CPU temps higher then they should on custom loop

[XVB]

So I had the good fortune of geting a 5950x, yeay, buat I only have a 120mm AIO, didn't cut it, so I ended up ordering a custom loop, it ran fine at first but then it starts to randomly shut down sometimes during idle and faster in heavy work loads also temps look weird, like in Hwinfo the CPU would be runing 80-90 watts but temps are at 63°C and climbing. Although I feel fortunate enough to be able to aford said custom loop and CPU, I am not made of money, so I humbly ask the forum what should I do? Any sugestions are welcome, thanx in advance

 

System info: 

Ryzen 9 5950x 

2x32gb trident z cl18 3600mhz

Asus ROG chrosshair VIII impact

RX 6700xt

Adata SX 8200 1tb as boot drive

Runing windows 10 update to latest

Stock setings, no manual OC

Cooling loop: CPU only

1x120mm ek coolstream XE

1x120mm ek coolstream SE

Barrow pump block resevoir combo

2x120mm noctua industrial ppc 3000 fans

1x120mm ekwb vardar fan

[Chris Pratt]

It looks like you still have a 120mm rad? Custom loop isn't going to do anything much more than an AIO would at the same rad size. It's more about looks than anything else. The issue is the rad size. You need 280mm or larger before you can outperform a good air cooler. 120mm isn't even good enough to cool something like a 5600X properly.

 

 

 

[XVB]

Yes, but you see I have 2 of those 120mm rad, one is 26mm thick and another 60mm thick. Is that stil not enough? Have you had any experience using a 5950x, to tell you the truth this is the first High end CPU I ever bought so I realy do not know what to expect

[XVB]

2 minutes ago, Chris Pratt said:

It looks like you still have a 120mm rad? Custom loop isn't going to do anything much more than an AIO would at the same rad size. It's more about looks than anything else. The issue is the rad size. You need 280mm or larger before you can outperform a good air cooler. 120mm isn't even good enough to cool something like a 5600X properly.

 

 

 

Btw, are you really chris prat? LOL

[Chris Pratt]

Just now, XVB said:

Btw, are you really chris prat? LOL

Yes, but not the actor.

[XVB]

3 minutes ago, Chris Pratt said:

Yes, but not the actor.

Haha, ok, oh and I see you have a 5900, tell me what are your idle temps nd load temps?

[Chris Pratt]

6 minutes ago, XVB said:

Haha, ok, oh and I see you have a 5900, tell me what are your idle temps nd load temps?

I run with a silent profile, so temps are higher than they could be if I accepted more fan noise. It pretty much has to be melting before the fans go more than 50% (not really, but you get my point). Anyways, given that, Im at around 42C idle and 78C under load. I'm also undervolted, for what it's worth, but I have PBO enabled, as well.

[XVB]

Hmmm, interesting those are similar to my temps but I have ramp up my fans, and yours doesn't shut down on 70 C?

[Chris Pratt]

1 minute ago, XVB said:

Hmmm, interesting those are similar to my temps but I have ramp up my fans, and yours doesn't shut down on 70 C?

No? I mean it shouldn't. 70C is nothing. Zen 3 should be able to hit at least 95C before it even throttles. You'd need to exceed like 100-105C before it will outright shutdown to save itself.

[XVB]

Just now, Chris Pratt said:

No? I mean it shouldn't. 70C is nothing. Zen 3 should be able to hit at least 95C before it even throttles. You'd need to exceed like 100-105C before it will outright shutdown to save itself.

Ok, well that is concerning, I might have to RMA something, (which I can't), I might try to switch ram or the motherboard

[TheGlenlivet]

Just now, XVB said:

Ok, well that is concerning, I might have to RMA something, (which I can't), I might try to switch ram or the motherboard

Yeah the temps aren't the issue, at least I don't think so.

The shutdown is.

Are you using an older PSU?

Can you check the Windows logs to see if you can find the cause of the shutdown? 

