First off, Hey all. I stumbled across this forum and have been reading lots of good info. I have a coupla questions.
I have a msi KT3 Ultra 2 running a barton 2500 clocked to a 2800 using default voltage. The box runs like a top 24/7 but I'm wondering if the temps shown in the bios and pc alert(both report same temps) can really be accurate. At idle the cpu temp averages around 96F. If I play sof2 or any other 3d game for hours, even if I just let q3 run all night long as a test, the cpu temp never gets above 105F according to the monitors in bio and pc alert. The case temp stays right around 80F according to pc alert.
I have a single antec fan with heat sensoring exhausting from back of case and for cooling cpu I use an old zalman copper hs, the one that has a black plastic shroud(sp) on top fitted for a 60mm fan sucking air off cpu. I've been using same hs since I first bought a t-bird 1400 coupla years ago, but I've since put a 60 to 80 adapter on it and have a 80 mm vantec stealth fan sucking air off cpu.
We do run the AC a lot so the house is usually 70F or lower, but I have a p4 2.0 in another msi box and it gets hotter than the barton with basically same cooling setup(p4 has similar zalman hs, but shroud is fitted for 80mm and it blows onto cpu) and case(smaller antec cases one exhaust, two intake ones, but I dont have intake fans) The p4 idle temps are about same, but on load it gets up around 120-125F. Both cases are surprisingly quiet to me considering I use to cool a t-bird 1400 on a7a266 heh. That box sounded like a helicopter.
But do the barton numbers seem right or do they seem just too low? I'm wondering if the msi kt3 is reading temps the same way the msi max2 is?
barton temp question
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
I use MSI KT3 Ultra-ARU (one model older than yours). I'd say the sensor in it is accurate, albeit being a socket sensor (MSI insists that is is a die sensor, but I don't think so). This should be the case for your AMD board too, whereas the PIV board shows the "true" CPU die/core temp. This should explain the difference for the most part. The CPU load temp seems to be bit on the low side for your Barton system though:
I use 1800+ Palomino, stock voltages. Should put out 66W max load, which should be close to your Barton. I use Swiftech's MCX462 with 26 dB/34 cfm Papst (8412 NGM). For case fans, I have two 12 dB/19 cfm Papsts (8412 NGL) for exhaust, one for intake.
While browsing the Net, and playing some Mp3's, room temp is 26°C, and CPU is 35°C, case 33°C. At load the CPU temps will go over 45°C (in winter I had CPU 43°C with 20-22° ambient). I recently played with undervolting the CPU fan, and under load I had max temp of 58°C (ambient was 30°C).
Or the Barton just puts out less heat. I thought of buying a T-bred, but the heat difference is too little to justify the purchase.
Cheers,
Jan
I use 1800+ Palomino, stock voltages. Should put out 66W max load, which should be close to your Barton. I use Swiftech's MCX462 with 26 dB/34 cfm Papst (8412 NGM). For case fans, I have two 12 dB/19 cfm Papsts (8412 NGL) for exhaust, one for intake.
While browsing the Net, and playing some Mp3's, room temp is 26°C, and CPU is 35°C, case 33°C. At load the CPU temps will go over 45°C (in winter I had CPU 43°C with 20-22° ambient). I recently played with undervolting the CPU fan, and under load I had max temp of 58°C (ambient was 30°C).
Or the Barton just puts out less heat. I thought of buying a T-bred, but the heat difference is too little to justify the purchase.
Cheers,
Jan