buttabean
Good to know. I'm about to start on mine now.

After updating the bios, the fans run a lot faster on my xc gaming 2080ti. Before max 100% fan speed I was getting 76-80c and then after I had to change my fan curve to 72% which resulted in below 70c Temps. At 100% it was very loud but I saw the temp drop down to 60c so that's at least good to know for anyone wondering about temps.

buttabean
So far with my h55 and the stock settings on the 2080ti I'm reading 59-60c with the radiator fan at 62%. The left memory is reading 68c. I put my top cover on which has a big 200mm fan that blows down. With heaven benchmark RUNNING, the temps dropped off to 58c on the gpu, memory 60c.

I think I'm going to design a small cooler for the left side that attaches to the nxzt bracket

Grimsin

buttabean
So far with my h55 and the stock settings on the 2080ti I'm reading 59-60c with the radiator fan at 62%. The left memory is reading 68c. I put my top cover on which has a big 200mm fan that blows down. With heaven benchmark RUNNING, the temps dropped off to 58c on the gpu, memory 60c.

I think I'm going to design a small cooler for the left side that attaches to the nxzt bracket

How are you reading the memory temps?

buttabean
I used my laser thermal probe. The left side is exposed. I'm sure the side with the bracket fan is very cool.

buttabean
Ha! That's awesome. I increased the memory to 605+. So far I've got an increase of 10fps in heaven over stock settings. I'm going to let it run and see what the temps max to before going higher

buttabean
Sitting at 76c. That's pretty good. I'm going to leave it at those until I get some heatsinks on it.

I think yours in oc higher because it's running cooler. I noticed when I started heaven while cooled it ran above 2010+mhz before it hit 60c. So definitely worth using a 240mm radiator for max oc

arestavo
To those that moan about too much paste....yeah, that's not a thing if it's non-conductive and non-capacitive. Jayz2cents did a video on it for a GPU application with SO much paste it wasn't funny and it actually didn't increase temps at all.
 
Here's a very rigorous test that Gamer's Nexus did for CPU application that also shows ...(wait for it)... within margin of error results.

GTXJackBauer

Grimsin

arestavo
To those that moan about too much paste....yeah, that's not a thing if it's non-conductive and non-capacitive. Jayz2cents did a video on it for a GPU application with SO much paste it wasn't funny and it actually didn't increase temps at all.

Here's a very rigorous test that Gamer's Nexus did for CPU application that also shows ...(wait for it)... within margin of error results.

Very informative, I did not know this so thanks!

While I respect some of his reviews (Gamer Nexus), this is ridiculous for these guys to say that too much thermal paste isn't an issue at all and their methodology could be questioned.  This was done hastily and the proper testing should have been done based on core temps to see if there were any possible air pockets or just too much paste in that part of the IHS and such, etc while using heavy benching like Prime95 to see how well it performed.  Of course those in the higher end packages and high OC benchers will be effected the most.
 
Yes, the closer you are to the block, the better heat dissipation.  The goal on thermal paste is to have an slim film of it between the IHS and the block as best you can but of course it will never be perfect.  I sometimes use the pea size method or cross.  Both have worked great for me.    Many have come forward and redone their thermal paste when there was too much applied originally to the most optimum way possible and have seen improvements, including myself.  By improvements I mean 5c-10c differences per core which imo is a great improvement.
 
I still believe too little and too much will effect performance at the end of the day.

arestavo

GTXJackBauer

Grimsin

arestavo
To those that moan about too much paste....yeah, that's not a thing if it's non-conductive and non-capacitive. Jayz2cents did a video on it for a GPU application with SO much paste it wasn't funny and it actually didn't increase temps at all.

Here's a very rigorous test that Gamer's Nexus did for CPU application that also shows ...(wait for it)... within margin of error results.

Very informative, I did not know this so thanks!

While I respect some of his reviews (Gamer Nexus), this is ridiculous for these guys to say that too much thermal paste isn't an issue at all and their methodology could be questioned.  This was done hastily and the proper testing should have been done based on core temps to see if there were any possible air pockets or just too much paste in that part of the IHS and such, etc while using heavy benching like Prime95 to see how well it performed.  Of course those in the higher end packages and high OC benchers will be effected the most.
 
Yes, the closer you are to the block, the better heat dissipation.  The goal on thermal paste is to have an slim film of it between the IHS and the block as best you can but of course it will never be perfect.  I sometimes use the pea size method or cross.  Both have worked great for me.    Many have come forward and redone their thermal paste when there was too much applied originally to the most optimum way possible and have seen improvements, including myself.  By improvements I mean 5c-10c differences per core which imo is a great improvement.
 
I still believe too little and too much will effect performance at the end of the day.

You're the reason why these videos get made.
 
Too much (non-conductive/non-capacitive) paste isn't bad. Too little paste kills chips, especially GPU cores. Changing paste types changes thermal conductivity - low quality paste, dried out paste? New, better thermal conductivity paste fixes that.
 
Find evidence proving me wrong and PM me it. I'd love to see the evidence to the contrary.