never buy oem gpu on ebay :D

[Evilsizer]

I got this Dell 2060 super for FAH was going to play some games on it. OMG this fan is so damn loud even at a low 60% controlled in MSI after burner. well this one for sure is going into the FAH folding only box. now the question is going to be how to keep temps low for long term on this card. i never went looking for "OMG THESE TEMPS" on this card till how loud this fan is. this thing is hitting 88c underload, did some stress testing to make sure the GPU was in good order. going to install FAH in a bit to see how things go.

after i did buy it, i started thinking what if the fan was bad. i started to look at some 80mmX15mm fans, mod the shroud a bit. should be able to super glue or silicon it in place. not sure what cfm these fans are rated for at full tilt, i have found some on ebay pushing 36cfm and 34cfm, ironically at the same dba ie 36dba and 34 dba. my computer is also on the desk, a arms reach away.

 

[ssjwizard]

Not surprising. Mass market OEM systems are build for consistency not luxury.
I would imagine a simple fan mod like you described to be just the ticket.
Feels like a flashback making shrouds and water cooling parts out of any random materials I could get my hands on

 

[stompah]

Yank the fans and shroud off then zip tie some quality fans on the HS. Then create a shroud with some blue painters tape and call it good!

Did this to my r9 270 and it worked out perfectly. But that system didn't have a window.

 

[Evilsizer]

well tried 2 games i couldnt really play before, after about a hour blue screen. Looking at the temp logs from GPUz it was hitting 89-90c thats with it dropping clocks down to around 1190mhz to 1240mhz. showing it was pulling about 135watts, fan was full tilt too. play something like smite or other end game no problem. i think this card was messed with(ie new paste or something else) or there is something wrong with it.

going to email the seller..

 

[Evilsizer]

so i setup a return after explaining what is going on. here are the two responses i got:

"Hi, do you have MSI afterburner? The card never ran hot for me I put the power to 0 and the fans on 65% at all times if you don’t have msi afterburner there is a chance you are running the card at 100% power which is way way to many watts even with the fan on 100% it can not handle a max load as any card couldn’t"

this makes no sense to me, if it can not handle it then why would Nvidia design it this way. much less Dell have a fan/heatsink setup like this. then i got this

"

Please watch this video on how to clock and let me know."
now i need to undervolt it? even the benchmark is not showing the right FPS much less the fact there is nothing going on but walking around. someone help me out on exactly is going on with this 2060 stuff? things are just not adding up to me, guess i will plug the card back in to check it out. then put it through its paces agian with fur mark and some tiny tina's wornder land and some BL3. maybe run the RTX Quake 2 demo with RT on..

 

[stompah]

Huh? All cards are plug in, install driver and play. Nothing else needed. The card is bad.

Push forward with eBay for a refund. Otherwise, this seller will just waste your time.

 

[Evilsizer]

got this too
"Gotcha I understand yeah it’s not good on cards to run without setting any clocks especially some of the newer cards temps they go all the way to 110° and run 265w lol crazy!!! If you have any questions or need help let me know."
yea, umm, NO on this one...

 

[EarthDog]

That dude is an idiot... wow.

 

[Zerileous]

I think we saw similar issues last time the GPU market took a dive. A bunch of janked up mining cards on ebay.

 

[stompah]

I think your assessment of this card having been messed with is correct. Are there any warranty stickers still intact?

 

[Evilsizer]

I think we saw similar issues last time the GPU market took a dive. A bunch of janked up mining cards on ebay.

i wonder what that means then for the asus 2060 super that is on its way.

I think your assessment of this card having been messed with is correct. Are there any warranty stickers still intact?

it is a dell oem, there are no stickers i see on it.

Post magically merged: Sep 28, 2022

 

[stompah]

A quick Google search came up with a small handful of pictures just like yours which show no warranty stickers.

If the heatsink was removed they didn't mar the screws.

I'd start talking to eBay about buyer protection.

 

[Zerileous]

I would say like anything else it's a gamble. Hopefully odds are in your favor, having already picked up one lemon. Some miners take care of their equipment, others don't. Mining in and of itself doesn't destroy cards, but extremes of heat for prolonged periods of time can. Often special mining BIOS are flashed to undervolt and increase fan curves and make other tweaks, but these don't work well for gaming. If the seller doesn't flash it back, it's possible to do it yourself, but then you're stuck trying to fix something you just bought. Never fun.

I'm wondering if your dell card doesn't have some modded BIOS. I can guarantee Dell did the bare minimum in terms of TIM, but probably best to adress with eBay rather than try to fix.

 

[Evilsizer]

seller has a 14day return window, got the label to send it back to them.

 

[Evilsizer]

the GB windforce came and did a bit of gaming in QII RTX, it hit 83c max. the fans were at 100%, livable but to loud for me. the thing i noticed is the boost clock was a bit under 2ghz, not by much maybe 30 to 40mhz. i want to figure out how to lower the boost clock, i would be fine with it being around 1.6 to 1.7ghz with good temps.

**edit**
with the clock speed near 2ghz with 100% fan speed it is still hitting 88to90c. if i use msi burner to turn down the clocks at 70% fan its at 85 to 86c. MSI is showing slightly different clocks then GPUz though, MSI is fluctuating alot on the gpu speed. anyone know of a bios or something that tones down the boost speed on the GB card? now this seller doesnt have a return policy so i want to cover my bases now. this card does appear to have the thermal pads replaced on the mosfets for the card. one thing i wont know unless i take it apart is how good the contact is. i know they are suppose to be squished down but depending on the cooler, finding the right thickness for the pads is tricky. for all i know the mosfets are not getting the right contact on the card.

can the 2060's really live around these temps?

 

[magellan]

83c sounds too hot to me (the hotspot on my 1080ti never even gets that hot). You could look up reviews of 2060's and see what kind of load temps they were getting.

What are your idle temps like?

 

[EarthDog]

i want to figure out how to lower the boost clock, i would be fine with it being around 1.6 to 1.7ghz with good temps.

Lower the power limit in the software.

 

[Niku-Sama]

its hot but might not be past critical meltdown.
its interesting this should resurface while i was stressing my new ebay card.

stock i've been running the furry donut for about 20 minutes and its leveled out at 89c which is crazy to think but looking the card up max is listed as 93c
but furry donut is synthetic and theres almost no cpu load too so eh?

i would still return it dude does sound pretty dumb, who knows what was done to the OPs card

 

[Evilsizer]

yea the OEM card i sent back already, the hot spot was over 100c. since there was a return window i did not tear it down to take a look at the cooling. now the GB card is running about that hot too, but no return window. Contact on the thermal pads looked to light, the paste on the gpu was not stock. looked like at some point it was repasted but the paste was dry. on top of that it looked like a mess for it getting spread around, looked like different size pools. LOL. why way i got some play-doh about to put some on the card to get the thickness for the thermal pads. my current debate is to paste with some Artic ceramic or PK3. i am more leaning toward the artic ceramic, i have so many tubes of it. i figured in this case once i get good contact, temps should be fine.

In MSI after burner there was no option to limit power unless there is a feature or box i need to check off. i get the card is 170ish watts but the gpu core should not be running around 90c. i did look around seems the gpu is going to be around 80-84c, once i get the GB tied up. i will need to check agian, but i also want it fairly quiet like the Evga card.