OC'd 5960X temps

I reran Aida 64 Extreme Stress test for @45 minutes. Ambient air temp was 75F or 23.88C. Water temp per my XSPC probe with LED water readout was 27.4C

The stress testing cpu/fpu/cache/system mem temps are min/max/average

MB 28/30/28.7
CPU Overall 28/55/52.4
Core#1 29/69/56.1
2 30/65/55
3 30/73/61.4
4 29/65/56.4
5 33/71/59
6 30/67/57.1
7 30/71/57.8
8 29/60/51.4

I hope with the ambient air temp AND the water temp for the loop it aids Earthdog.


CPU core Voltage min 1.312 max 1.328

Clock
cpu 4400
nb 3000

I'm slowly bumping voltage up to get a 4.3 stable. Currently trying 1.315.

At idle, CPU is about 32. Ambient is 23.8C, case temp is about 26C (bottom half of case). Water in is 27C and out is about 25.5. This is with everything at its lowest/quietest. Only 50L/Hr and fans just under 600 RPM.

So running at the above voltage at 4.3 bumps up core peaks to 80C. Overall of 67C. Ambient remaining the same and water in/out is 26.8 and 25.1. VRMs are nice at about 53C.

Seems I'm near the limit I want to bring this thing to. Guess I didn't win the lottery on a good chip. Oh well. Batch 3423B529 for inquiring minds.

You need 1-1.5GPM for most efficient heat transfer...though your 1-2c difference leads me to believe that it's is ok.

I thought we covered that above. That's where I am now.

Perhaps my math was wrong....

But 50L/hr is around 12GPH. That isn't remotely close to 1GPM....?

That's just at idle and for low level things to keep it as quiet as possible. Nothing under load runs at that rate.

Railgun, I feel the same way about the "silicon" lottery though I can OC to 4.4 Ghz. Oh I can go into Windows 8.1 at 4.5 with ease but need in excess of 1.36v to keep stable when running serious benchmarks like ROG RealBench. I really cringe when I see reviews of the 5960x where the reviewer is using an AIO and OCs to 4.6Ghz and says it's stable after running Aida for 15 minutes. Horse Hockey!

Asus ROG developed their RealBenchmark test to put the cpu through the ringer. I'm still in the infancy stage of memory tweaking (have 4 chips of Corsair LPX DDR4 listed at 2800) I'm running them at auto (2133) to keep the cpu stable at 4.4Ghz. My vcore is manually set to 1.31 but climbs to 1.328 after hrs of ROG benching. I find this acceptable and low enough for running 24/7.

I now have the equipment (mb and cpu and cooling capacity ) to start some serious tweaking BUT I refuse to jack up the speed just for bragging rights. I'm sure you are the same. 4.3Ghz is still DARN good super stable for an 8 core/16 thread beast that at stock is 3Ghz!

BTW, did you buy the Intel Protection plan? I did for this and have it for my 3930k and 3770k. Cheap insurance ($25 to $35 depending on chip).

Good luck and great info.

Railgun said:

I'm slowly bumping voltage up to get a 4.3 stable. Currently trying 1.315.

At idle, CPU is about 32. Ambient is 23.8C, case temp is about 26C (bottom half of case). Water in is 27C and out is about 25.5. This is with everything at its lowest/quietest. Only 50L/Hr and fans just under 600 RPM.

So running at the above voltage at 4.3 bumps up core peaks to 80C. Overall of 67C. Ambient remaining the same and water in/out is 26.8 and 25.1. VRMs are nice at about 53C.

Seems I'm near the limit I want to bring this thing to. Guess I didn't win the lottery on a good chip. Oh well. Batch 3423B529 for inquiring minds.

Click to expand...

I also got confused as well as Earth pointed out about your flow. So that is your idle flow readings and your water flow gets bumped when temps rise I assume and your water temps in/out, now is that on idle or full load? Scratching my head a bit here when I see your load temps fluctuate so much and having such low water temps if say, this was on full bore. At first I thought its possible there wasn't enough flow being pushed with that much raddage but now I am thinking its possible your seating might be an issue since I feel, knowing that the 5960x do run hotter than their 6 core sibling but I heard getting at 4.4 GHz is no issue. Its when making the jump to 4.6Ghz is when its wants some more juice.

