***The Official 5900X \5950X owners thread***

[Chud_Jenny]

I had no problems installing everything,5600X was there before 5900X,great temps and everything.

You mean 4 screw which are goes to the Mobo,or 4 screw which are securing the block

The 4 that secure the block. That's where there the tension should happen. There's a real change the bars under the block have bent slightly so the tension isn't there.

 

[Vigorito]

Yes 4 screw which are goes from block and securing cpu are goes till end no problem

 

[Chud_Jenny]

Yes 4 screw which are goes from block and securing cpu are goes till end no problem

That's the issue then I'd think.

 

[Vigorito]

Hah i dont think so,if that is problem then 5600X shouldnt work this good.

 

[Chud_Jenny]

Hah i dont think so,if that is problem then 5600X shouldnt work this good.

Removing it for the change may well of caused it. It's just a matter of bending the arms up very slightly to push the cold plate down more. I've had this issue years ago with a coolermaster aio.

 

[Vigorito]

Okay so at this moment,which revision is safe to get it

 

[Chud_Jenny]

Any new stock should be version 4.

 

[Vigorito]

Only revision 4 so is safe,ok thx

 

[ianh]

Offcourse i am,are you running stock? And what excatly temp value are you reading Core temp or different value

I'm using both Ryzen master and CPUID HWMonitor (CPU temp taken from the MoBo rather than the Chip) and both CPU temps were matched through-out the test, unlike TMPIN7-8 & the 5900x temp which were 30+ seconds off before they matched. I'm running my 5900x completely stock, all 12 cores running at 4.13Ghz on CineBench R23 (slowly dropping to 4.1Ghz over time) and temps on this 10 min run only hit 60'c (got some windows open so more of a cool breeze blowing through) and again took about 2 mins to settle at that temp, with it very quickly rising from 30'c idle to 50'c within a few seconds and then a slower crawl to reach 60'c.

My Arctic Freezer 360 V2 is revision 4 also, with the heatsink mounting plates (the two black curved bars) which have offset mounting holes specifically for the 5900x & 5950x so you get better contact on those chips.

 

[Vigorito]

Are we talking about this bars?
If yes i have them on REV.3

https://ibb.co/HPSngDJ

This is what i have from screws and the rest,12 screws for radiator are on rad.

https://ibb.co/0jwmLH7

 

[555BUK]

This is the best my 5900 can do with memory timings. Should I chase lower-CAS or higher MHz?

 

[ianh]

Are we talking about this bars?
If yes i have them on REV.3

https://ibb.co/HPSngDJ

Not those ones, it's the ones in the image below -



Here's a link to the manual which describes which holes to use and which orientation the pump is supposed to be in too - https://support.arctic.de/lf2-360r4

 

[Vigorito]

I check everything today,i have identical equip. In rev.3 to the same detail,i dismounted everything and start from a scratch,i think i mounted today better,testing now cine23 this morning was 1501 stock single core at 75c,now is 1590 at 67c,there is a progress,but im not into numbers and testing,will test in games

 

[Vigorito]

@ianh did you mounted it like this

https://ibb.co/QPv2XBs

 

[ianh]

@Vigorito Yeap, then the pump goes in with the built-in mini fan aimed at the bottom Top of the case like the image above.

I top mounted my rad with the pipes at the front of the case, used kyronaut paste and have a very good airflow case (thermaltake full tower something) and I only fit 75'c+ with PBO enabled jumping me to 4.4ghz all core in Cinebench R23 IIRC.

 

[Vigorito]

I mean did you select and use a screw holes like i did,arctic paper tutorial and you tube video for rev.3 and rev.4 for ryzen 3000/5000 are a mess,thats why im asking you is this a right mounting offset for 3000/5000,i use Tg Kyronaut paste too,lian li o11d case so ok airflow,now the thermals of 5900x are much better in cine23 temps are around 68,in games average 75c,idle is good around 33 or 40 but somethimes and 50 (im only monitoring core temps not any other value) i have often a high spikes at 68-70 and there are longer spikes not just for few seconds,ant there are little strange im looking right now at 50c idle core temp of 5 minute spike,and its not going down but its weird,its interesting that every time i open ryzen master temps from 50c idle goes down at the same momemt at 30c

 

[Chud_Jenny]

Are we talking about this bars?
If yes i have them on REV.3

https://ibb.co/HPSngDJ

This is what i have from screws and the rest,12 screws for radiator are on rad.

https://ibb.co/0jwmLH7

It just looks like a poor mounting system to me. It the screws easily bottom out it shows there's not enough mounting pressure hence the poor temps. If you where to slighting bend the bars that screw under the block that would help. So bend the ends up to push the cold plate down into the CPU.

 

[ianh]

I mean did you select and use a screw holes like i did

Yeap, I used the same screw holes #2 (for AMD 3000/5000) as you have also done.

It just looks like a poor mounting system to me.

Quite literally tens if not hundreds of thousands of other Arctic Freezer users without any temp issues would probably disagree, compared to a handful with issues

 

[Trox]

Reading this thread it's clear to me that people with good temps have restricted the performance of the CPU by one way or the other.

70c on load during Cinebench is fine, the CPU generally jumps around. 30-40 c at idle is fine.

It's good for 90c according to the AMD spec sheet afterall...

 

[Vigorito]

"Reading this thread is clear to me that people with good temps have restricted the performance of the CPU by one way or the other"

After investigating for several days al the forums im thinking the same.