Normal Prime95 temps for 950 X4 B50 ?

[ramrodcar]

I'm currently running an S1283 heatsink with push/pull fans, 1.44v at 3.7ghz, hitting 58-59c. Is this normal? I want to up the voltage and frequency but I'm afraid with rising temps.

System was built last week, do I just need to wait for the AS5 to fully cure?

[richierich]

Sounds like you need to reapply the paste. Only place two thin lines (little bit longer than 1/2-inch) on the two middle rows where the heatsink meets the cpu.

[ramrodcar]

Still getting the same temp. I filled in the gaps (like i did before) between the aluminum and the heat pipes and laid 2 strips down the aluminum on both sides of the center heat pipe.

hitting 61/62 with case closed (58/59 with side panel open) at 1.488 @ 3.7.
19x multiplier fails (looks to be from heat as it gotes to 62, as i had to raise vid to 1.52)

Do i need to lap this heatsink? I know i haven't installed it wrong...

I'm using P180 case with thermaltake push/pulls on it along with 2 120mm exhaust and 1 120mm intake, all running high.

maybe it's ambient temperature of the room? it's about 78F in here...
idle temp is 42C

[Zeus]

How much degrees Celcius is 78F? I guess around 24ºC? Your temps are bad, how is the imprint of the cooler in the tim?

If that's not good you should lap the cooler.

You should be looking at high 40's low 50's i guess, not high 50's.

[Evil]

Looks OK to me. I´m using Thermalright HR-01 Plus, got pretty much the same temperatures. Right now running prime95 max heat 3,6ghz 1,45v, using 2000rpm TT fan and loading 51c. Room temperature 24c. If i change the fan, getting 5c higher temperatures under load. I have mounted heatsink several times but the result is still the same.:shrug:

[demonkevy666]

Looks OK to me. I´m using Thermalright HR-01 Plus, got pretty much the same temperatures. Right now running prime95 max heat 3,6ghz 1,45v, using 2000rpm TT fan and loading 51c. Room temperature 24c. If i change the fan, getting 5c higher temperatures under load. I have mounted heatsink several times but the result is still the same.:shrug:

it's not the heat sink or paste it's the IHS it's about 5mm thick off copper.
I don't think they're flat they are usually concave right in the middle.
lapping might help the cpu, or check the heat sink too if it's flat

[St Devious]

I'm not sure, but isn't 60C too high for AMD CPUs. isn't it like the critical temp ?

[Tom128]

My 955BE with the same cooler was hitting low 50-53C at 3.8Ghz and 1.45v if that helps. So your temps do seem high, especially since I feel my chip runs alot hotter than most (under water it's hard to keep cool above 1.45v even).

[Maregg]

I'm also running an unlocked B050 X4 cooled by a Xig 1283.
However, I'm at stock voltage and of a lower OC 3.6Ghz.

All I can say is that the room temp has a HUGE impact on my temps.
Where I live, the temp is all over the place. Sometimes I can get the room at 22C, sometimes it goes to friggin 30C :mad:.

At 30C, I can get my CPU to hit 55C. If I would be to push the vcore harder then stock, I'd guess I would it 60 pretty fast (65 is a throttling temp in bios for my config). Not that I would ever want to go over 3.6 Ghz quad.

If my room temp is at 22C, my CPU hardly reach 48C under load. Its THAT huge of an impact for me.


Not to derail your thread, but I would say I'm not particularly happy with the Xiggy purchase. I didn't know I would be stuck with the very awkward top-down fan configuration. My case isn't configured to allow strong airflow toward the top (no top fan). All it does is shooting the hot air from the CPU directly into my Power Supply.... !??? On top of that, but partly my fault for not researching, this orientation flaw is preventing me access to 2 of my 4 ram slots. Whatever was in their mind ?

I'm planing on getting a Mugen 2. Its supposed to be a pretty solid cooler and it comes with AM3 compatible backplates. Depending on the form factor, it might even give me access to tall heatsinked ram in my 2 first ram slots !

[shadow_419]

Idk if it's specific to my Cooler Master Hyper 212 Plus, but I got better temps just applying a thin layer spread over the entire ihs. Old school, maybe, but I don't have to worry about how well it spreads when I mount the cooler.

