misplacement of the cooler, intel has a horrible hooks, whether these are misplaced one simple tap to get out and generates this temperature.
Another problem is the motherboard, misplaced or slightly “bent”.
“lack of power source Psu” is very rare can happen.
Problems manufactures the micro is very rare can happen.
Also try to test whether bios sensors make the same or similar. If so have some kind of problem in previous years dimensioned points.
Esa temperatura de CPU pueden ser 4 problemas.
mala colocación del cooler, intel tiene unos enganches horribles, si estos estan mal colocados un con simple golpecito se salen y genera esa temperatura.
otro problema es la placa madre ,mal colocada o ligeramente “doblada”.
“falta de potencia de fuente Psu” es algo muy raro puede pasar.
Problemas de fabrica del micro es algo muy raro puede pasar.
Igualmente intenta de probar si en bios los sensores marcan lo mismo o similar. Si es asi tienes algun tipo de problema mensionado en os puntos anteriores.
I looked up your cooler master centurion 6 atx
And it comes by default only with one fan.
Now I do not know much about that case - I have a 2008 MacPro and with simple fans and well engineered
airflow you can keep the CPUs under full load below 70 degree - no need for expensive water cooling.
So did you look into adding more fans? You can have up to 9.
I don’t much about cpu temperatures or overclocking, infact, I’ve only had computers for 31/2 years, in which time I quickly became a Blender fanatic,and I have learned a lot about computers,but there is so much more to learn. Anywho, I’ve just performed a 5 minute stress test on my cpu with the intel extreme utility overclocking software, and for the whole 5 mins. the cpu’s core temp. was hovering between 100c and 103c , and when the stress test was completed, the utility said that the cpu passed the test… So if there is someone out there who knows about overclocking, I would like to know are these kinds of temperatures normal for an overclocked cpu? As I said in a prior post of this thread, the support team at Origin computers who manufactured my laptop, set my overclocking lower than what is was when I purchased it because I was having troubles while running Blender, at that time, which was 2 1/2 years ago, I knew next to nothing about computers and overclocking, all I knew when I purchased my computer was I needed something powerful to do what I wanted on blender, as after a year I burnt out a HP laptop (my first ever computer), but I’m sure the support team who set the overclocking must have run the stress test and seen the temperature it was running. But what I’m understanding from this thread is that these kind of temperatures are dangerous! These temps, of course are occurring when the cpu usage is at 100%, which is what it is when blender is rendering! All replies will be greatly appreciated, as I am concerned about my cpu temperatures. The mother board is having the same range of temperatures as well.
Generally you don’t want your CPU to be at >80c for extended periods of time (ie rendering 24/7 etc…) however even pushing it a bit over should still have the CPU running fine for longer than its worth keeping.
Having it throttling at 100c however isn’t ideal at all.
The stress test that XTU have isn’t as good as Intel Burn Test, however Intel Burn Test makes your CPU far hotter than a normal stress test, or rendering does. I have had my PC crash in an unstable overclock with Cycles, and pass the XTU test.
I run my overclock at 4.8GHz on the i7 4790k, and it gets to about 70c in the stress test, and about 73c in Cycles. Running something like Intel Burn Test gets it to about 93*c, which is fine to test stability.
@zeealpal Thanks for your reply. I have just downloaded speccy, which gives temp. readings of the different components, and individual core temperatures, and it seems its only two cores that exceed 100c. with an average of the 4 cores being 97c during those peak temps, the temp. being giving by XTU is the highest temp. of any of the 4 cores. But during a 2 1/2 minute render, that high temperature is sustained for about a minute of the last half of the render, and when the picture is being build on to the display. And yes, its throttling at 100*c+. I am sometimes rendering for like 5 hours at time, as I am working on an animation project, with scenes exceeding 1200 frames or more, and somtimes lots of vertices, the present scene I am working on has like 17838130 verts. and particles and stuff, and I am able to render a frame in like 2 1/2 minutes. I have done some breaking things up and compositing with a background with alpha overs where I can. Anyways I have contacted my support team at Origin computers and explained my concerns to them, and am waiting for a reply. Thanks again for your reply!
@the animator If the screenshot of the XTU is correct (i personally don´t know or use the program so i could be wrong) - the Support team has set the CPU back to default values (up to 3.8Ghz on a single core, shown by the multiplier 38x in the upper right corner) and not to 4.4 Ghz. The CPU should work fine and don´t get to hot without the overclocking - considering they even sold it with 4.5 Ghz and it should do fine even with that frequency… Anyways, as the CPU throttles down to 2.8 Ghz there is something wrong.
