UNSOLVED
Aurora R11, RTX 3080, my review
Got the system 3 days ago.
First impressions; with 1 case fan, 2 case fans if you want to call the top radiator fan on liquid cooling for the CPU a case fan is purely a joke.
Next up, the ventilation for the RTX for the air to be pushed out of the card, though a metal mesh, then through plastic cut out slats, that doesn't even allow for airflow of the entire length of the video card is a joke.
Putting two fans on this 10.5" card is joke. I don't know what Dell was thinking (its obviously they weren't). These cards require minimum 3 fans.
If all you want to do with your Dell is do some lite gaming, buy it, as for some reason, the card manages to run cool enough not to trigger 100% GPU fan. I've only tried a few games on this, like Far Cry new dawn and they seem to run fine without affecting the 100% death fan effect.
My problem comes in, that if I want to do anything more with my computer like say, some distributed computing, like and anything else, than the fans instantly rev to 100%. The noise is unbearable, unless you have the box in another room. Even with the side cover off, for much better air distribution, and the GPU temps in the low 60's, the fans never ramp down, it's 30% to 100%. That's the options you have.
I tried to install a bunch of fan controllers, and the only one that worked was MSI afterburner, but, even then when setting the fan curve to come on to start ramping up at 55C, I notice it thermal throttling, for some reason even when at 58C. I've come to the conclusion, that this really short card with only 2 fans installed is factory set to ramp up fans when it's being loaded up to 100%. It can only handle so much heat since with no large backplate to soak up and dissipate the heat, the only option is to run the fans at max speed or slow itself down by thermal throttling to not fry itself. Such a nightmare. If you are thinking of using your card for research or playing future games that will load up a 3080 in the future, running the GPU at 100% isn't working out for me.
So just keep this machine for lite gaming or return it? I bought this for one reason only, for the RTX 3080. I really didn't know it was going to be a custom Dell OEM card that will probably die an early death because it was never designed to be put in such a compact case. It was also not designed for a 2 fan setup. Bad move. In my opinion, I can not recommend this combination to anyone. Unless you need a really small footprint, and aren't going to push your computer to the max.
8193
1
Posted January 10th, 2021 02:00
Return it and build one you will thank me 100000x over.
8108
0
Posted January 15th, 2021 14:00
I have approx. 10% throttling on my Dell Aurora R10 with RTX 3080 running at a temperature of 46 C
7476
1
Posted February 25th, 2021 21:00
After some research, the issue is not the CPU temps, but the memory temps. You can now see them in HwInfo or GPU-Z. As soon as they hit 110c, the fans revs up to 100% to protect itself. The 3080 Founders Edition has the same issue. A way to fix it would be to had slightly bigger thermal pads on the memory so the backplate can take more of its heat.
7428
1
Posted February 26th, 2021 11:00
I have applied custom fan curves and split them up between both GPU fans, giving each fan it's own curve and speed. I run fan #2 500 RPM higher than fan #1.
My Tjunction has never gotten higher than 92 Celsius when the card is fully loaded up for an extended period of time, and normally hoovers around 70-75 during gaming sessions, with the occasional increase to 85 ish due to extended periods of load.
I do think the default fan curve is not sufficient for proper cooling, and you might want to try playing around with that before taking a +$1,000 dollar card apart to try and improve the thermal properties.
7427
0
Posted February 26th, 2021 11:00
I am interested, what's the Speed in Mhz when it has dropped 10%?
7413
0
Posted February 26th, 2021 12:00
If your a serious gamer then I would highly recommend you send it back ASAP. I tried running iracing at max setting and had great fps (280). But that only lasted about 2 weeks. It started shutting down after 30 mins of play and increased to where it would only last 2 mins. My solution while I wait for a new card from Dell it to have the side cover off and a household fan blasting at high speed directly at the GPU. With settings at lowest setting I can game, but the heat still gets to 72c. My guess is the new card wont be any different and wont last either. I have ordered 3 corsair ml pro fans to install. One for pull at the bottom and two stacked up top in a pull push fashion. Im not holding my breath that it will work though. And, after the 30 day cut off it will be like moving heaven and earth to get a return authorized.
4041
2
Posted February 26th, 2021 12:00
72 should not shut the card down. In fact, it should down clock when it get's hot, not shut down. But 72 is honestly not bad temperature to have during an intensive gaming session.
Are you sure there's no other reason why it's doing that? Mine does not exhibit that behavior.
I have been playing for 3 months on my R10 RTX3080, with zero issues. It has never crashed or shut down unexpectedly, or overheated or anything like that.
Hopefully your replacement card provides a better experience.
2691
2
Posted March 1st, 2021 00:00
+1
"I have been playing for 3 months on my R10 RTX3080, with zero issues. It has never crashed or shut down unexpectedly, or overheated or anything like that."
6036
0
Posted March 1st, 2021 05:00
The fan curve won't fix the overheating memory, or barely. It seems like the memory chips is not even connected to the fin stack.if you are doing very memory intensive tasks, like the person in the thread started, or mining, it's not gonna be enough. Just playing games, nothing crazy, the memory is around 100c, which is 5 degrees higher than the recommended temps, and 10 degrees under throttling. At this temp, the core is at a very confortable 50-60c. There is an issue with the design.
LDDiamond
1 Rookie
•
24 Posts
8250
0
Posted January 8th, 2021 07:00
Your gpu shouldn't be thermal throttling until it starts hitting low 80s ... I think 83 degrees but could be wrong on the exact number.
Now if your gpu is showing 100% usage in task manager... that is what you want. It shows that the gpu is being fully Utilized before the cpu.
That is actually ideal . Now if your card throttling as in downclocking a lot at temps under the 80s, you have an issue with your gpu