Brand spankin new 1.33GHz K7 (it's a toaster)

[Guest]

So, not really to brag, but more as to inform and question.. I just got my new 1.33GHz Athlon. In a word fast. In two words, hot! I bought a retail box knowing it would be of the best quality (and it really is it came without having the L1 bridges cut). I installed the stock heatsink/fan combo that comes with it figuring that it's under warrenty for 3 years to run with it. 62C is the avg temp of the processor. I'm thinking that might be a tad toasty. If anyone knows if this is normal for the stock heatsink/fan combo please let me know. Also if there are some better ways to cool this think I'd love to hear them.

PS: Installing that heatsink was hard it felt like you were going to break everything in half

 

[Kelledin]

That's definitely higher than normal. My 1.33GHz runs at 40C idle, 46C under load, with a GlobalWin FOP32. My best guess is that the ambient case temperature is too high, which means that either your room is too warm, or the case has rather poor airflow.

Btw...no L1 bridges cut...lucky you! You got yourself a jewel there.

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

 

[gregorarch]

Technically that temperature is well within spec - AMD says their chips can run at up to 95 deg C. Personally, I wouldn't be happy with it.

Still, retail boxes don't usually come with very good cooling - just good enough to keep the processor within spec. If I were you I'd shell out for a high-end HSF and some artic silver.

 

[CALV]

1st rule of fitting a new cpu- get rid of the naff hsf that comes with it. I'd also recommend using a thermal grease of one sort or another. If you play on overclocking i'd DEFINATELY get a new HSF.Also, is the 62C under load or idle? I'd also suggest looking at your case cooling.
My axia 1 gig @ 1350 runs about 50C under load which is too high for my likings, this is with a "bit bigger than average" heatsink, I replaced the crappy little fan on it with a 60mm sunon until my thermo engine arrives. Run it with the case open and see if you see significant drops, if so, your case cooling needs some attention too.


Next time you wave - use all your fingers

 

[Guest]

That's under full load (rc5des). The case temp is around 76F, I have two huge fans, one for intake and the other for exhaust, pulled them off an old rackmount setup.

 

[Guest]

<<<1st rule of fitting a new cpu- get rid of the naff hsf that comes with it.>>>

why his cpu is covered for 3 years with that heatsink

 

[kavbear]

Covered for 3 years mean running within factory specs w/o overclocking. Like someone said, AMD's specs are up to 95 degrees celsius. But to get lower than around 60 degrees, he needs a betted HSF.

Are we there yet?

 

[kavbear]

Sorry about the extra post, but, is your power supply AMD approved? They make a big difference in how well the cpu HSF cools. AMD psu's have vents right by the HSF that draw air ove the HSF and then out of the case...

Are we there yet?

 

[Guest]

who cares if it runs hot if its under warranty, if amd ships inadaquete hs/f then the deserve to replace it if the chip burns up.
btw i don't think he mentioned overclocking.

 

[CALV]

Bob, I get your point, but evein if it IS under warrantly, ill bet AMD want the "old" one back before they send you a new one (if they dont try and blame the customer- Ive never dealt with AMD so I dont know)- but either way its TIME SPENT with a dead pc. Qbert posted here asking weather or not people thought it was hot, and asked for alternate ways of cooling, of course everyone has opinions, and if you are happy with a standard hsf then fine, I personally am not but thats my choice (ok, Im overclocking-but even so I wouldnt like 62C)- I just gave an opinion on what was asked.

Next time you wave - use all your fingers

 

[Guest]

i agree that its hot as hell but i think it sucks that somebody would need to upgrade the heatsink and fan with a retail boxed processor. he paid for one heatsink i'm sure he doesn't want to fork out more $ for another one. i'd just let it run hot and if it dies tear amd a new [-peep-].

 

[munkey]

i would tend to agree with bob. but if he doesn't want his processor to go belly up in a little over three years, after his warrenty is out. becuase that is what the processor life expetantcy is with that HS/F. then i would definatley go with an aftermarket HS/F.

is this reality... i thought it would more realistic.

 

[Guest]

can someone explain to me the L1 bridges thing. Why would they be cut? I'm a bit foggy on this.

Aklein

Life is hard...Live with it.

 

[Guest]

>i would tend to agree with bob. but if he doesn't want his
>processor to go belly up in a little over three years,
>after his warrenty is out. becuase that is what the
>processor life expetantcy is with that HS/F. then i would
>definatley go with an aftermarket HS/F.

Really? Do you have any references for that?
I've never seen any data on CPU life expectancy vs. temperature (or HSF type).
Thanks



In theory, there is no difference between theory and practice.
In practice, there is.

 

[munkey]

i don't have any refernces that say that excttly but if you think about it, why not. its common knowledge that running an electronic device hotter than its optimal temprature can shorten its life. plus even though i am an AMD supporter you have to keep n mind that they are still out to get as much money as possible. so why wouldn't they sell a product with a warrenty a little shorter than its expected life span. even Dell does it with there portable batteries, no matter what warrenty you buy with your machine the warrenty on the battery is only 1 year, when the batterys are expected to work properly for only 18 months.

but if someone has evidence otherwise please let me know.


is this reality... i thought it would more realistic.

 

[Guest]

>i don't have any refernces that say that excttly

That's what I thought.

I don't dispute that higher temps will reduce CPU life. But to claim that he'll have to replace his CPU in 3 years if he sticks with the stock fan is complete speculation.

To suggest that AMD is designing the CPU such that it will fail in 3 years with the fan supplied with the boxed CPUs is ludicrous. They don't have enough control over the operating environment (case ventilation, room temp, hour/day on, loading, etc.) to even begin to do something like that.

Besides, the CPU will be obsolete in 3 years anyway. If it dies at 3 years, you'll pick one up on eBay for $5


In theory, there is no difference between theory and practice.
In practice, there is.

 

[Guest]

Was that a K7 Master for a motherboard? If so, its thermal sensor is right up next to the cpu and is more accurate than most motherboard sensors. Hence the seemingly higher temperature.
Zoom

 

[kavbear]

I know he didn't mention overclocking, I was merely giving a reason why someone would upgrade. Also, so many here do oc that if he wasn't planning on it already, he may want to later. Additionally, I agree that AMD is responsible if he operates his cpu within specs, but to me, it's worth the extra few dollars to not have to go through the hassle of taking my rig apart, shipping the cpu off, and waiting 'till it returns. That's just my opinion...

BTW -I just looked at his first commment again and he mentions that the L1 bridges weren't cut. Why would he care, unless he's gonna oc?

No man stands so tall, as when he stoops to tweak his rig.<P ID="edit"><FONT SIZE=-1><EM>Edited by kavbear on 05/18/01 04:55 PM.</EM></FONT></P>

 

[kavbear]

The factory would cut the L1 bridges ao that the cpu is "locked." This way you can't change the multiplier only FSB, if your mobo allows it. I don't think the L1 bridges don't have anything to do with L1 cache, if that's where your confusion comes from.

No man stands so tall, as when he stoops to tweak his rig.