Liquid pro and liquid ultra arrived today. Used Ultra, result within.

[Ferzerp]

I'm delidded, so I have to decide what TIM to use for each interface.

Any suggestions for both?

edit:

Here is the result with an H100 on medium, using the ultra on both interfaces.

 

[Idontcare]

Technically, liquid ultra is superior to liquid pro. See tw33k's thread for the long story.

 

[Ferzerp]

Technically, liquid ultra is superior to liquid pro. See tw33k's thread for the long story.

I'll use it on both interfaces then! Now to decide if I'm up to it tonight to do it. hah.

 

[Vectronic]

Liquid pro and liquid ultra arrived today. Suggestions?

Body art.

I'm delidded

That's the spirit.

... so I have to decide what TIM to use for each interface.

Whoa there, Rated R.

See tw33k's thread for the long story.

Coollaboratory Liquid Ultra vs Liquid Pro

 

[LagunaX]

OCN seems to say same or better results with Liquid Pro (I could be wrong), but they have a huge delidded thread there.

 

[Idontcare]

OCN seems to say same or better results with Liquid Pro (I could be wrong), but they have a huge delidded thread there.

The CEO of the company that makes both products says the Ultra has superior thermal conductivity to the Pro.

I'm not about to argue with him, and it leaves me inclined to view the "pro Pro" reports as being more of a placebo effect than the result of actual apples-to-apples robust testing studies.

How else can you explain the discrepancy? I assume the guys who developed Pro and Ultra are the experts in the discussion, the enthusiasts doing uncontrolled tests in their homes and basements are, presumably, the amateurs in the discussion.

 

[Ferzerp]

Not bad. Updated OP with result.

 

[Idontcare]

Nice :thumbsup:

Did you re-optimize your Vcc for that specific GHz now that the operating temperature has been lowered?

 

[Ferzerp]

Nice :thumbsup:

Did you re-optimize your Vcc for that specific GHz now that the operating temperature has been lowered?

I started to. I'm at + .015V offset in that shot. -0.005V gets ecc errors. 0.005V appeared ok, but I didn't have time tonight for full testing. There is no 0 offset. That sets it to auto and it jacks the voltage way up, then. -0.005V or less with more LLC might do it though if +0.005V turns out to be stable. My CPU is a bit of a lemon. +0.015V was stable with the nt-h1.

 

[footballrunner800]

Do you have any before temperatures to compare?

 

[tw33k]

OCN seems to say same or better results with Liquid Pro (I could be wrong), but they have a huge delidded thread there.

This is exactly what prompted me to run tests between the two. There was a lot of misinformation being posted. People claimed to have used both and found Pro performed better but no one posted any results. A guy posted in Hard forums a while back that Liquid Pro was rated at 82w/mk and people simply believed him and concluded that Pro was better. It wasn't until I received an email from the CEO of Coollaboratory with the actual figures that people started changing their tunes.

I recently received 3x Pro and 3x Ultra so I can run more tests but I have to say to say that my initial findings are that there's nothing between the two (not enough to justify switching from one to the other anyway)

 

[KingFatty]

I guess it's somewhat interesting to know what results amateurs achieve with the actual product, because that's the typical use scenario?

If it takes a professional lab to install the product to obtain the theoretical results, then the end-user can practically ignore that info.

I would be interested in knowing what a user like myself can get from a product, such as a consensus among actual/amateur users messing around with installing the stuff in home office environments.

 

[Idontcare]

I guess it's somewhat interesting to know what results amateurs achieve with the actual product, because that's the typical use scenario?

If it takes a professional lab to install the product to obtain the theoretical results, then the end-user can practically ignore that info.

I would be interested in knowing what a user like myself can get from a product, such as a consensus among actual/amateur users messing around with installing the stuff in home office environments.

The issue there is you have so many uncontrolled variables in each case when it comes to the at-home user.

Way to many opportunities for things to go tits up from user error, only to then be blamed on the product because everyone likes to rate themselves a "4 out of 4" for technical expertise.

It takes someone with dedication and resolve to do exactly what tw33k did, to compare the two products in a controlled setting in which the only variable that is being changed is the TIM itself. Same hardware, same environment, same end-user making the application.

It makes the Pro vs. Ultra results as apples-to-apples comparable as it can get.

 

[Ferzerp]

Do you have any before temperatures to compare?

Not formal because of the pump-out effect. The best I ever got with NT-H1 was high 60's (looking at screenshots, 67 is the lowest I can find, but often up to 70 depending on which application, also the ambient is warmer right now), but it would only last a few weeks and get progressively worse until it was in the 80's. I'm hoping that doesn't happen here. I've reinstalled so many times that I can't remember.

So, worst case scenario, a 5C improvement over my best with NT-H1

 

[LagunaX]

Yeah that's why I use MX-4 or IC 7 - no pump out effect.