Run Memtest (or equivalent) to verify it's not the RAM.  72C shouldn't be causing that.

[XVB]

Just now, TheGlenlivet said:

Yeah the temps aren't the issue, at least I don't think so.

The shutdown is.

Are you using an older PSU?

Can you check the Windows logs to see if you can find the cause of the shutdown? 

Run Memtest (or equivalent) to verify it's not the RAM.  72C shouldn't be causing that.

No, I use a newish corsair SF750 platinum connected to a 1300 va APC PSU, darn it if is the RAM.. 

[XVB]

14 minutes ago, Chris Pratt said:

No? I mean it shouldn't. 70C is nothing. Zen 3 should be able to hit at least 95C before it even throttles. You'd need to exceed like 100-105C before it will outright shutdown to save itself.

Ok, one more question please, in this screenshot, does everything look normal to you? Specifically the temps. You see I had to shutdown 8 cores for it to be stable, but say its attached to a 240mm AIO.. is this normal?

[TheGlenlivet]

9 minutes ago, XVB said:

You see I had to shutdown 8 cores for it to be stable

?!?!?!  

Why?  

Your temps are cold.  That is NOT the issue.

[XVB]

9 minutes ago, TheGlenlivet said:

?!?!?!  

Why?  

Your temps are cold.  That is NOT the issue.

Cause if I run all 16 cores, the shut downs were more frequent, at first i though it was overheating.. dang, my head is spining now at the revelation that its geting more likely it was a ram issue.. Ok, now I have to test the ram, I will post my results 

[XVB]

So I tested it with windows memory diagnostic tool and it shut down about 7% in. Twice now.. now going to try memtest86

[XVB]

Memtest86 faild to complete, now removing 1 stick or ram

[XVB]

17 hours ago, TheGlenlivet said:

?!?!?!  

Why?  

Your temps are cold.  That is NOT the issue.

Ok so this happened,  i tried stick no1, no erors, then tried stick no 2 no erors, tried changing slots, no erors. Then I put them both in and suddenly shutsdown after 16 percent. Why is that?

[XVB]

21 hours ago, Chris Pratt said:

No? I mean it shouldn't. 70C is nothing. Zen 3 should be able to hit at least 95C before it even throttles. You'd need to exceed like 100-105C before it will outright shutdown to save itself.

just to let you know, it turns out, there is a problem with the ram, but not the ram sticks them selves, the bios just does not play nice with 64gb of ram. for some reason even though the mobo suports it. but at least my system is now runing stable on 32gb of ram. thqs for the help

[Chris Pratt]

2 hours ago, XVB said:

just to let you know, it turns out, there is a problem with the ram, but not the ram sticks them selves, the bios just does not play nice with 64gb of ram. for some reason even though the mobo suports it. but at least my system is now runing stable on 32gb of ram. thqs for the help

Well, that's still not great. Running single channel will drastically hurt your performance. Capacity is meaningless, other than the amount of DIMMs and/or ranks it takes to make it up. Sometimes running four sticks or 8+ ranks can be difficult for the IMC, leading to instability. However, with only two sticks, this should absolutely not be an issue. What slots did you have them in? It should be 2 and 4 counting from the CPU side.

[XVB]

11 hours ago, Chris Pratt said:

Well, that's still not great. Running single channel will drastically hurt your performance. Capacity is meaningless, other than the amount of DIMMs and/or ranks it takes to make it up. Sometimes running four sticks or 8+ ranks can be difficult for the IMC, leading to instability. However, with only two sticks, this should absolutely not be an issue. What slots did you have them in? It should be 2 and 4 counting from the CPU side.

No no, I used a diffren ram kit, so i'm now runing a 2x16gb ram at 3600mhz cl18 so I'm running 2 channels only lower capacity.. the motherboard only has 2 slots, and the previous 64gb kit is at a friend of mine being tested for potential ram stick issue not detected in the ram test. Vut, tnx for your concerne