>It seems that FireStrike has really pushed these cards to their thermal limits, as the primary CrossFire card reaches 87C and the secondary 82C.
>1288MHz & ALREADY HOUSEFIRES
>AYYMDPOORFAGS ACTUALLY THOUGHT THIS WOULD EVEN OVERCLOCK TO 1500MHz
IT'S OVER, AYYMD IS FINISHED & BANKRUPT
>>55235126
87 degrees isn't bad for a reference card. Especially when it's on the stock "silent" fan curve and pressed up against a second card that's just as hot.
Don't forget the 1080 was 85 degrees in single gpu tests.
HAHAHAHAHA NVIDIA WON
>fan speed 49%
Nice try
>>55235126
>49% fan speed
>44% fan speed
Nice meme.
>>55235126
>reference cards
>fans not even at 50% duty
Looks like fucking nothing. My 8800GT used to break 100C playing Clear Sky and that was "okay".
>>55236241
>>55236289
>>55236382
Look at the temperature curve again, you'll notice the MAX being selected there while not on fan?
>>55235126
I speculated yesterday that a single 480 running a power virus would thermally equalize to about 80-81C, based on the heating rate from that leaked thermal camera test.
Guess I was pretty close.
>>55236788
And if you look closer you'll see that the fans didn't go much faster than at the time of the screenshot - perhaps 55% going by PWM voltage rules.
>>55236963
Doesn't change the fact that this will be buyers remorse of the year.
That is, if there's actually anyone who bought it and not just a pack of wild shills.
>>55236981
And just why do synthetic bench temps matter anyways? Running a game without vsync or frame rate control sees my card top out at no more than 75-76C, but in Furmark the card goes above 82. Actually I just this morning modded the BIOS to a power limit of the connectors + 8% and thats why it hits 75C, let me run furmark again and show you.
>>55235126
now show us your shit build, pajeet
>>55237008
>let me run furmark again and show you
RESTRAIN THIS MAN!
>>55237048
Got busy with something else.
>>55235126
This is crossfire right?
I mean, if the lower card blocking the upper's fan, off course it will be higher temperature.
I think a single card will reach around 70-80 C°
>>55236220
What did they win?
>>55237960
More unpaid shills
>>55237519
>I think a single card will reach around 70-80 C°
With a 22mhz overclock.
There are people seriously shilling that this shit will overclock to 1600Mhz+
>>55235157
>it's okay when AMD does it
>>55238690
Nvidia did it first, did it at higher fan speeds, and did it with a single card.
>>55235126
Notice how the fan speed never reaches over 50%.
You can set temp profiles in the new overclocking tools, as to keep the fans set so the card never goes under 80c or 70c or 60c, and so on.
Basically, the card was set to "silent" mode. Plus it is in CF, so the two cards are pumping heat on each other.
You can set the fan profiles to keep the card at, say, 60c under max load. Of course then it will be much noisier.
This is all cause they use a 5$ heatsink with no heatpipes in them. Once vendors start doing their own non-reference coolers, this won't be an issue anymore.
>>55236382
>Looks like fucking nothing. My 8800GT used to break 100C playing Clear Sky and that was "okay".
It's okay when Nvidia does it. My 8600gt used to IDLE at 62C.
>>55235126
Amd cards may run a little hotter but atleast they don't literally housefire on you. I've had 2xGTX 650 and a 9500 burn on me. Also I don't know why temps matter. Different chips behave differently to heat. I think I'll be getting a RX 480 simply because of the work on the newer Linux AMDGPU and AMUGPU-PRO drivers. Linux support on Amd is in a fantastic shape with UVD, VCE, DPM, Vulkan ported over
>>55238667
1600MHz on GCN is obviously nonsense, but temps on a reference card at half fan speed tell us almost nothing about the overclocking ability of the card. The tiny thing in pic related is the entire heatsink for the RX 480 so if anything 87C without the fan on blast is a promising result.
>>55238810
Heatsink relative to the whole card.
