RX 480 Shuts Down Motherboards
...
As someone that doesn't know a lot about video card cooling, is this bad?
No. It's not a cooling issue either, it's a power one. A weird one at that. Basically it's something that'll probably be irrelevant in a few months, and doesn't apply to many people.
If motherboard == older, 480 might have problems from lacking power pins. It's something third parties will clear up, as the guy says in the video.
The heatsink is too small.
>old motherboards stroking out
still terrible but atleast im in the clear
also heard that amd recently passed pci-sig or some shit to ease reddit, is that true?
actually its only because newer mobos act nice to you and predicts such situation - by no means they're obligated to do that - AMD is obligated to conform to PCI-E standard, mobo vendors are not obligated to break standards for sake of AMD because they're lazy fuckers.
in short, do not assume any new mobo will work - they may or may not - if not tested by anyone else you have to make decision yourself and be ready to face the fact that this mobo actually follows PCI-E standards and does not predict situation where you want to pull continuously 30% more power than standards define with spikes rating up to 230% more
Ah I see. Why did AMD jew out so much on the cooling? I guess they had to skimp somewhere to keep it so cheap.
>in short, do not assume any new mobo will work - they may or may not
Yeah this is the problem. Hell even if it works for a while, it might randomly crap out at some point.
>all that shroud
>that itty bitty babby heatsink
the fuck.
basicly anything above asrock as a motherboard is pretty much fine even a fucking north korean motherboard is better than asrock..
the problem with the card is this one
the card has a boost mode going up to 1266mhz and the voltage goes up to 1.3
now when the card goes down to normal mode aka 1080mhz base clock the voltage doesnt go down and it gets stuck on 1.3 instead of 1.1
yet the nvidia paid shills somehow managed to say that the card from a 0.2 increase manage to get a 300% watt increase (294% to be nazi math) while in reality the card draws 80-84 watts from the pcie while it should have getting 75-76
no explosions
no china syndromes
no fucking gtx 480 arsonist edition
no gtx 590 that was drawing 93 watts from the pcie
no nothing
the card is fine as long as you dont overclock it to the hell and back an increase to 1380 is at 190 watt draw from the connector and the pcie reallyyy stretching the limit..
But the OP video shows it shuts down motherboards even when not overclocked
>We need to cut the price of the new 480 down, what should be we do?
>Let's get rid of a little bit of aluminium and copper from each heatsink. That will save literally pennies.
This is definitely the case.
math nazi here
pcie absolute low standar is 75 watts
the 480 as we saw on tom was getting 80 to 84 at some cases from the pcie alone
no i dont know how you manage to find that they card is getting 30% more (97.4 watts from the pcie alone) while by toms number is 11% and that being the spike while the constant was on 80 watts (84watts on pcie was the 300 watts on torture test)
now correct if im wrong but you are literally suggesting that 480 was drawing even more power than gtx590 ever did and not a single card has passed it..
It actually pulls up to 50% beyond spec (from pcper article)
To be fair, once you OC you should have zero expectation for your card to run in the proper PCI specification, it is after all not running stock settings and they can't exactly guarantee you will be able to OC it to certain standards, that's the whole point of setting stock speeds.
Agreed, but it has the same issue at stock
>using cheap mother board
>shit doesn't work
How come my honda doesn't come with shofer?
the op video doesnt show this its shows that the guy is saying the pc was randomly crashing and then he goes and say he had a 400 watt power supply...literally ok lets be honest here amd recommends a 500 psu and he goes and plays with a 400 watts...this is just wrong and he actually dont even care saying it public..
lets clarify something
pcie since it became a standar draws 75watts as the absolute minimum for a card to be powered after the POST (bios boot) its something people need to understand
now since the pcie have a 300 watt limit as a spec there isnt (to my current knowledge) any consumer motherboard out there that has reached even 100watts of constant power delivery only some server custom ones has that feature
now pcper showing that max power of a pcie is 66 it contradicts everything we know that basicly says that the people who created it is a bunch of liars and pcper is the true and only ..
no this never happens and while his 480 results are correct (since we saw already screenshot of the card at 1380 drawing 190 watts combined) having the pcie 9 watts lower to create false results is worse that having kyle saying that the card wont pass 850mhz and raja will coup lisa and take rtg and leave....
at some point either nvidia shills or amd shills we really need to slap those smartasses so hard that they will not dare to make such a intentional shill again
>he op video doesnt show this its shows that the guy is saying the pc was randomly crashing and then he goes and say he had a 400 watt power supply
Nope watch the whole video, he swapped out the power supply and same issue.
