So I just got a 2500k and I am wondering what to expect for temps? I am going to use my h50 with 2 fans push/pull ( the stock fan and Antec http://www.newegg.com/Product/Product.aspx?Item=N82E16835209011). It will be in a CM690 with one 120 in the front and two 120's in the top. I plan (hope) to run in the 4.6-4.8 range.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
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.
2500k temps?
- Thread starter trick0502
- Start date
Welp, can't give you a direct comparison, but can get you in the ballpark.
Lian Li PC-9F with 2 front 120mm fans (1 stock, 1 Scythe S-Flex rev. F), 1 rear case fan (Scythe S-Flex rev. F) and 1 stock TY-140 fan on my Thermalright Archon yields idle temps at around 30C (~26-28C on colder cores, 30-33C on hotter cores, measured with RealTemp). Load temps for gaming tend to hover between 40-45C on all cores. Stress testing will push that up into the low 50s, with particular test loads (e.g. Prime95 SmallFFT) reaching nearly 60C. TIM is Thermalright CF3, all fans currently connected to the motherboard and left in auto. CPU is currently at 4.5GHz, with load voltage up to 1.36V.
(Edit: Hitting 4.5GHz stably on P8P67 Pro required nothing more than setting LLC to Ultra High and turbo multiplier to 45x, so voltage will vary from ~.9V at idle to anywhere between 1.15V and 1.36V under varying load conditions)
H50 temps (even with 2 fans push/pull) tend to run a few degrees hotter. I worry a little about the airflow in your case as well, as it's strongly bottom-to-top rather than front-to-back, which is unlikely to do your rad any favors with where it will be mounted.
Edit2: So the OCP side of [H] stopped misbehaving finally, and could piece together some numbers based on Marc's tests (to keep it in the family, so to speak). At stock clocks on an i7-920, he showed temps of 27.8C (idle) and 36.2C (load) for the Archon. Overclocked, 30.9C/47.2C. For the H50, 29.2C/40.3C idle, and 32.9C/52.9C overclocked. Between his apples-to-apples test results, and my real-world numbers, you can probably make a fair guesstimate of what you'll see.
Lian Li PC-9F with 2 front 120mm fans (1 stock, 1 Scythe S-Flex rev. F), 1 rear case fan (Scythe S-Flex rev. F) and 1 stock TY-140 fan on my Thermalright Archon yields idle temps at around 30C (~26-28C on colder cores, 30-33C on hotter cores, measured with RealTemp). Load temps for gaming tend to hover between 40-45C on all cores. Stress testing will push that up into the low 50s, with particular test loads (e.g. Prime95 SmallFFT) reaching nearly 60C. TIM is Thermalright CF3, all fans currently connected to the motherboard and left in auto. CPU is currently at 4.5GHz, with load voltage up to 1.36V.
(Edit: Hitting 4.5GHz stably on P8P67 Pro required nothing more than setting LLC to Ultra High and turbo multiplier to 45x, so voltage will vary from ~.9V at idle to anywhere between 1.15V and 1.36V under varying load conditions)
H50 temps (even with 2 fans push/pull) tend to run a few degrees hotter. I worry a little about the airflow in your case as well, as it's strongly bottom-to-top rather than front-to-back, which is unlikely to do your rad any favors with where it will be mounted.
Edit2: So the OCP side of [H] stopped misbehaving finally, and could piece together some numbers based on Marc's tests (to keep it in the family, so to speak). At stock clocks on an i7-920, he showed temps of 27.8C (idle) and 36.2C (load) for the Archon. Overclocked, 30.9C/47.2C. For the H50, 29.2C/40.3C idle, and 32.9C/52.9C overclocked. Between his apples-to-apples test results, and my real-world numbers, you can probably make a fair guesstimate of what you'll see.
Last edited:
- Joined
- Dec 3, 2006
- Messages
- 2,129
Another indirect comparison with a h60
i5 2500k 4.5ghz 1.310v
I have an Antec p182 with 1 front 120mm fan (stock) and the h60 is mounted on the rear fan in push pull formation pushing air from the case out and a top fan being an exhaust.
On idle surfing the web I get around 28-30C with coretemp.
