I've Been Trying To Find Out Acceptable Tempurature Range Is For Overclocking A Cpu.

I have a 4th gen Xeon E3 1271 V3 cpu that runs at 3.6GHz with a turbo of one core at 4.0GHz. I'm seeing if I can overclock it to 3.7GHz and I'm using Prime95 to stress it and it keeps all cores at 100% usage. The CPU reaches a max temp of 79 celsius. From there it seems to fluctuate from 73c up to 79c.
I've read that 75 - 80c is considered a good temp range when overclocking, but I've also read that it shouldn't go over 70c.
So I"m hoping I could get some help here on just what an acceptable temp range should be.
Thanks

 

Where did you read that? And I sure don't know what they mean by a "good" temp range. You don't have to strive for the coolest temps possible, but a "good" range would be something considerably lower than the maximum allowed.

So IMO, all those temps are too hot - at least for me. The only official Intel document I can find for the 1271 V3 does not list maximum temps but another source says 75°C is the "absolute maximum". But note that source also says,

So they are recommending keeping your max when under load between 60 and 65°C - which is where I would be looking for too.

And note when I look at the ARK for the E3-1230 V2, which is in the same family of processors as your 1271, it clearly shows the maximum Tcase allowed is just 65.8°C. So then the recommendation would be between 50 and 55°C.

And note those are maximums. It generally is not recommended to run any CPU at or near its maximum allowed for extended periods. It is okay for a few seconds, but then it should quickly drop back down to a comfortable temp.

IMO, you need to immediately drop your clocks back down to get your maximum temps closer to 60 - 65° (or below). And then you need to address your case cooling, and then perhaps your CPU cooling before you attempt to increase your clocks again.

Remember, it is the case's responsibility to provide an adequate supply of cool air flowing through the case. The CPU cooler need only toss the CPU's heat into that flow. So if you don't have adequate case cooling first, you need to look at adding more fans, or swapping your existing case fans for better ones.

 

I stopped overclocking the cpu so it's at 3.6 and then ran Prime95 which proceeded to max out all the cores and threads of the cpu.
I'm using a Gigabyte H81M-DS2V motherboard in a Gateway Dx4860 tower case which has very little air flow from the front, so I modded the front with an opening and added a 120mm fan in the front. Then I added a Thermaltake Contact Silent 12 for a cpu cooler.
Even after doing all that and just running the cpu a stock speed it still seems to be running kinda high.
I purchased the cpu off of ebay and I'm beginning to think I got a lemon, though it's not crashing nor is the cpu throttling.

After about 20 minutes the following core temps still seemed kinda high.

Core Curent temp Max Temp
0 65c 69c
1 68c 71c
2 67c 71c
3 62c 64c

Doest this mean it's just a crappy cpu or is this what happens to a cpu when it was abused in a prior life through much overclocking?

 

Neither. It suggest poor case cooling. Are both front fans drawing cool air in? If your CPU was a lemon, it would be crashing or fail to boot at all.

Did you apply a good, thin, but thorough layer of TIM (thermal interface material) between the CPU and heatsink?

 

I have the 120mm fan I placed in the front of the case ( I actually had to use tin snipes to create a big enough opening for it so there was no obstructing air flow. I have yet to place the actual plastic cover on the front of the case as I'll have to make a hole in that too and I got a large 140mm filter cover to place over the hole I'll be creating in the case and the filter will also provide some dust from getting through.

Then the Thermaltake cpu cooler also has a 120mm fan on it blowing out the back of the case (which I also made more openings via drill holes in back of the case to allow the Thermaltake 120mm fan to provide airflow out the back as the back of the case only had vent holes to fit a 92mm fan.

I used MX-4 thermal compound and use a razor blade to spread and thin it out over the cpu. A method I've had good luck with before so I think I did a good job - at least I hope I did
I checked to make sure I had the fan attached right so it was blowing through the heatsink fins and out the back ( a mistake I did once beforeand fortunately caught it before any problems occured ).

My power supply brings in air from the back and blows it down into the case and pretty much right onto the fins of the thermaltake heatsink tower ( there's only about a 3/4" space between the psu grill and heatsink ). I'm wondering if this would hinder the cpu heatsink from dissipating the cpu heat and maybe by adding a case fan on the back would help.

