30°C Drop in Temperature after repaste (MSI 2080 Super Gaming X Trio)

[Deleted member 193596]

i bought a 2080 super gaming x trio from a panic seller for 400 bucks (it was 3 weeks old and the sticker was still untouched on the heatsink screw)

after i installed it and overclocked it the card ran at 84°C at 80% Fan Speed.

i opened the Card, (thermal paste application was perfect) and repasted it with normal NT H1.

temps dropped by 22°C...

after an full hour of my new favourite GPU stresstest (wow in 4k with 2x scaling and 8x msaa pulls more power than furmark in certain areas.)

the card peaked at 64°C at 51% Fan speed (instead of 80%)
and then it settled at 62°C at 54% Fan speed.

at the same 80% Fan Speed the card was at 55°C.

this makes MSI (for me at least) the worst brand for pre applied thermal paste and dethroned ASUS and Zotac.

i Highly recommend to repaste MSI Cards aswell as ASUS and Zotac.

 

[mstenholm]

I owned +10 MSI from 970 to 2070 Super and have never had a problem. My 2070 Super runs at 32 % fan at 62°C. OK my load is never more than 99%, folding. Good that a new paste job did the job for you. I have a felling that the seller knew it was poor specimen.

 

[jayseearr]

Thank you for your recommendation but I've never seen my card go over 68 degrees so i think i'll pass on re-pasting. Much like mstenholm I've owned several msi cards and never had temperature issues to speak of on a single one of them.

Let's not try to put a blanket over msi based off of you single isolated used card buying experience

 

[Deleted member 193596]

Thank you for your recommendation but I've never seen my card go over 68 degrees so i think i'll pass on re-pasting. Much like mstenholm I've owned several msi cards and never had temperature issues to speak of on a single one of them.

Let's not try to put a blanket over msi based off of you single isolated used card buying experience

that wasn't the only MSI Card that had these issues for me.

i had a RX580 which ran ridiculously hot, and a 2060 Ventus (XS? or something) that peaked in the 90s and throttled (sent it back to amazon)

this is now 3 out of 3 for me where two where brand new and sealed.

 

[jayseearr]

that wasn't the only MSI Card that had these issues for me.

i had a RX580 which ran ridiculously hot, and a 2060 Ventus (XS? or something) that peaked in the 90s and throttled (sent it back to amazon)

this is now 3 out of 3 for me where two where brand new and sealed.

Well that is just flat out unfortunate and i can't necessarily blame you for feeling that way based off your own experience. My only point was, suggesting that everyone go and take apart and repaste any gpu that says msi on it is just an asinine thing to do or say

 

[micropage7]

that wasn't the only MSI Card that had these issues for me.

i had a RX580 which ran ridiculously hot, and a 2060 Ventus (XS? or something) that peaked in the 90s and throttled (sent it back to amazon)

this is now 3 out of 3 for me where two where brand new and sealed.

then you point MSI from your experience it's a bad brand, you need to look at some threads about other brands too

 

[EarthDog]

I've reapplied on MSI cards and didn't see much of a drop. I doubt this is a systemic issue as your click bait title implies, but appreciate you sharing your experience nonetheless.

 

[Deleted member 193596]

I've reapplied on MSI cards and didn't see much of a drop. I doubt this is a systemic issue as your click bait title implies, but appreciate you sharing your experience nonetheless.

changed the title for you. even if it was just my honest opinion after working with hundreds of graphics cards and systems since it was my job for several years now.
and it is a fact that MSI has absolutely awful thermal paste.. just as bad asus and zotac.
the only good cards i've seen so far were from Gigabyte and Inno3D


i dropped over 30 degrees and all of my MSI Cards were absolutely awful.. and i am not alone with that issue.

 

[EarthDog]

I've worked with just as many... haven't seen what you're seeing... not even close. I'm not taking away from your experience... just adding mine to the pile.

Thanks again for sharing your experience.

 

[droopyRO]

this makes MSI (for me at least) the worst brand for pre applied thermal paste and dethroned ASUS and Zotac.

I have a second hand GTX 1070 Ti GamingX that barley reaches 70ºC during the summer with about 1200 rpm. I did have a bad experience with a RTX 2080 Ventus that was hot and loud, that i bought new and i returned it. So yeah, we all have bad or good luck sometimes with hardware.

 

[Bubster]

That's fantastic .... Cooler and less noise...even though Nvidia keeps lying about 30 or 40% cooler every year on year

 

[John Naylor]

this makes MSI (for me at least) the worst brand for pre applied thermal paste and dethroned ASUS and Zotac.

I think you have to consider something other than the paste ... look at any TIM comparison .... never seen a best to worst anywhere near double digits. So where do we begin ?