[ELItheICEman]

I have the 550BE (lapped) in a GA-MA790X-UD4P cooled by the Xig DK (which is the S1283). I had a custom mounting bracket made for it so that it will blow out the back of my case, and it seems to work well. As a dual, CoreTemp gives me 28C/34C and EasyTune6 reads about 12C higher than that. When I unlock the other cores, ET6 is the only utility I know of that will read any temps. I assume that ET6 still reads about 12C higher than actual temps since it says my idle is 40C for dual and quad. Anyway, my estimated load temp at 3.7GHz with 1.450V is around 50C. Leads me to believe something's just not right with your IHS. I never really measured temps on my chip before lapping, but I'm sure it helped me out a ton as my IHS was extremely concave :eek:

[Dopamin3]

Looks slightly high but nothing to really worry about.

[ramrodcar]

My temps above are from ET6, i guess i should see what the temp is with coretemp/AOD/ET6 and subtract the difference between coretemp/AOD and ET6....
is ET known to be consistently incorrect?

[ELItheICEman]

Just be sure to take it with a grain of salt. I don't know for sure whether ET6 is wrong or CoreTemp is wrong, or both. Don't let the reported temps on either run away on you though, just in case ET6 is actually correct.

EDIT:

Did a quick google search and found the following:


Ascaris
n00b



Joined: 11 Jul 2009
Posts: 2


Posted: Sat Jul 11, 2009 12:13 pm
I have the Gigabyte GA-MA790X-UD4P motherboard. Until today, it ran an Athlon X2-6400+ mildly OC to 3.3 ghz (100 mhz over stock) with the stock HSF (the bigger one with the copper base and the heat pipes). It idled (cpu case) according to EasyTune 6 (the Gigabyte utility) at around 37 and got up to about 60 when running Prime95. This was certainly within expectation. Both cores under CoreTemp 0.99.4 read slightly different numbers, but they were higher than the cpu(case) reading, as expected. (I am operating under the presumption that the reading reported simply as "CPU" by the EasyTune6 program is interchangeable with the cpu(case) temp, even if that is not technically true).

I didn't take much note of the exact readings of the cores when I was running the 6400+, since the AMD max heat spec of 63C refers to the case temp, with no core temp given. They were, to the best of my knowledge, about 10-15 degrees higher than the cpu(case) temps.

Now, I have upgraded to the Phenom II 550 X2. It is OC to 3.6 ghz (500 mhz over stock) and running with the Arctic Cooling 64 pro HSF (about three times bigger than the stocker). With the much bigger HSF and the thermal specification of the chip at 80w vs. 125w for the 6400+, I would expect lower temps even with the bigger overlock.

Now this chip only reports one temp for all the cores, not individual core temps as the 6400+ does. The reading at idle was 32C on the cores (as reported by CoreTemp) and 41 on the cpu(case) (as reported by EasyTune6).

Isn't a case temp higher than core temp actually... impossible?

I started up Prime95 and both the core temp and the case temp increased. The core temp topped at 42C while the cpu(case) temp topped at 54C. The delta is a little larger, with the cpu(case) temp again higher than the core temp!

This same motherboard was able to report what I think are likely to be accurate cpu(case) temps with the other cpu. Now, though, something is fishy with the core temps being lower than the cpu case temps, and I tend to suspect the cpu(case) temp is the screwy one. With the much bigger HSF and the chip with 45w less TDP, I would expect idle temp to be lower than with the 6400, not higher.

Now I know the load temp is really what matters, and that my load temp of 54 is well within the safe range. However, I do have a reason for trying to nail this down beyond curiosity!

This motherboard is one of the ones that is able to unlock the extra 2 cores of the Phenom II dual core. When I do this, the chip effectively becomes a 125w quad core Phenom (and probably a bit more since it is OC). But this bigger HSF should be able to hold the 125w Phenom temps down better than the relatively small stock AMD cooler did with the 125w 6400+.

With the quad cores enabled, unfortunately, I lose the ability to read the core temps! This is an unfortunate side effect of enabling the extra 2 cores... actual as-released Phenom quads don't have this issue.