As it is a now 3 year old notebook you could try to use a vacuum cleaner to get the dust out of it. Is there enough space around the air intakes/escapes while you use the machine?
If you checked these two things, you could take a look if there are any settings that control the voltage of the CPU within the XTU. In your Screenshot i can only see an additional turbo voltage which is already set to 0.0 mV. Maybe it runs with to much voltage which could be a reason if the CPU gets to hot - as it came overclocked this could be a possibility. Even if a CPU runs with factory default voltage it is often able to run with lower values - basically every pc i´ve build is undervolted a little bit (if it is not overclocked of course:-)
Sometimes it’s not the fan speeds, but the cooling capacity of the heat sink on your chips.
In my case the system was stable for everything except some particle sims and the smoke sim in Blender. (No problems with games or some other rendering in Blender that took hours.) Only those two particular things would trigger the motherboard’s thermal protection shutdown.
Turns out the stock cooler that came with the CPU wasn’t anywhere near enough for the application. (I didn’t ever overclock, but apparently running at 100% on all cores for an extended duration wasn’t something the original package anticipated.) Went and got a Hyper212EVO heatsink. Fan speed and power load is about the same, however cooling capacity is about 3X in terms of heat being moved (surface area is that much bigger), thus no more crashes.
I’d say a lot of manufacturers don’t anticipate the computer equivalent of running at full throttle for extended periods. They may include some nice maximum specs, but the intended level of continous operation is about 70%. For rendering there typically will be some components that need to be upgraded vs. the stock configuration.
Damn you guys know your computers. But, back to @ezra3131 for a minute. My understanding is a computer needs a minimum of eight (8) inches on all sides for maximum cooling. And, a unobstructed flow of air. Which means many computer desk are designed wrong for our purposes. Also being on the floor, or close to it, will mean picking up more dirt faster then being up high. This I found to be true and moved mine to the desktop. I had mine sitting on a rug which was a guarantee for picking up lint and dirt.
And, you never mentioned an ambient temperature, ezra3131. After changing out the CPU paste mine now plays nice at an ambient of 78F for animated renders. But, really prefers 77F or a little below. Which is still above the optimum temperature for a CPU under full load. Now please keep in mind that statement is based on a cheap digital desktop thermometer. Meaning it might just be a reference for me. However, in the Winter time the ambient can creep up to unacceptable levels. Then again my old ass might have something to do with that. At seventy six my thermostat isn’t exactly up to tolerances.
At those temperatures I would suggest redoing the CPU paste. Honest to god you have to suspect the integrity of the paste and the heat transfer after that. That and a through cleaning along with putting it up high with a unobstructed airflow might just be all you need. Along with watching the ambient temperature. I sincerely hope so.
Much good tech info posted on this thread. I have no idea why I would weight in.
CPU paste: In my working days I once found myself changing out components on PC boards albeit in a shipyard school. Those that have done it know a small soldering iron can be your best friend or worst enemy. Regardless, I went to YouTube looking for tutorials on redoing the paste since it was my only computer up and running. Jesus what a waste land of misinformation that is given the subject. But, I finally found one which didn’t defy common sense. And, was for the configuration I had.
Just have all your materials at hand. And, very carefully break the existing seal if necessary. @ezra3131 I doubt if you even have a seal given those temperatures. And, for all I know this is a evolution you might have done numerous times. If so I apologize but just do it one more time to humor a old man. And, a guy who posted earlier the paste needs to be renewed. Regardless of what caused the original problem your CPU paste might be a useless blob sitting there now. Mine was when I finally got in there. Happy New Year and happy Blendering, guy.
Not so long ago I finally got permission from the store the PC came from to change the CPU cooler. I replaced it with a cooler master hyper 212 EVO and got a 40°C improvement. It makes you wonder why intel even supplied a stock cooler because it was clearly not sufficient to cool the CPU.
With time and dust collection or heat sink thermal paste gets removed.
My rendering with 8 threads of blender was going to 88 degrees when not running AC. I have installed cooler master seidon 120v+
Now temps reach only 73 degrees while rendering in Non AC. In AC it is average 69 degrees. I think I have to connect my seidon fan to molex instead of motherboard and it will cool it further. Plus I have used only one fan, another fan at the outside for push and pull might help.
I think Intel might realize this as you can buy their CPUs without a cooler if you wish. But if you purchased a pre-built computer they will probably slap on whatever is cheaper. I bought a Cooler Master Hyper 212 EVO too for my 5820k CPU which runs normally at 3.3 Gz. I have mildly overclocked it to 3.8Gz, and while rendering in Blender with all cores It doesn’t get above 68C. The cooler is a monster though, so you need a very large case to fit it in.