Right now the Thermaltake heatsink and 120mm fan for the cpu leaves a clearance of about 2 inches from the back of the case and I thought that because it was so close that I wouldn't need a case fan out the back, but with the psu adding it's heat to the case and right on the heatsink tower should I add a case fan on the back? There is enough room for me to attach a case fan to the back grill which is set to take either an 80mm or 92mm fan size. Even though I drilled a lot of holes in the back of the case around the grill holes already there ( about a 4" X 4" size area ) there isn't enough room for a 120mm fan.
If you think it would help with the air flow I can attach a 92mm fan to the back case grill, but I also have an 80mm high speed cooling fan 1000-4500 rpm with an airflow of 46 cfm at 34db - do you think it would be a better choice?

 

That does not answer the question. The question was, "Are both front fans drawing cool air in?"

I don't recommend the use of a razor blade because it would be easy to scratch a Grand Canyon size trench (microscopically speaking) in the metal of the CPU die. If you were very careful to avoid that, you are probably okay. In the future, a credit card works fine, or I snip the end off a plastic shaft cotton swab, then bend it about 1/2 inch in to form a little hockey stick shaped applicator/spreader. Then spread a small dab of TIM like spreading icing on a cake.

Why would it be doing that? That would not be normal as every PSU I have worked with exhausted its own heat directly out the back. You would have to mount it backwards some how to have a PSU pump its heat back into the case. If top mounted, a PSU should actually help exhaust heated air out the back from inside the case. If bottom mounted, they will either pull air from inside the case and push it out the back, or they will pull cool (and hopefully filtered) air from a vent under the PSU then push the heated air out the back.

You generally want a little over pressure with your air flow. That is, you want fans to be pulling a little more air in than pushing heated air out - though I quickly add that is primarily with filtered cases. With a little over pressure, cooled air is pulled in from the vents (and hopefully through dust filters). With under pressure (which creates a slight vacuum inside the case) unfiltered air gets sucked in through every crack and crevice - including through optical drives, USB ports and other places you really don't want dust to get into.

I also am not in favor is drilling holes in cases. I've done it. In fact, before case makers got smart and starting designing cases with lots of fan options, I even drilled 92mm and 120mm holes in the tops of cases for "blowhole" exhaust fans. The problem with drilling holes is it is impossible to ensure no, not even one metal filing does not fall inside the case and possibly onto the motherboard or elsewhere where it might short out a circuit. So to be safe, all the components need to be removed, and then the case thoroughly cleaned before putting it all back together. That's a PITA.

Frankly, if just adding new fans (and insuring the front fans are pulling in and rear fans are exhausting out) does not provide sufficient air flow through the case, I recommend you shop around for a new case. One with lots of large (preferably 140mm or larger) fan support and removable, washable air filters. I like Fractal Design.

In the meantime, blasting a desk fan into the open side of the case should keep your components cool enough until the budget allows for a new case.

 

I did remove everything from the case before drilling and even taped in a sheet of plastic to control where the shavings go for easy vacuuming. I've had the cover off before when running Prime95 so this time I added an 8" desk fan angled to blow on the cpu heatsink tower/fan so it goes towards the back. I ran Prime95 for about an hour and even then Prime95 registered high temps as:
Core 0 65c
Core 1 68c
Core 2 67c
Core 3 62c

Is it possible that this particular cpu simply runs at these temps all the time when under full load?

 

It is possible but it is not normal for CPUs to run at full load all the time. That typically only happens under simulated scenarios, like when running benchmark programs like Prime.

Still, with those temps registered as the peak temps when running Prime, I would call them good!

 

Okay thanks, but that was with a desk fan blowing on the cpu heatsink tower. So I figured the case could use another input fan and made it so I could attach one to the side cover the of the case. I then closed up the computer case and ran prime again. The temps were only one degree higher then when I ran prime with the side off and the desk fan on it. So I guess I'm good to go. Thanks for your help.

 

I am glad that worked. For the record, I am not always a fan (no pun intended) of side panel fans. Typically, you want a front-to-back "flow" of air through a case. We have found in some cases, a side panel fan actually disrupts that flow with turbulence and resulted in temps increasing over those achieved with just front and rear fans.

The notable exception was when the side panel fan fired into a tube that then channeled its air flow directly onto the CPU (or GPU - depending on fan placement in side panel) without interfering with the main flow through the case.

Another problem we found with side panel fans is fan noise due to fan vibrations reverberating through and causing more vibrating noise in the large side panels. Front and rear panels, being much narrower and securely "fixed" to the main frame of the case are ridge and tend to suppress fan vibrations.

That said, in some cases, cooling improved and noise was not a problem so it is definitely a case-by-case basis to see how it works in each scenario. I am glad it works in yours.

Thanks for the followup message.