• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

AMD recommend minimum of a 280mm AIO liquid cooling for the 3950x. That's the trouble with 7nm it heats up much quicker and in smaller areas on the heatspreader. Don't think an air cooler is gonna remove the heat quickly enough on that monster CPU.

https://www.amd.com/en/processors/3950x-thermal-solutions

Though Noctuas website does recommend their cooler so obviously they have tested so.
 
Minstadave said:
It won't degrade enough to cause that much of an issue. Have you reseated the heatsink?
Click to expand...
Phate said:
Without wanting to condescending, is the heatsink seated 100% correctly? Sounds like it isnt making a good contact. When you take the heatsink off check to see how the thermal paste is spread across the top of the cpu. :)
Click to expand...

I did try re-seating once early when I was having issues (that turned out to be related to the BIOS version ..) and contact looked OK - started with a pea sized amount and the whole surface was covered, screws are all the way down. Maybe I'll try re-seating it once more just in case.
 
Ironic Namesake said:
I did try re-seating once early when I was having issues (that turned out to be related to the BIOS version ..) and contact looked OK - started with a pea sized amount and the whole surface was covered, screws are all the way down. Maybe I'll try re-seating it once more just in case.
Click to expand...
I'd try a bit more thermal paste?

Horses for courses and that. But for my 3900x I put 3 good blobs on. One over each die and the I/O die as well. If you look online you'll be able to figure the layout. I have zero issues with temps in my system (granted I am WC but still) - what have you got to lose?

The behaviour you describe though are the typical symptoms with bad HS contact to the CPU
 
ChrisD. said:
Haha. I laugh but I put a waterblock on my 1080 Ti and only after putting the HSF back on a few years later to sell it did I realise all of the memory strips had the plastic protective cover on. :D
Click to expand...
I nearly did the same in my recent build. Put the GPU block on, it wasn't until I was about to start fitting the tubing it dawned on me I had not removed the plastic protective tape from the thermal pads :D

Whipped the blocked off and removed. Phew! Glass tubing in my rig, so I am glad I caught that one! :eek:
 
Phate said:
I nearly did the same in my recent build. Put the GPU block on, it wasn't until I was about to start fitting the tubing it dawned on me I had not removed the plastic protective tape from the thermal pads :D

Whipped the blocked off and removed. Phew! Glass tubing in my rig, so I am glad I caught that one! :eek:
Click to expand...
That's lucky! I have to say the card still OC'd well and I never had any memory related problems.
 
Phate said:
I'd try a bit more thermal paste?

Horses for courses and that. But for my 3900x I put 3 good blobs on. One over each die and the I/O die as well. If you look online you'll be able to figure the layout. I have zero issues with temps in my system (granted I am WC but still) - what have you got to lose?

The behaviour you describe though are the typical symptoms with bad HS contact to the CPU
Click to expand...

Yeah I'll give that a go. Worst case I still have an EK water kit with 360mm rad but I was lazy to maintain it last time and recently took it out (2 years + without cleaning, not great!) and changed back to air cooling.

Woodsta888 said:
That MOBO isnt recommended for 3950X, time for an X570

https://www.reddit.com/r/Amd/comments/c7qj5e/am4_vcore_vrm_ratings_to_help_you_decide_on_a/
Click to expand...

I was trying to avoid upgrading the Mobo so I didn't have to buy a new Win 10 license to be honest .. but might upgrade that too in the not too distant future!
 
Right ... tried re-seating heatsink again, added a little more thermal paste and pretty much the same - close to 60c without load, and single core maxed out it goes to about 80c within a few seconds.

Might just clean the heatsink / CPU tomorrow and do it again from scratch to rule that out as a problem , pretty sure I have another tube of different paste somewhere too.
 
W10 licence is only £12 these days for a pro key. Does say in that list that your MOBO should run the 3950X at stock but not OC'd but seeing as these CPU's basically OC themselves (unless you want an all core OC) then really your MOBo should be fine I guess. Hope you sort it soon.
 
Woodsta888 said:
W10 licence is only £12 these days for a pro key. Does say in that list that your MOBO should run the 3950X at stock but not OC'd but seeing as these CPU's basically OC themselves (unless you want an all core OC) then really your MOBo should be fine I guess. Hope you sort it soon.
Click to expand...

£12 for a legit key? Damn, I think last time I bought one from MS online directly ... can't remember how much that was but a lot more than £12.

Anyway let's see how I get on with a third re-seating, will leave that for later or tomorrow.
 
Ironic Namesake said:
VRMs overheating - this shouldn't affect the idle temps I assume?
Click to expand...

What is your idle frequency across all of the cores? What power plan are you running in Windows?

If there is a legitimate yet high current load being requested by the CPU at idle, then the VRM is working harder than it needs to be thus getting hotter, and when you try an all-core load the VRM is getting too hot can't regulate the voltage properly and causes the CPU to go in to over current protection mode.
 
Journey said:
What is your idle frequency across all of the cores? What power plan are you running in Windows?

If there is a legitimate yet high current load being requested by the CPU at idle, then the VRM is working harder than it needs to be thus getting hotter, and when you try an all-core load the VRM is getting too hot can't regulate the voltage properly and causes the CPU to go in to over current protection mode.
Click to expand...

Power Plan is set to High Performance ... task manager reports 4.08 GHz, Ryzen Master 2.6 GHz and fluctuating .. HWMonitor reports 4099 MHz

I just noticed in HWMonitor, the individual cores are reading as 1.1V approx ... but CPU VCORE shows 1.52V : https://imgur.com/ibqs2zw
 
Ironic Namesake said:
Power Plan is set to High Performance ... task manager reports 4.08 GHz, Ryzen Master 2.6 GHz and fluctuating .. HWMonitor reports 4099 MHz
Click to expand...
using Balanced power plan may help with high idle temps. Also many Zen2 users are reporting high idle, which is a monitoring software side being incompatible with quick boost algorithm.

Ironic Namesake said:
Prime 95 on a single thread, temp goes up to about 77-80c. Not hot enough to force a shut down at least
Click to expand...
Single thread load spiking to 80 seems about right to me.
Are you certain that shutdown on full load is caused by temperature? It feels like VRM/power delivery thing.

If you have time, experiment in prime95 gradually increasing number of workers. See how temperatures and power consumption (in hwinfo) change. See if you hit the limit on either
 
You must log in or register to reply here.
Back
Top Bottom