so... I came across a liquid ultra tube today.......

I was in micro center where I feel like a kid in toys r us. They usually don't have liquid ultra there but today they did, and liquid pro too, and i decided that I was going to pick a pack of ultra up because I read liquid pro is worse in every way. I'm wondering how its holding up for everyone that has tried it after a few months of use? I want to try it but I wonder if the benefits are worth the risks that I have heard. The risks of hardening and staining cpu dies, one guy even said it got so hard that he needed to use a knife, sand paper, and polish compound to remove it after a long period of use. Then of course is the risk of a short.... Has anyone come across any of these issues I have read? Is it worth the risk? I haven't seen it mentioned around here again in a while. Thanks!

 

I repasted with Liquid Ultra about a month ago, that might not be as long as long a test period as you're looking for but you'll get an idea of what it's like after a month at least. I can say that my temps are still where they were after repaste.

I've just cleaned it off both GPUs and heatsinks (in preparation for GPU replacement, wanted the heatsinks to be clean), it was still mostly liquid on the die and heatsink, easy to get off. However, the heatsink did have a small layer that wouldn't come off easily. The included abrasive sponge got it off easily on the heatsink, but there is still silver-coloured staining on the heatsink. It shouldn't affect thermal performance by a detectable amount (as it is still metal), but be aware of that.

With that said, the CPU sees far higher thermal fluctuations that could cause the paste to wear out faster. (I experienced that with MX-4 - it literally degraded in two days, with a 10C jump in temperatures, after application on the CPU. The GPU applications of MX-4 on the other hand were fine, although I still replaced them)

Honestly, what I'd say is that it seems to be worth it on the CPU but not so much on the GPUs. On 7970Ms at least, thermal performance was no better with Liquid Ultra than with IC Diamond or even MX-4. All of them peaked at about the same temperature, within margins of error. The situation might be different on nVidia's GPUs, particularly if you're overclocking them, but you will see far greater gains on the CPU.

Personally I'd just use Liquid Ultra on the CPU and use something like IC Diamond on the GPUs.

 

I don't think youre going to get Liquid Ultra off of the heat-sink, which to me isn't a big deal. While investigating a temperature difference between my primary and secondary GPU's a few weeks after applying this TIM I attempted to clean the heat-sink of the secondary (the GPU that was running hotter) and after a few wipes with a cotton rag and alcohol I realized that it was not coming off, which was fine for me, I simply re-applied a thin layer of Liquid Ultra back onto the GPU die and re-fitted and thoroughly tightened down the heat-sink.

There is a temperature reduction in fact, but again, it makes no difference if youre conducting heat from the components to the heat-sinks more effectively if youre not conveying that heat out of the system more effectively, the importance of elevating the laptop in conjunction with manual fan-control.

With IC Diamond I was seeing temps of 75-76 C primary and 69-72 C secondary while playing Far Cry 3 all settings maxed with GPU clocks at +250 core/+500 memory @ 1.025 V. Now I just see 67-70 both primary and secondary. Tbone has reported similar results.

http://forum.notebookreview.com/ali...-liquid-ultra-any-tips-before-i-start-37.html

Were seeing temps lower than with the factory clocks and voltage (+250 core/+500 memory @ 1.025 V, Tbone on 1.05 V).

So there is a difference, and the risk is relatively minimal, you just have to take your time and not get any on the exposed surfaces surrounding the GPU die.

 

talking about shorting risks, just follow my guide of silicone-dam around the CPU die to prevent any excess running off and down the "valley of motherboard components" Additionally I think making a silicone-dam like I have explained in the thread linked by vulcan ^^^ may extend longevity of the compound as it would almost act isolated from atmospheric conditions, could probably last longer, age slower.

I too have experienced a heatsink staining but not that it made temps be any worse after re-application. Must thing is you have to apply it to CPU die and heatsink area of die-impact for the best results, and you should not need to replace it until you begin to see major raises in temperatures. I reapplied month+ later just out of curiosity (and wanting to waste some time, had nothing better to do). I switched back to ordinary paste as GC gelid Extreme, which wasn't a bad paste, but nothing compares to Liquid ultra, so I have switched back to Ultra again with a silicone dam method for a total piece of mind. A tip: add some clear sealing tape on CPU pcb around the die where you're gonna apply the silicone (if you are) so it doesn't cure to the CPU if it will, as some brands do harden and attach pretty well to surfaces, where as to sealing tape it will not cure as hard and it will be easy to separate in the future.
Hope this helps

 

That or a foam shim would do the job.

 

hmmm. thank you all for the replies. I think I may give this a shot when I'm feeling brave next weekend. Till then I'll leave in IC DIamond. I was thinking though, how could it dry out if its 100 percent liquid metal. If there isnt any water or other solution in it to evaporate, how could it ever dry?

 

in theory it shouldn't and it doesn't, like few people stated somewhere before it should last years, the only culprit to blame is the purity of the heatsink copper ratio as its never just a "virgin pure copper" hence it acts and bonds sometimes to the metal heatsink. if your heatsink would be nickel plated that would most likely totally eliminate curing to heatsink issues