He even put in a 980 Ti and it worked fine.
Issue turned out to be the motherboard, it couldn't handle the 480 going out of spec.
Pcper is right though, they're only talking about part of the draw (rails?). Here's the full article:
pcper.com
What is an 'old motherboard'?
well the rails of the connector is totally different than the pcie itself
he cant claim 66 wats for pcie its below to what we know like 7% below this cant happen
now if its the rails of the connector and he "mistakenly"forgot to say well the numbers are correct to what we already have seen from oc 480
so basicly its what he says on the comment section
dont fucking plug it on older systems without pcie 3.0? that is interesting thing to test seems logical since pcie 3.0 can dynamicly manage the load
so a small tl dr
pcie 1/2 rev you risk having system shut down because the slot cant compensate
on pcie 3.0 it draws more power because the slot can compensate "enough" for it
would love to see more tests on pcie 2 tho since there might be people with older mobos that is well known that there is a backwards compartibility on the gpus
It looks like the pcie works on two rails, here they explain it:
"The +12V rail has a tolerance of +/- 8% (11.04V – 12.96V) and a maximum current draw of 5.5A, resulting in peak +12V power draw of 66 watts. The total for both +12V and +3.3V rails is 75.9 watts but noting from footer 4 at the bottom of the graph, the total should never exceed 75 watts, with either rail not extending past their current draw maximums."
One of them is 66W and the total shouldn't add up to more than 75W
Yes but it adds up.
5c per card, 1 thousand cards sold, 50 dollars saved.
Now imagine that on a global scale.
For a company as broke as AMD, every cent counts.
>rx480 grew legs and put itself in the trash
I'm sorry, did you mean a chauffeur?
What are you like 12? This board is for 18+ only kid.
Are you Jewish?
so why he still showcase only the one rail while clearly he measures the combined output of the pcie?
that would kick in the OCP and shut down the machine to protect it from inrushing current (what we see on op video)
the problem is the 12v rails that comes through the 24 connector of the mb is secured on the OCP any typical psu with a 40A ocp point (literally even shit noname tier)can run 40A times 12v=480 watts
on a sli setup the 12V1 should suppoly 150 watts to both cards and the rest they can get it from the PEGS
the main problem is he is lying
youtube.com
look what happen to a single rail setup
He did measure both rails:
"The highest power draw I measured with the RX 480 at stock settings showed 80-85 watts of power draw at over 7A on the +12V line and 4.5-5.0 watts of power draw on the 3.3V line. "
So total wattage was 90W and that's not even overclocked.
8.29x12is almost 100 watts..
7 x 12 is 85
5 x 3.3 is 15
the total wattage was 100+watts+-2-3watts...
unless we dont know how to calculate wattage and voltage again he is either not doing something right or he just slaps bigger numbers together anyway doesnt matter im too bored and i dont have any card to go all technicall on its ass
forums.overclockers.co.uk
sapphire nitro 480 at 330 euros mild oc dvi ports backplate
kinda hoped for a toxic one but oh well
>amd
>broke
Wouldn't an 8pin connector solve the pci-e power draw?
Yep
Yes, but AMD jewed out so hard on this board they didn't even do that.
So when nvidia shits the bed, there are always mandatory shitposts all around but when it's amd/ati threads are civil.
Goes to show the userbase of each team.
>buy low budget video card
>now you have to buy a high budget mobo
Way to miss the point.
Even worse it can damage even a high end mobo over time
Imagine if they cut back on R&D, memory or PCB design?! They'd save millions!
>implying they didn't
why is he using rx480 with that stone age athlon cpu + mobo?
to keep it closer to a real amdpoor setup
So you need a high end Pc to run poo in loo card?
>Logic
Everyone should be undervolting the RX480
Literally no drawbacks
Is your motherboard from from 2007 and before? Then maybe. Other than that I have it OC'ed on a MSI board from 2010 and nothing happens.
>shofer
my god