Load temp with gaming (I was playing black ops) around 40-45C all cores
Prime95 after 3 hours around 60-61C
i5 2500k 4.5ghz 1.310v
I have an Antec p182 with 1 front 120mm fan (stock) and the h60 is mounted on the rear fan in push pull formation pushing air from the case out and a top fan being an exhaust.
On idle surfing the web I get around 28-30C with coretemp.
Load temp with gaming (I was playing black ops) around 40-45C all cores
Prime95 after 3 hours around 60-61C
A little jealous that you can be Prime95 stable on 1.31V.
When I first tested 4.5GHz, I tried leaving LLC at high, which would yield around 1.32V at full synthetic load testing. Didn't have bad enough errors to BSOD, but did get rounding errors on SmallFFT.
Also, for the OP, you've got a nearly 100% chance of getting a 2500K that'll hit 4.5GHz without issues. Probably a little better than 50/50 of getting 4.6-4.7GHz. If you go take a look at the P8P67 overclocking thread in the Intel Mobo forum, ASUS' internal testing of 2600Ks showed a pretty steep drop off between chips able to sustain 4.6-4.7GHz and chips able to sustain 4.8GHz+. Sample size is only around 100 chips, and that's 2600K not 2500K, so certainly not definitive results, but some food for thought regardless.
When I first tested 4.5GHz, I tried leaving LLC at high, which would yield around 1.32V at full synthetic load testing. Didn't have bad enough errors to BSOD, but did get rounding errors on SmallFFT.
Also, for the OP, you've got a nearly 100% chance of getting a 2500K that'll hit 4.5GHz without issues. Probably a little better than 50/50 of getting 4.6-4.7GHz. If you go take a look at the P8P67 overclocking thread in the Intel Mobo forum, ASUS' internal testing of 2600Ks showed a pretty steep drop off between chips able to sustain 4.6-4.7GHz and chips able to sustain 4.8GHz+. Sample size is only around 100 chips, and that's 2600K not 2500K, so certainly not definitive results, but some food for thought regardless.
i'll take 4.5. that is a big improvement from my 07 q6600.
N
nösferatu
Guest
Hmm....lot's of people can keep their i5 2500k at such a low temp, anyway, here's mine(proof) :
ambient is at 27.5 - 28 C, rest of the rig is in sig.
ambient is at 27.5 - 28 C, rest of the rig is in sig.
Last edited:
Asked on another thread but didn't get the exact answer i was looking for.
I have a 2500k on a Asus P8P67 deluxe board with a 9900max zalman cooler. Good cooling and plenty of airflow in the case.
Ran great at stock speeds and using the AI suite 2 autotune i got it to 4356 at the fast OC setting. (Been running it at this speed last 3-4 days doing steam downloads and such)
Testing it on the extreme for a bit i had it go to 4944 and probably could get it to go past 5000.
What i wanted to know is that with using the autotune, should it be safe to run it at the extreme settings as long as it's cooling well?
Have not done much OC'ing so this is why i'm asking. Pretty sure it should be okay but would rather have info from people who know what they are doing ....
I have a 2500k on a Asus P8P67 deluxe board with a 9900max zalman cooler. Good cooling and plenty of airflow in the case.
Ran great at stock speeds and using the AI suite 2 autotune i got it to 4356 at the fast OC setting. (Been running it at this speed last 3-4 days doing steam downloads and such)
Testing it on the extreme for a bit i had it go to 4944 and probably could get it to go past 5000.
What i wanted to know is that with using the autotune, should it be safe to run it at the extreme settings as long as it's cooling well?
Have not done much OC'ing so this is why i'm asking. Pretty sure it should be okay but would rather have info from people who know what they are doing ....
stock idle is around 89f for you Americans out there.
Full load with stock cooler stock speed is 130-139f with stock cooler.
I dont do the C even though I can all day long. F is a more fine tuned with more temperature intervals in between. It has a higher resolution.
My mini-ITX is 2500K with stock cooler and its for sale .
Full load with stock cooler stock speed is 130-139f with stock cooler.
I dont do the C even though I can all day long. F is a more fine tuned with more temperature intervals in between. It has a higher resolution.
