the fans until over 60 C.

Saying a chip @ 52 C will last 1/2 it's life is a little drastic in my opinion. Please, shed some light here if I'm way off.
Click to expand...

You aren't way off. It all depends how long you want your stuff to last. "LIFETIME" for a CPU can be as long as .. well, I don't know! My 8086 is still running. 18 years old. I don't think it's ever been over 40C. My 486DX4/100 was constantly over 70C.It died after 11 years. Anecdotal evidence, but there are empirical formulae to back it up.

The only reason the thermal specs are so high on these chips is because intel and amd want them to die within a reasonable amount of time, the same way that Microsoft releases a new version of windows every 2 years. Do we need XP or Longhorn or Vista over 2000? Not really! They won't die in two years, maybe not even 4, but ten? Yeah, they won't last that long, unlike the Dinosaurs of the Beginning, when no one even knew what was coming down the road.

But if you'd like to test the hypothesis, run a nice new chip at its thermal spec limit 24/7 and see how long it lasts. I'd do it for science, but I'm fresh out of money to blow, thus I keep my shit below 50C.
 
I'm sure that temps reduce CPU life, just like high core voltages - it's what the experts say. From my experience though, it won't matter to any of us. We'll have tossed that CPU in the trash years before it fails or sees any degredation in performance. I've been doing DC stuff since way back in the RC5-64 days and have overclocked just about every processor I've ever used in the office (hundreds of 'em) and at home and I've never seen a CPU failure - even after 4 or 5 years at 100% usage. I've still got a couple of Duron 700s @950 running in the office with no problems whatsoever and with the crappy coolers we had back in the day they run just about hot enough to burn you.
 
Mwarps, I was a huge fan of WIN2K, but I dare not say that XP wasn't a needed improvement. Besides, let's compare apples to apples here. Put the tinfoil hat away, please. :)
 
Comparing almost anything to a 486DX4/100 I think is really not a fair comparision.

Heat kills computers, thats a fact. How much, for how long who knows. Out of hundreds and hundreds of CPU's, I've have very few fail. I think I can count on one hand the number which have failed, and 1/2 of those were my putting +50% more voltage through the thing than I had planned.

Huh, just thought of this. I think I killed a 3.2 Intel chip, 0 C were the temps. I think there might have been other issues going on there.... :p

XP rocks! Well, for the most part.
 
AtomicMoose said:
Mwarps, I was a huge fan of WIN2K, but I dare not say that XP wasn't a needed improvement. Besides, let's compare apples to apples here. Put the tinfoil hat away, please. :)
Click to expand...

The phrase of the day is "Planned Obsolescence" :p Now, given, If I'm running wireless, I'm damned well gonna have an XP box. Also, XP might be a bit speedier. Otherwise? Dunneeedit.

Tin Foil Hat? That's just plain ig-nant, dawg. :D

Mary said:
Comparing almost anything to a 486DX4/100 I think is really not a fair comparision.

Heat kills computers, thats a fact. How much, for how long who knows. Out of hundreds and hundreds of CPU's, I've have very few fail. I think I can count on one hand the number which have failed, and 1/2 of those were my putting +50% more voltage through the thing than I had planned.

Huh, just thought of this. I think I killed a 3.2 Intel chip, 0 C were the temps. I think there might have been other issues going on there....

XP rocks! Well, for the most part.
Click to expand...
Eh, not really, silicon is silicon is silicon. Heat does it all the same with silicon. a 486 has the same basic diffusion and electromigration equations as a Conroe as an A64.

Like I said, lifetime? Who the hell knows how long it'll last, my 8086 still lives. I have no clue how. Maybe a chip at 40C will last for 40 years, I really don't know what the predestined lifetime is, but I do know the hotter the shorter.

0C? Condensation much? Tried that once. Peltiers are not as much fun as I'd have thought.
 
I'm happy to report that my MP's are not running at 80 degrees, there seems to be an inaccuracy in the sensor detection under linux. They are running around 66 degrees with a full load.

 
I have a P4 2.26 in an ASUS Terminator barebones. Poorly designed, BIOS says the CPU is at 94C I'm sure its wrong, but it is VERY hot.
 
You must log in or register to reply here.
Back
Top