4300 Temps

[vanilla_guerilla]

i just got a 4300 to fool around with while my e6600 is being rma'd (machine check errors when running 64bit os's). its a week q705 from newegg. my idle temps are about the same as the 6600, 38c, but they go up to 80c during full load testing like tat or orthos. this is with vcore at 1.46v bios, around 1.41v with vdroop. i am using an ultra 120 and as5, which hasnt cured yet (got cpu last night) but i will only pick up 2 or 3 degrees from that at most. and out of the 6 dual core cpu's i have owned, this is the only one where core 1 runs hotter than core 0 at idle. about 3 degrees. i can play games and do 3d testing up to 3600mhz, but only orthos stable to 3435mhz so far. is 80c too high to run at for long periods, like 4 or 5 hours of gaming?

 

[Battl3fr0nt]

I would think 80C would be a little to high for running for 5 hour's.. I would never do it that is for sure but tommarow I am going to try and get my C2D E6420 to 3.2 with the stock HSF I know alot of people do it but I dunno I gotta do it while keeping my vcore 1.4 I will check back here and give the temp's If I do it and it work's out ok but as for you.. when playing game's I don't think it will get that hot Orthos will have you're CPU running much hotter then while in game. I am still kind of a noob at this so Im sure other people will have a better anser for ya..

 

[unclewebb]

It's quite possible that CoreTemp and TAT are misreading your Tjunction temperature. Some of the later E4300 processors have a Tjunction of 85C and not 100C so core temperatures are being reported 15C too high.

How do your temps compare to your E6600 if you had it in the same system at the same voltage? What does SpeedFan report for core voltage. It's more accurate than CPUz once the voltage goes over 1.40 volts.

When you're gaming your temps are not going to be anywhere near your temps when running Orthos so you'll be fine.

 

[montypythizzle]

It's quite possible that CoreTemp and TAT are misreading your Tjunction temperature. Some of the later E4300 processors have a Tjunction of 85C and not 100C so core temperatures are being reported 15C too high.

How do your temps compare to your E6600 if you had it in the same system at the same voltage? What does SpeedFan report for core voltage. It's more accurate than CPUz once the voltage goes over 1.40 volts.

Still anything at 63C is still pretty high considering he has an ultra 120. This sounds very disturbing, I would try reseating the HSF, or put a Thunderblade on that sucker

 

[unclewebb]

Still anything at 63C is still pretty high considering he has an ultra 120.

Depends on room temperatures and case air flow as well. Many people are running their computer at 80F now so higher temps can be expected.

The numbers that AnandTech were getting were NOT while running Orthos so some people's expectations are too high. AnandTech's test procedure was running FarCry, a single threaded game, so temperatures are going to be higher when both cores are maxed out running Orthos.

 

[vanilla_guerilla]

Still anything at 63C is still pretty high considering he has an ultra 120. This sounds very disturbing, I would try reseating the HSF, or put a Thunderblade on that sucker

its not the ultra120. i checked and its making great contact. but i am a little disapointed with it, as i havent seen any cooler temps than when i used an xp90.

 

[cornelious0_0]

I'm sure unclewebb would stand behind the research just as much as anyone...but you should really check out this thread for a little more info on E4300 temp monitoring...as there's a fairly "up in the air" chance that you have either an 85C or 100C tJunction and should be using either 0.94 or 0.95 of Core Temp.

 

[vanilla_guerilla]

if you check my screenshot you will see that i am using coretemp 0.95. when i tried to use 0,94, it said that my cpu didnt support temp monitoring. 0.95 says my tjunction is 100c. but i have read in several places that coretemp reads this incorrectly on all 4300's. besides, the tjunction temp, is the temp limit that can be reached at the outside edge of the cpu chip, where it touches the card it is mounted on. its not the limit of cpu temps before throttling or shutdown. i was just wondering if the large temp diff between idle and full load (over 40c) is normal, and if its ok to run at or around this temp for several hours.

 

[cornelious0_0]

It's not that Core Temp 0.95 reads E4300's incorrectly at all, it actually does just what TAT does...and "guesses" at the tJunction (throttle temp) being 100C...because some L2 revision C2D's are 85C and some are set at 100C...but there's no way software can tell atm.

What you need to do is clock your CPU down to 200x6 if possible and lower your vcore to about 1.100v. Take note of your room temp and see what your idle temps are at the lowered speed/voltage. At that slower speed/voltage the chip gives off next to no heat, and if Core Temp 0.95 and/or TAT are saying that you're running ~15C or so above your room temp then they're wrong and you actually have a tJunction of 85C and Speedfan/Core Temp 0.94 will do the trick for you. The other thing to check is look at your idle temps while underclocked/volted with Speedfan/Core Temp 0.94, if they report your idle a couple degrees or more below room temp then it's obviously incorrect and you know that you can rely on Core Temp 0.95 or TAT.

It seems silly, but it honestly took me less then 10 minutes to do the switch and let it sit idle...but I did confirm that in the above scenario I idle about 4C above my room temp, which is roughly what should happen, and I was able to come to the conclusion that TAT and Core Temp 0.95 are wrong in my scenario, as I actually have a tJunction of 85C.

Even if the chip is just a temp until the E6600 gets back, I'd take the time to do the above testing and see what happens, so that at least you'll know 100% for certain if you're going by the right temps...because there is no other way to know for sure, as Intel hasn't released any tJunction values for different stepping/batches of L2 revision chips.