1. If there was a bad paste, it's hardly logical that your 1 (or 3) anecdotal experience is representative of the entirety of the GPU market. Consider that the panic seller knew something wrong. The likely cause here has little to do with the paste and everything to do with a bad / uneven pressure mount. We have been building PCs for almost 30 years. From 2005 - 2013, we were almost exclusively an Asus Shop (GFX and MoBo). .... now most our builds are getting MSI cards because users tend to rank cards in order of 1) performance 2) noise and 3) temps, In recent years MSI has been putting up the best numbers .... we don't make the choice, we point the users to TPU reviews and suggest they choose their components accordingly.

The 2070 was the most poular card chosen and here's TPUs numbers in order of fastest card at top
Zotac AMP Extreme 145.4 fps 69 C 36 dbA
MSI Gaming Z 144.5 70 30
PalitGameRock 143.1 70 33
Asus Strix 142.7 65 37
Nvidia 2070 FE 138.7 77 34
EVGA Black 137.4 64 34

Of particular note in this grouping ....
-EVGA card was the poorest performer, worse than FE
-The top for close in performance
-Asus stood out with 65C / MSI stood out with 30 dbA
-Between those two ... what would you rather have 5C cooler of 7 dbA ? ... 7 dbA is 1.62 times as loud.
-With the Zotac, you're trading 1C for 6 dbA, 1.52 times as loud

Looking at the 2080 .....
MSI Gaming X Trio 180.2 fps 71 C 36 dbA
Zotac AMP Extreme 177.7 84 36
Asus Strix OC 175.0 63 36
Gigabyte Gaming OC 172.2 69 34
Palit Gaming Pro OC 172.7 73 36
Palit Super Jet Stream 171.7 71 34
Nvidia FE 169.0 74 35

This is discussed further in Item 3 ... but at the top end squeezing that extra bit of performance out if a card results in exponential increased in power. Drop the OC a lil bit and the impact on power consumption is much larger. If ya want that top spot, it's gonna cost ya ... only one AIB 3080 beat the FE card and that was MSI

We often do custom painting and custom loop builds which involve removing the coolers and painting the shrouds or adding water blocks. We bench the cards to make sure they meet advertised specs outta the box in case it's a bad card and needs to be sent back, recording temps and voltages under stress testing. It's tested again after painting and reassembly, and we only use premium TIMs after an Indigo Xtreme surface cleaning . Same with water blocks; if they are removed and blocks / cards sold separately, we reinstall the cooler and test again. We have never seen more than a 1.0 C swing.

2. If you look at any of those 80 way TIM tests, the max difference is like 2-3 C from best to worst ... even changing from peanut butter wouldn't drop 20C ... and yes, it's been done

You got your peanutbutter on my cpu! You got your cpu in my peanut

So you popped off you heat sink without realizing you are out of your favorite thermal compound. Before you go running around screaming and hoping to god that your local late night electronics store is still open sit down and be stationary and quiet while you panic. You have a few options, wait u...

forums.overclockersclub.com

3. There's always a balance between sound and temps ...when setting up default sound curves, different manufacturers will have different targets for their base sound curve. many vendors also set up two BIOSs , one with lower temps / more noise and another with lower / noise / higher temps. Noise has a variable impact .... it's matter of how much; temperature does not ... if you are 65 or 75C, who cares ? , there's no impact on performance or the user experience. That does not hold for noise ... the impact on the user is different whether it is 28, 33, 35, 38 or 42.

Look at the 3090

ASUS RTX 3090 STRIX OC 68°C 42 dBA 2175 RPM
MSI RTX 3090 Gaming X Trio (Low Temp BIOS) 67°C 37 dBA 2016 RPM
Gigabyte RTX 3090 Eagle OC 71°C 36 dBA 1809 RPM
ZOTAC RTX 3090 Trinity 76°C 37 dBA 1700 RPM

MSI is cooler than all others and 2nd in sound, 2 have a quiet mode ... more rpm = more noise and lower temps

ASUS RTX 3090 STRIX OC (Quiet BIOS) 75°C 34 dBA 1615 RPM
MSI RTX 3090 Gaming X Trio 77°C 33 dBA 1543 RPM

4. TPU Hit 72C w/ max load + OC on your card... you are 20C below that ? ... seems odd no ? .... if WoW test is so grueling ? How is it that when TPU, tested the same card, they couldn't break 72

5. At full fan speed (1225 rpm) , under Furmark, my GPUs are at 39C ...I can hear them tho. When they drop below 850 rpm, hear nothing and GPU temps are 44C ... they never break 625 rpm and 42C in gaming, .... why go cooler if it means more noise.

6. If you look at any of those 80 way TIM tests, the max difference is like 2-3 C from best to worst ... even peanut butter doesn't add 20C

You got your peanutbutter on my cpu! You got your cpu in my peanut

So you popped off you heat sink without realizing you are out of your favorite thermal compound. Before you go running around screaming and hoping to god that your local late night electronics store is still open sit down and be stationary and quiet while you panic. You have a few options, wait u...

forums.overclockersclub.com