Used mx-4 after a delid

[Watashi]

I know that recently liquid metal has been becoming increasingly popular. However, I can't really say with confidence that it is worth the risk of using. Considering it is conductive, that is enough to make me a bit nervous honestly. I just completed my first successful delid using mx-4, and ostensibly dropped 10-15 C under load. If these temps are confirmed, it makes liquid metal seem like a bit of an unnecessary risk to me. Thoughts? 

[JoostinOnline]

If you're happy with it (and you should be) then there is no reason to risk something conductive.

[Watashi]

3 hours ago, JoostinOnline said:

If you're happy with it (and you should be) then there is no reason to risk something conductive.

I agree with that. I risk a custom water-cooling loop. But to be honest, it really does not feel like such a risk. Only 2 out of my 6 fittings would drip upon components if a leak did occur, and there is always the possibility that if a leak occurs, and is properly cleaned that no damage will even occur. 

[Guest]

Well, liquid metal is known for its significant temperature reductions, so I see no problem. Liquid metal is definitely better then any other aftermarket thermal compounds and the risk seems unecessary in your use case, considering the improved temps under load.

[porina]

Keep an eye on temps over time. I used MX-4 and MX-2, as well as the Noctua one for delids and found temperature increases over time. MX-4 and Noctua I put it down to them being runny, but even the thicker MX-2 did similar. I've only started trying liquid metal, it certainly provided a further improvement over conventional pastes, but too early to comment on longer term stability.

[JoostinOnline]

5 minutes ago, porina said:

Keep an eye on temps over time. I used MX-4 and MX-2, as well as the Noctua one for delids and found temperature increases over time. MX-4 and Noctua I put it down to them being runny, but even the thicker MX-2 did similar. I've only started trying liquid metal, it certainly provided a further improvement over conventional pastes, but too early to comment on longer term stability.

That's why the Intel TIM is actually good.  It doesn't perform as well as most other pastes, but it lasts way longer.

[Watashi]

3 hours ago, TheBeastPC said:

Well, liquid metal is known for its significant temperature reductions, so I see no problem. Liquid metal is definitely better then any other aftermarket thermal compounds and the risk seems unecessary in your use case, considering the improved temps under load.

That's a good point, and I can understand people wanting the best possible performance out of their components. I know it is something I enjoy striving towards

[Watashi]

3 hours ago, porina said:

Keep an eye on temps over time. I used MX-4 and MX-2, as well as the Noctua one for delids and found temperature increases over time. MX-4 and Noctua I put it down to them being runny, but even the thicker MX-2 did similar. I've only started trying liquid metal, it certainly provided a further improvement over conventional pastes, but too early to comment on longer term stability.

About how long would you say this takes? I would guess a few months?

[Watashi]

3 hours ago, JoostinOnline said:

That's why the Intel TIM is actually good.  It doesn't perform as well as most other pastes, but it lasts way longer.

That is a good point, but do you think it is because of that black goop they use to hold the substrate and the IHS together aswell? 

[JoostinOnline]

18 minutes ago, Doomerson said:

That is a good point, but do you think it is because of that black goop they use to hold the substrate and the IHS together aswell? 

The adhesive is a huge part of the reason delidding makes a difference.  They use a ton of it.  Take a look at the link in my signature for more details.

[porina]

1 hour ago, Doomerson said:

About how long would you say this takes? I would guess a few months?

I didn't make precise notes but it was of the magnitude of several months.

[zberry7]

There’s a few factors which cause Intel CPUs to run so hot. The paste they use and there’s too much silicon which means the IHS is further away from the die. 

 

So so removing the extra silicon will net you some improvement, and better paste will also help. I saw a chart where someone compares different TIMs between the IHS and die, and the difference between liquid metal and normal paste was very significant. Almost 10C.

 

tldr: a lot of your improvement in thermals could be from just having less silicon adhesive and less thermal paste between the die and spreader. You could still probably pull another 5-10c out using a high quality liquid metal.

[DarkSmith2]

My experience with MX-4 is that it lasts longer then a year until it even starts drying. Its one of a few paste's thats known for its longevity.

But its cooling performance is a couple degrees worse then f.e. Thermal Grizzly Kryonaut.

 

For delidding conductonaut is the best LM you can use. Since it wont dry.

 

 

[ImNotThere]

liquid metal will just short parts of the CPU and just needs cleaning off and will work straight after afaik, pretty sure GN got liquid metal on a delidded 7900x and then just wiped it off.

[Watashi]

On 10/17/2017 at 9:52 AM, JoostinOnline said:

The adhesive is a huge part of the reason delidding makes a difference.  They use a ton of it.  Take a look at the link in my signature for more details.

Solid link in the signature, that picture you used actually helped paint a pretty clear picture why delidding can be so effective. I thought that the CPU PCB was called the substrate, is this a misunderstanding on my end?

[JoostinOnline]

1 minute ago, Doomerson said:

Solid link in the signature, that picture you used actually helped paint a pretty clear picture why delidding can be so effective. I thought that the CPU PCB was called the substrate, is this a misunderstanding on my end?

It is, but nobody knows what that means.

 

I also didn't make the picture myself.  I found it a long time ago (before I wrote that post) and it was in the Downloads folder on my phone.

[Watashi]

3 hours ago, JoostinOnline said:

It is, but nobody knows what that means.

 

I also didn't make the picture myself.  I found it a long time ago (before I wrote that post) and it was in the Downloads folder on my phone.

well done friend.

[dizmo]

You misused the world ostensibly.

[Watashi]

57 minutes ago, dizmo said:

You misused the world ostensibly.

You have no way of actually knowing that, considering you have no way to know if I had a 100% solid understanding of my temps (I don't) or if I used a proper testing methodology (I didn't)