My mini-ITX is 2500K with stock cooler and its for sale .
I am getting those kind of temps with the stock Intel cooler w/ stock TIM @ stock clocks and my case in a desk enclosure with Q-Fan enabled ("normal" on CPU, "turbo" on case). Ambient is around 22C but higher in the enclosure. IMO, you should be doing better than that with the cooler and fan settings you have. My prime temps might be in the mid-high 50s, but the point is still valid. Maybe I have the world's best stock cooler, or Canadian 22C air cools better than everywhere else. lol
seems like you were right. just reran some tests. 67 in the house right now. temps are mid to upper 40's.
I was asking myself questions about that proc temps also.
I've just bought one, and mounted it in a FT02 case with Megahalem rev B, push pull config.
It's overclocked at 4.5 stable, 1.33 Vcore.
Ambiant temp is 25 / 26 °C.
At idle, the proc gets up to 30 - 35 °C.
Under load (Crysis, 1200p, very high settings), it reached 55 °C.
It seems a bit high to me, but maybe it's perfectly ok. What do you think ?
I've just bought one, and mounted it in a FT02 case with Megahalem rev B, push pull config.
It's overclocked at 4.5 stable, 1.33 Vcore.
Ambiant temp is 25 / 26 °C.
At idle, the proc gets up to 30 - 35 °C.
Under load (Crysis, 1200p, very high settings), it reached 55 °C.
It seems a bit high to me, but maybe it's perfectly ok. What do you think ?
paperwastage
[H]ard|Gawd
- Joined
- Jul 12, 2009
- Messages
- 1,131
?
Any temperature has infinite measurement/intervals/resolution... it can be 25.64898623165649865 C or 96.465323165 F...
It just depends on how the chip's sensor works... and since Intel reports everything (Tjunction) in Celsius, most likely the sensor reports in celsius, use Celsius?
@bbross, 55C seems fine at 4.5ghz 1.33 vcore....
my 212+ at IntelBurnTest goes to 60C at 4.5ghz ~1.32vcore
Last edited:
65 or less is fine for 24/7 use?
F or C whats the difference. Its all math.
Anyways 60c at full load stock speed is perfectly fine. Just worry if you are going over that at stock clock speed. That might mean your HSF is crap, fans bad, or goop is not applied right. H60 should cool on par with most average aftermarket coolers. You can easily push 4.6 with H60. So many people here do it.
Remember that you should be using offset voltage as well so you are not just creating a ton of wasted heat and power bill running the cores at 1.XX oc voltage continuously.
Anyways 60c at full load stock speed is perfectly fine. Just worry if you are going over that at stock clock speed. That might mean your HSF is crap, fans bad, or goop is not applied right. H60 should cool on par with most average aftermarket coolers. You can easily push 4.6 with H60. So many people here do it.
Remember that you should be using offset voltage as well so you are not just creating a ton of wasted heat and power bill running the cores at 1.XX oc voltage continuously.
- Joined
- Dec 3, 2006
- Messages
- 2,129
I got mine up to 5ghz but stable only at 1.48v. I don't think the extra voltage is worth the extra 500 mhz
Last edited:
N
nösferatu
Guest
You should try small FFTs imo to see the real 100% load max temp, not blend. Blend test does not heat up your CPU as fast as small FFTs especially when you've only run it for 7 mins.
Last edited:
N
nösferatu
Guest
I dont think so, there are numbers such as 50.5 C, 50.75 C etc etc....
The point is, 50 C is still cooler than 124 F and 140 F is still cooler than 65 C...
Offset is mostly garbage if you're overclocking, you tend to need more voltage to compensate for the fact that you're constantly fluctuating power delivery in conjunction with different clock speeds, ending up with a less than stable system at what would otherwise be a lower manual constant voltage.
Thing is, without a load on the chip, 1.4v is not doing much to the chip because there's hardly any current going through it. Other than some minuscule power savings, there's really no reason to use offset mode.
Well this might be true and might be grounds for an entire thread in it's self.
