Hey there!
So, has anyone used a RX 580 Powercolor red dragon v2 (8gb) so far? I read it gets super loud, is this true?
I was thinking to get this card since it is on discount with limited stock for 150 euros at my local store.
PS: I hate noise computers, would you still recommend it?
Just replace stock tim with a pea sized of AS5 in the middle. The stock tim is of abhorrent quality that they glob on to the point where it actually starts to insulate some heat. Noise problems shouldn't be an issue then.
ah, i see. i did not think about changing the thermal compound.
Have you experienced noticeable change?
Not him but I went from 77C @ 60-80% fan to 73C @ 20-40% fan which definitely helped reduce noise. I haven't even tried undervolting yet, all settings at stock.
on an rx 570 btw
Sounds good. Thank you for the input.
I own one. It gets pretty damn loud. Not the loudest cooler I've ever heard, but we're talking 2500RPM+ on the fans under load.
Completely wrong. I've repasted mine with Thermal Grizzly and it did very little for noise. Maybe shaved a couple of hundred RPM off the fan speed, but it's still very loud. Ultimately, even the best thermal paste on the market can't compensate for a small, not particularly great heatsink.
AMD cards are nothing but shitty toasters. Go buy a real card. One made by NVIDIA.
Did you clean BOTH the die and heatsink contact plate thougroughly with rubbing alcohol and then apply a small pea-sized amount on the center of the die?
Nvidia has nothing to compete in the mid-range. gtx 1060 costs more, has the same performance, AND less vRAM
>Mid-range.
People like you deserve to die.
Don't use the pea method on the die! Only use it on chips with a heatspreader.
On dies you have to spread the TIM.
No you don't and doing so increases air bubbles which prevent heat transfer. Unless your GPU die is the shape of a hotdog (do line method for that) a pea sized amount in the center is more than enough and will spread evenly once you screw down the heatsink.
back to /lgbt/
depending on the viscosity of the paste that's retarded advice. if just one corner of the silicon isn't covered by paste it's gonna degrade faster than necessary.
I'm and I can assure you the pea sized method which I've used has caused me no problems after a year. As long as most of the GPU surface area is covered by TIM once the heatsink presses down then heat will evenly be distributed.
What you DON'T want is manually spreading the TIM on the die which will create significantly more air bubbles which cause higher temps or smearing a metric ton of TIM on which will ooze outside the GPU and insulate heat on the GPU which will also cause higher temps.
Congrats, you've tested this for a year. I made my experiences with direct die cooling back when CPUs didn't have heatspreaders.
The problem you don't seem to understand is that both silicon and air conduct heat a lot worse then well-spread paste. If one corner of the die overheats the thermal diode wouldn't even know and thermally throttle. Something at the edge of the die like the memory controller is just gonna die. And you can't guarantee contact unless you manually spread the paste.
To see this in action look at overclockers handling cards. Those are of course more extreme conditions, but the same problems will occur for you. Just later. If you want an example look at the card that buildzoid killed, I think it was an older AMD model. Or see Vega cards with hot spots at >100°C because people don't understand why the pea method is even recommended.
The pea-praeching was just a way to make pc building easier for idiots because it literally doesn't matter when there's a heatspreader involved.
Reading about all of this makes me wonder id I should just spend 30€ more and get a Sapphire rx 570 for 180€.
I'm in the same boat as OP but I don't have the budget for a 1060 or an rx 580.
If it's really just bad paste then I wouldn't worry about it. Just google how to repasted a GPU.
Not sure how good the cooler really is though.
I'd just pay the 30€ difference and not worry about disassembly. I'd feel like shit for breaking a 150€ product. I'm not buying it for myself, after all.
If it's not for yourself then yeah, spend a bit more of you're worried about noise. The nitro cards are very quiet.
Can confirm, my Nitro+ Vega 64 is dead silent, I assume Polaris models are similar
Though changing thermal paste is something easy to do and something you have to do every couple of years anyway
Not the 480s sadly, they had way too tiny coolers. The 580s and Vegas are great.
I have their Red Dragon Vega 56, and can vouch that the fans aren't too loud (not quiet, but not loud at all) while keeping the device at a good temp. Assuming that's true of their RX580 as well
I've been wondering about this because a few days ago I spent nearly all day fucking testing temperatures on my GPU. They were getting quite higher after I decided to do a PSU+mobo upgrade. Of course I could have blamed those but to me it seemed as though something else was wrong.
I opened up the thing and noticed that after cleaning it thoroughly, there was a diagonal line that I couldn't clean off and in fact, I doubt it's TIM at all, it doesn't come off. It's almost as though the fucking thing got scratched by the cooler itself as it matches one of the 2 lines that separate the area with copper on the center from the rest of the heatsink (hard to describe).
Anyways, that's not all. When I opened it up, even though it covered the corners, there was a c shaped part at the center that wasn't as covered as the rest, on both the cooler and the chip itself. I applied it again, tested temps, took it out and a similar shape was found after I took it out (only I put more paste in there). I am confused about that shape since it doest look like the cooler could do that.
Another thing that worries me is the thermal pads around. I reuse them but frankly they're not the cleanest. I'm wondering if that alone could fuck with temperatures. Reason I'm so bothered is because a week ago I was getting something like 74C at 1400RPM and now I'm getting 75C at nearly 1900RPM, and the fucked up thing is, I'm applying the paste exactly as I did before (thick vertical line at the center, a bit on every corner to make sure it covers the whole area).
you shhould've spread it, now expect cracks in the die. Proof: GamerNexus