With mine, Coretemp reports 0C as a quad core. However, EasyTune 6 still reports the cpu(case) temp. However, now it idles at 45C and gets up to 64 while running Prime95.... that's toasty, and is a degree higher than the max temp of the quad core Phenoms (and since that is effectively what I have with the cores unlocked, I think that to be the more pertinent max temp, rather than the 70C listed for the Phenom with 2 locked cores).

I think the rise in the cpu(case) temps when the other 2 cores are enabled is normal. I expect that! But that's too warm for comfort, and the question about whether the cpu(case) temps were accurate in the first place remains.

It's relatively cool in the room where the PC is located right now, but due to circumstances beyond my control, during some parts of the year, the room temperature gets 10-15 degrees F hotter than it is now, and I would like to know that my computer can take it!

What is my cpu temp, really? Anyone?


Ascaris
n00b



Joined: 11 Jul 2009
Posts: 2


Posted: Sun Jul 12, 2009 8:07 pm
Ok, I obtained a digital multimeter with a temperature probe. On my particular HSF, the copper base plate that contacts the cpu is slightly smaller than the heat spreader built into the cpu case, so it leaves a nice little "step" where I can stick the probe to get a ballpark reading of the case temp of the CPU (on the left side of the cpu as I face the motherboard).

I recorded the following numbers with my Phenom II 550 with all cores unlocked, with my Arctic Cooling Freezer 64 Pro cooler:

At idle:
Cpu(case)(as reported by EasyTune 6): 42
As reported by my multimeter: 33
(CoreTemp does not work with all 4 cores unlocked)

Under full load, Prime95:
Cpu(case)(ET6): 60
Multimeter: 44

Then I removed the Phenom II 550 and put my Athlon X2 6400+ cpu back in. This cpu has a TDP rating of 125w, the same as a quad core Phenom in the 3.0 ghz range (which is what my Phenom effectively is with all cores unlocked).

At idle, I saw these temps with the 6400+:

Cpu(case) as reported by EasyTune 6: 33
CoreTemp: 42,34 on the 2 cores
My multimeter: 33

At max load, prime95:

Cpu(case)(ET6): 49
CoreTemp: 58, 52
Multimeter: 44

I double checked my notes and it's true: both chips recorded the exact same idle and load numbers with my meter. Given the similarity of the CPUs in TDP, and that I was using the same cooler, it's not surprising. What is surprising is the wild differences reported by EasyTune 6 while the Phenom was installed!

The idle temp I noted with my meter on the 6400+ was identical to the EasyTune number, and the load number was fairly close, with a delta of 5 degrees. I would hazard a guess that the EasyTune numbers are right on the money with this CPU.

The results reported by EasyTune 6 with the Phenom are much higher, even though I recorded identical numbers loaded and idle with my meter. The idle temperature was overstated by EasyTune 6 by 9 degrees, and the loaded by 16. Now, if we presume that the 5 degree difference between the multimeter temp and the EasyTune 6 temp would be consistent with that of the Phenom. then the loaded reading in ET6 looks to be overstating the actual cpu(case) temperature by 11 degrees.

It does seem as if EasyTune 6 (which matches the temp reported in the BIOS at bootup and the temp in Speedfan) is reading high by about ten degrees. It may not be exactly accurate to subtract ten degrees from the reported temp, but doing so would (by these numbers) get me within 1 degree of the presumed accurate temperature.


Source (http://74.125.155.132/search?q=cache:PIAHChXXh44J:www.alcpu.com/forums/viewtopic.php%3Fp%3D2844%26sid%3D7bf2b2546a3e2ca32 ed516404f8454c2+easytune+6+ma790x-ud4p+temps+reporting+high&cd=4&hl=en&ct=clnk&gl=us)
The source is a google cached page because the actual link was broken due to a server being down. If you want the original source, google the following:
easytune 6 ma790x-ud4p temps reporting high

it'll be the fourth hit.

[ramrodcar]

The source is a google cached page because the actual link was broken due to a server being down. If you want the original source, google the following:
easytune 6 ma790x-ud4p temps reporting high

it'll be the fourth hit.
That's very interesting....