Question Need Help With Liquid Metal

[Nero1024]

I want to use liquid metal on NVME SSD between chips and copper heatsink. Is it safe?

 

[Nero1024]

Point of diminishing returns. With a cpu it's @ 70°C. Under that temp, there's zero benefit to a cooler running chip.

And no, you are making a basic mistake. There's more than 1 temp going on in any chip. Surface temps affect core temps, depending on the level of cooling, but the two are not the same. Even with extreme measures like LN2 thats -196°C, the cpu is still running 30-70°C core temps, the liquid nitrogen doing nothing but absorbing excess heat from the workload and speeds.

A cpu at 40°C will last just as long as if that same cpu spent its entire life at 60°C. Having the lowest possible temp is just as much a falicy as having the highest possible fps, once you pass the point of diminishing returns, it's pointless.

I know that radiator is doing it's job by removing excess heat. That's why you need radiator in the first place.
And no, lower temperature is always better, because it will affect all other hardware temperatures in your PC box. Also, where is the proof that "A cpu at 40°C will last just as long as if that same cpu spent its entire life at 60°C". Do you have physics/engineer degree?

 

[Karadjgne]

Do you need an ASE Certification to put air in your tires, gas in the tank or change a sparkplug?

And yes, I did go to ITT Technical instute and have an AS in electronics engineering.

And yes, I still own a PentiumII 350MHz, OC to 400MHz. The realistic lifespan of a cpu based on silicon ic's manufacturing has yet to be determined as software renders a cpu obsolete far sooner than actual elemental damage under normal operational guidelines. It's only stupidity in implementation, such as running cpus above recommended voltages that shortens lifespans, and even that cannot be accurately determined since physical properties on the silicon and manufacturing process differ for every cpu. What's recommended by Intel or amd is a recommendation, to cover any cpu, not an absolute fine-line maximum for every cpu.

Up until a few years ago, airflow, temps, psu output voltages weren't even a consideration, if the cpu worked and didn't crash, that's all that mattered. It's only recently that anyone has bothered to explore further into overclocking from a performance standpoint, airflow from a cooling efficiency standpoint, psu output voltages and rail ripple from a stability standpoint etc.

I've been building, diagnosing, repairing, customizing pc's for just shy of 40 years. You are going to have a rude awakening one day, on the day you realize that ppl often know a whole hell of a lot more about a hell of a lot more than what some piece of paper certificate from a school that didn't teach them much of anything they didn't already know, says they do.

But you go one believing in your piece of paper. You've been advised by multiple sources, some far smarter and more knowledgeable in that particular subject matter than me, that using liquid metal on a m.2 drive is a fools endeavor. But you go on believing that they know nothing either, since they probably don't have degrees in thermal dynamics and physics relative to the chemical and electromechanical properties of metals that occur naturally in a liquid state at room temperature.

If all that is above your level of comprehension and understanding, I'll put it bluntly. I don't have any schooling in Karate whatsoever. But if I punch you in the face, that lack of a colored belt isn't going to make it hurt any less.

As Vader so succinctly put it "I find your lack of Faith, disturbing..."

Take their word for it, LM on an M.2 = very bad idea.

 

[Nero1024]

Do you need an ASE Certification to put air in your tires, gas in the tank or change a sparkplug?

And yes, I did go to ITT Technical instute and have an AS in electronics engineering.

And yes, I still own a PentiumII 350MHz, OC to 400MHz. The realistic lifespan of a cpu based on silicon ic's manufacturing has yet to be determined as software renders a cpu obsolete far sooner than actual elemental damage under normal operational guidelines. It's only stupidity in implementation, such as running cpus above recommended voltages that shortens lifespans, and even that cannot be accurately determined since physical properties on the silicon and manufacturing process differ for every cpu. What's recommended by Intel or amd is a recommendation, to cover any cpu, not an absolute fine-line maximum for every cpu.

Up until a few years ago, airflow, temps, psu output voltages weren't even a consideration, if the cpu worked and didn't crash, that's all that mattered. It's only recently that anyone has bothered to explore further into overclocking from a performance standpoint, airflow from a cooling efficiency standpoint, psu output voltages and rail ripple from a stability standpoint etc.

I've been building, diagnosing, repairing, customizing pc's for just shy of 40 years. You are going to have a rude awakening one day, on the day you realize that ppl often know a whole hell of a lot more about a hell of a lot more than what some piece of paper certificate from a school that didn't teach them much of anything they didn't already know, says they do.

But you go one believing in your piece of paper. You've been advised by multiple sources, some far smarter and more knowledgeable in that particular subject matter than me, that using liquid metal on a m.2 drive is a fools endeavor. But you go on believing that they know nothing either, since they probably don't have degrees in thermal dynamics and physics relative to the chemical and electromechanical properties of metals that occur naturally in a liquid state at room temperature.

If all that is above your level of comprehension and understanding, I'll put it bluntly. I don't have any schooling in Karate whatsoever. But if I punch you in the face, that lack of a colored belt isn't going to make it hurt any less.

As Vader so succinctly put it "I find your lack of Faith, disturbing..."

Take their word for it, LM on an M.2 = very bad idea.

Ok. Bla, bla, bla. You still didn't say why it's a bad idea? Too much words and very few meaningful ones relevant to the topic. If you know physics, you should know that matter decays slower on lower temperatures

 

[USAFRet]

You might want to look up the term "operating temperature range".

 

[Nero1024]

You might want to look up the term "operating temperature range".

Facepalm

 

[USAFRet]

Also, we're not necessarily discounting cooling down the drive.
Just that Liquid Metal probably isn't the best choice for that application.

 

[Nero1024]

Ok. I didn't knew I have to be this specific to get the answer for my question. I am not here for YOUR opinion or bla, bla, bla, discussion, I don't give a ****. I just want technical, professional answer if it's safe to use it or not. Thats all

 

[USAFRet]

Ok. I didn't knew I have to this specific to get the answer for my question. I am not here for YOUR opinion, I don't give a ** . I just want technical, professional answer if it's safe to use it or not. Thats all

Liquid Metal - No. I wouldn't.
Some other thermal paste - Sure.

 

[Nero1024]

Liquid Metal - No. I wouldn't.
Some other thermal paste - Sure.

Then, WHY?

 

[USAFRet]

Then, WHY?

Being a conductive material, if/when some of it oozes out onto the PCB...it may short something out.

That material was made for a whole different application. Under the heat spreader on a CPU. Trapped inside.
Not where it could potentially short something out.

Any good quality non-conductive material would work just as well.

 

[Nero1024]

Being a conductive material, if/when some of it oozes out onto the PCB...it may short something out.

That material was made for a whole different application. Under the heat spreader on a CPU. Trapped inside.
Not where it could potentially short something out.

Any good quality non-conductive material would work just as well.

You missed that nothing can be as conductive as liquid metal = best option ever if you take every danger into account and try to mitigate or void it

 

[USAFRet]

As was said much earlier...you seem to be dead set on this path.
You've gotten multiple opinions.

Go for it.

 

[Nero1024]

As was said much earlier...you seem to be dead set on this path.
You've gotten multiple opinions.

Go for it.

In your opinion thermal paste would be as fine option on NVME SSD chips?

 

[USAFRet]

In your opinion thermal paste would be as fine option on NVME SSD chips?

If I were so inclined to do this, yes.

 

[DSzymborski]

As it seems clear that the original poster intended to do whatever he wanted under any circumstances, I'm closing this thread so as not to waste the further time of moderators and posters.