NH-U14S-TR4 running hot on stock 1920X

[eidairaman1]

They look very similar except for the cold plate, with very similar dimensions (145x65mm vs 150x52mm - I didn't find fin counts), so if the Ashura had a cold plate suitable for TR4 I'm guessing it would perform pretty much the same - dependent on the fans, of course.

True, it is AM4 compatible, unsure if it is TR4 though

 

[Valantar]

True, it is AM4 compatible, unsure if it is TR4 though

Doubtful - the mounting mechanism is entirely different, with threaded holes in the socket retention plate. Also, plenty of tests have shown that you need a cold plate designed for TR4 to cool it properly due to the distributed dice. Even relatively large round Asetek cold plates are too small, and they're bigger than all other consumer IHSes. Threadrippers are huge.

 

[eidairaman1]

Doubtful - the mounting mechanism is entirely different, with threaded holes in the socket retention plate. Also, plenty of tests have shown that you need a cold plate designed for TR4 to cool it properly due to the distributed dice. Even relatively large round Asetek cold plates are too small, and they're bigger than all other consumer IHSes. Threadrippers are huge.

Yeah I know of the sheer size. Scythe hasnt jumped in to provide a water kit or TR4 style coolers.

 

[HTC]

@Valantar : did you test fan #1 VS Fan #2?

You were suspecting a bad performing 1st fan: comparing it VS 2nd fan should give you an answer to whether or not the 1st fan "requires" RMA.

 

[Valantar]

@Valantar : did you test fan #1 VS Fan #2?

You were suspecting a bad performing 1st fan: comparing it VS 2nd fan should give you an answer to whether or not the 1st fan "requires" RMA.

I'm suspecting that the spec sheet for the cooler has an error, as the new fan runs at almost exactly the same speeds as the one that came with the cooler, and its spec sheet says 1200rpm, not 1500 like the cooler. It would be odd of them to make a faster version of the fan just for this cooler - I'm guessing that would require an entirely new motor, which sounds expensive. So I'm going for "spec sheet misprint" as the explanation. I didn't bother to do A/B testing when I saw how close the RPM numbers were (in fact, the new fan is ~30rpm slower). Still, I guess I ought to throw an email Noctua's way to clear this up.

 

[HTC]

I'm suspecting that the spec sheet for the cooler has an error, as the new fan runs at almost exactly the same speeds as the one that came with the cooler, and its spec sheet says 1200rpm, not 1500 like the cooler. It would be odd of them to make a faster version of the fan just for this cooler - I'm guessing that would require an entirely new motor, which sounds expensive. So I'm going for "spec sheet misprint" as the explanation. I didn't bother to do A/B testing when I saw how close the RPM numbers were (in fact, the new fan is ~30rpm slower). Still, I guess I ought to throw an email Noctua's way to clear this up.

1200 VS 1500 is 20% slower: i call that significant.

Wouldn't hurt to ask.

 

[Valantar]

1200 VS 1500 is 20% slower: i call that significant.

Wouldn't hurt to ask.

It's definitely significant, but as I said, I'm chalking it up to a typo and my expectations thus being off. I'll toss them an email after work just to be sure, though. Who knows, maybe they're like Logitech customer support, and they'll send me a new cooler just to be nice (Logitech sent me a replacement retail package, no questions asked/proof of destruction required, for a >3-year-old, out-of-warranty G602 mouse with a failed middle click (mouse wheel) button. And I've had similar experiences with them earlier also. Their support is truly second to none.)