I do remember that my temps were actually the same with or with out offset. Im pretty stable at offset so im not worried. I do understand what current is but you do not think that running 1.4+ 24/7 will generate more heat overall even at idle as well as not wear out the transistors faster?
I saw no difference in idle temps between manual and offset, with offset giving me a ~1v idle and manual a constant 1.4v. There are many threads regarding manual vs offset on other forums with most people coming to the same conclusion; manual is more stable at a generally lower voltage. You could even argue that manual is safer due to the lower chance of a BSOD at an inopportune moment, possibly corrupting files.
Just got my i5 rig going and OC'd to 4.5 at 1.28/5v.
Temps are around the low 50s in Prime 95, maxing out at 60 (but staying at low 50s 90% of the time)
How is that, for an FT02 case with a H212+ with a push/pull setups and shrouds?
I did mess up the thermal paste though, I think (first time) so I'll probably reseat that if temps don't drop in a week.
Temps are around the low 50s in Prime 95, maxing out at 60 (but staying at low 50s 90% of the time)
How is that, for an FT02 case with a H212+ with a push/pull setups and shrouds?
I did mess up the thermal paste though, I think (first time) so I'll probably reseat that if temps don't drop in a week.
Those are great temps, and really low voltage for 4.5. I get similar temps at 4.4GHz and 1.32V with a CoolIT liquid cooler (240mm rad), although it's stuffed in a small mATX case with the fans in a push config with dust filters all around.
I would recommend everyone try doing custom FFTs starting at 1280K though, as that's where instability shows up the earliest for me (if you run Prime95 blend, those tests don't run until 8+ hours in).
I would recommend everyone try doing custom FFTs starting at 1280K though, as that's where instability shows up the earliest for me (if you run Prime95 blend, those tests don't run until 8+ hours in).
okay scratch that haha. After about 4 hours core 3 failed, so I upped the voltage one more bit and am re-testing (seems to be doing fine on your FFT 1280Ks).
In the bios It's set for 1.270 and CPU-Z reports 1.288V under prime load. Not sure which to believe but yeah around there. I hear most people have to crank it to 3 or higher to get 4.5 so it's all good.
Temps are still low 50s and the like, and in a week or so I'll try reseating the heatsink with AS5 and see how that goes. If temps climb I still have a large amount of ICD7 so I can reapply that.
This is the first time I've ever done a computer build so I can be paranoid about things I shouldn't be.
In the bios It's set for 1.270 and CPU-Z reports 1.288V under prime load. Not sure which to believe but yeah around there. I hear most people have to crank it to 3 or higher to get 4.5 so it's all good.
Temps are still low 50s and the like, and in a week or so I'll try reseating the heatsink with AS5 and see how that goes. If temps climb I still have a large amount of ICD7 so I can reapply that.
This is the first time I've ever done a computer build so I can be paranoid about things I shouldn't be.
I have been running prime 95 for a little under an hour and my highest core temperature is 54 degrees Celcius.
I am running 100% stock right now with a Coolermaster Hyper 212 Plus single fan heat sink.
I am going to let it go for another hour and see if temps hold steady. Then its time to try a little over clocking.
edit: got impatient. Highest it got was 54 degrees C.
As soon as I stopped prime it dropped right back to 36C
I am running 100% stock right now with a Coolermaster Hyper 212 Plus single fan heat sink.
I am going to let it go for another hour and see if temps hold steady. Then its time to try a little over clocking.
edit: got impatient. Highest it got was 54 degrees C.
As soon as I stopped prime it dropped right back to 36C
Last edited:
Sounds roughly the same I got using that setup. OCing to 4.5 using aorund 1.280v brought my max temps to around low 60s.
N
nösferatu
Guest
I did 4.5 Ghz with 1.28v and small fft ran fine for 6-7 hours but playing BF2BC would freeze/lock the pc in just around 2 hours, fyi.
-_-
nösferatu;1037281631 said:
you need to run blend or custom test starting at 1280K (my suggestion). small FFT's are not likely to show instability on SB
My temps are around there too.
Disagree. I am using a offset voltage and so far so good. Sure I need 1.38V for 4.5Ghz, and only 1.33V with fixed voltage. But who cares, the system is idle most of the time, so it uses around 1.0V.