Usually the temps fluctuate per core within 10c-5c of each other. Maybe I am reading this wrong but to have more than a 10c difference from your average to your highest core, I would look at reseating and reapplying the TIM on the CPU block in case of bubbles and or bad contact. Can I ask what TIM you're using? I think its irrelevant if you're using a reputed company regardless.

In general, I think you should be getting better temps on that CPU in a system like that @ that speed and voltage but I could be wrong.

Nonetheless, nice rig OP.

Railgun, can you run Aida 64 Extreme for @1/2 Hr under stability test checking cpu etc ( don't need storage or gpu) and see what you are showing for min/max average?

Sorry about that. Yeah, I could have been a bit more clear here.

We'll leave idle out of it since that's irrelevant for the most part. At load, when cpu or GPU hits a specified temp, my pumps ramp up to about 200L/Hr, which is somewhere where it needs to be as discussed above.

The overall cpu temp is whatever the non-individual core temp is. 80 is what any individual core reports. I'll stop reporting on anything other than individual core temps. That said, discount the 67C.

I'm using CLU. Though it was applied and seated and left for at least three weeks if not more before it was ever turned on. Whether that has anything to do with anything, I don't know. First time I'm using this TIM. It's possible I didn't prepare it well.

I need to drain and replace one of the tubes up top now anyway as its blocking a fan, so I'll have an opportunity to remove the block and clean and reapply. Just need to get some more tubing. Being a noob with acrylic, I went through three orders of this stuff. Most of it wasted.


Gkline, I'll run that in just a few.

240L /hr is 1 GPM.

That stability test is horrible. It needs to be rewritten a bit. I don't think it taxes the CPU all that well...max temp was 66, next highest core was 62. Lowest max core was 54. Plus, it ate 28+GB of RAM. There's no need for that. I have my page file disabled as there's no need for it with this much memory, and wouldn't have expected it to eat that much. Seems like a horribly written little test.

EarthDog said:

240L /hr is 1 GPM.

Click to expand...

It's been that kind of a weekend. My bad...it's set for 260 L/Hr. I thought I had it set to 200, but it's actually set to 260. That's 1.15GPM.

Running FPU as the only cpu test should give it a hurting.

See, this is what happens when you don't concentrate on what you're doing and try to multitask too many things.

So doing that now...NOT running the memory test (duh) and I'll report back shortly.

Ok...ambient here is 21C. Water is about 24.5C after 60m of running.

CPU min/max/avg over 60 min of running...

CPU 29/68/63.5
Core 1 32/74/66.5
Core 2 34/72/67.3
Core 3 32/78/73.5
Core 4 32/73/67.4
Core 5 33/79/74.0
Core 6 32/72/67.3
Core 7 33/76/70.7
Core 8 32/68/64.6

Currently at 1.328V

Railgun, on average you seem to be @ 5C higher per core. I bet the cpu bl;ock and/or the TIM is the culprit. The change in temp of the loop for both of us rose @3.5C. I'm betting on the TIM and cpu block seating. I am using the newest EK Supremacy EVO block which some reviews says accounts for @ 2C difference.

Gk.... He's run a different test than you now, different rads and raddage, different ambient (but can account for that), etc. Can't compare between two systems!

Anyway, things seem about in order to me. Reseat the block and replace the Tim, but, seems about right to me honestly.

Perhaps I'm confused EarthDog, but I thought he also ran Aida64 Extreme Stability test?

If you only run the FPU test, like I believe he did the second time, it's a different test than with having the others checked. Notice his difference in temperatures from the first time?

Yes, I see it now.

Railgun said:

Ok...ambient here is 21C. Water is about 24.5C after 60m of running.

CPU min/max/avg over 60 min of running...

CPU 29/68/63.5
Core 1 32/74/66.5
Core 2 34/72/67.3
Core 3 32/78/73.5
Core 4 32/73/67.4
Core 5 33/79/74.0
Core 6 32/72/67.3
Core 7 33/76/70.7
Core 8 32/68/64.6

Currently at 1.328V

Click to expand...

Wow, great Delta-Temps.

You might shed a few degrees if you reseat and replaced the TIM with MX-4, for example but its not a guarantee.