For fuck sake
Why can't anything ever be good?
What do we wait for now?
Meanwhile I'm sitting here with my GTS250
>>54597215
We pray that Polaris and Vega can deliver us from Nvidia's price gouging and incremental performance updates. If they can't, the GPU market will literally become the CPU market, with Nvidia pulling the exact same Jewish tricks as Intel. Pascal is merely a test run. It can still be stopped, but only if AMD don't fuck up. Again.
What's the problem if it can handle the heat? My 6700k ran at 100 degrees for an hour (fucking gigabyte fan controller, piece of shit) and didn't have much of a problem with that.
If you want a high end GPU that runs at like 60C, you're going to be waiting a long time.
>>54597215
>torture loop
what is that?
>>54598307
Fuck off retard
>>54597215
nvidia fusion reactor technology
>>54598656
An hour isn't an issue. Ten hours, a hundred hours, a thousand hours and it starts to become one. Components can survive temperatures like that short term, but not long term.
>>54598685
>If you want a high end GPU that runs at like 60C, you're going to be waiting a long time
>>54597215
>filename
>>54598749
top kek
>>54597215
>foundry edition
>>54597215
Lol senpai
>>54598841
Reference buzzword
>>54598849
>literally measuring a "hot spot" on the PCIe bracket
Tom's Hardware, ladies and gentlemen!
nvidia knows how important memes are to us
they really care :3
>>54598843
Nvidiafags confirmed for kinophiles
>>54598901
>kinophiles
I'm not googling that
You do realize everything has an operating range. If it hits 80 C who cares as long as it is built to deal with that much heat. They won't realize something that breaks because it overheats.
>>54597215
How the fuck did they manage this with 16nm? Oh thats right they overclocked the 980ti
again
>>54598888
>gpu 104degrees
fuck knows how how the actual card was im guessing in excess of 110+
>>54598923
>to 2.1 ghz
I shiggy
>Reference cooler
So, no real surprises? I bet with aftermarket coolers it'll be 70C max
>>54598922
People are writing about it because they showed off a demo running 2114 mhz while being 67 degrees C, what did you expect people were gonna do ?.
Running at 75-85 degrees is not pleasant to have in your room at all, i got enough of that running 7970 CF.
>>54598922
They won't realize something that fucks up realtime applications because the last 0.5GB are not as easily accessed as the rest of the grafix memory.
>>54599001
They will be around the same as current 980ti custom coolers, because its clearly as hot as 980ti stock cooler was.
>>54599003
>running 2114 mhz while being 67 degrees C
It can do that, when the fan is set to 100% at all times.
>Nvidia releases a NEW CARD
>its just a 980ti thats been overclocked
>probably not even 16nm
>still no async compute
>new, 'faster' memory even though its the same speed as the 980ti
Gee, this 'new' card sure is amazing.
>>54598756
The R9 Fury is no longer considered high end now that the 1080 is about to be released to the public.
>>54598976
clock for clock its also weaker as well and still has no real async
>>54599066
>its just a 980ti thats been overclocked
>to 2.1 ghz
I shiggy
>>54599066
>cant even run at 61 degrees at 2.1ghz without thermal throttling or power throttling
This card is a turd
>104 degrees
>>54599066
Still faster than everything by AMD and probably 1070 is as well, until 2017 minimum.
>>54599106
>1070
Why would anyone want to buy a GTX980 with a different name?
>>54599116
it's probably faster than gtx 980 and could be good card for very small itx builds or for lower end builds.
>>54599116
Because "muh nvidia"
>>54599142
>it's probably faster than gtx 980
Well duh, its the same card but overclocked.
>>54597215
Why buy the high end models anyway? So you get the steam achievement for running the game on ultra 60fps?
I usually go for the mid line, new techs.
Waiting for the 1060. Running a 660 right now.
>>54599116
because:
https://www.youtube.com/watch?v=O7fA_JC_R5s
Nvidia will gimp the 980 and 980Ti
>>54599165
Well duh, Nvidia doesn't actually make new video cards anymore, they just rebrand. The 1080 is a OC'd and rebranded 980ti, which is in turn an OC'd and rebranded 780ti. They just gimp their old cards so it looks like they're improving but they're just sitting at the same place.
>>54599044
No it can't holy shit.
>>54599272
Then how the fuck did they get those numbers?
>>54599284
They probably had it in fucking refrigerator-tier room. I mean there is a lot of ways to just cool the card. I'm not saying it can't get those temps with aftermarket cooler but just read the benchmarks and temp tests, no one has been able to mantain 60-70 while under load with the stock cooler at 100%.
>Ignoring the benchmarks that destroy AMD.
>>54597313
oh cool they even come with hotspot
>>54599336
the human eye can only see 30 fps anyways
>>54599066
You must be confused AMDrone. It's AMD that does the rebranding. And Polaris is just another rebrand. Pascal is a brand new architecture.
Do you know why they call it the GTX 1080? Because when you use it you turn 1080 degrees and walk away.
>>54599336
What? I get 70-100 fps at 4k with a 980ti
>>54599336
Good news everybody! New cards are better than old cards!
WHY THE FUCK WOULD ANYTHING NICE HAPPEN?
>>54597215
>What do we wait for now?
AMD to go out of business after the GTX 1070 and 1080 comes out.
>>54599366
Do you know why they call it the GTX 1080 ? Because it runs at 1080 degrees and melts away
>>54599412
fuck these 'benchmarks' are fake as all fuck
>>54599412
[citation needed]
>>54599624
Skip to 4:30
https://www.youtube.com/watch?v=bql4x9YIPZU
>Based Nvidia making these AMDcucks quivering again after releasing the fastest single GPU on the market.
Nvidia's cooler has always been terrible.
>>54599350
>30
Fuck off retard
>>54599780
>taking bait this hard
>>54599339
underrated
>>54599791
>I was only pretending to be retaded
Fuck off
>>54599868
i wasn't the one who said it shitlord
>>54599336
>new card better than old card
holy smokes
>>54599336
New cards are better than old ones!??
Wait for Gigabyte
Then wait for Polaris
Remember not to fall for Nvidia's jewish tricks.
>>54599165
No, because this time it's still maxwell
Optimizations for pascal will carry over like amd GCN, and unlike keplar
>>54597215
Founders Edition? Moar like Furnace Edition am I right? :^)
>>54597215
how much is that in 'merica temps Mikhail?
>>54600357
f=(c*9/5)+32
>>54600390
He'd need an indian to perform that operation. Luckily there are a lot of them in the US now.
>>54597215
>Founder's Edition
Wait for a card that does not have a blower-style cooler if you want lower temps.
>>54599336
new gen after a die shrink beats old gen!!!!
>>54597215
I don't know that much about computer/electrical engineering in particular, but your temperature profile is determined almost entirely by your boundary conditions and this gradient towards the IC edges is irrelevant when the only cooling is on the external convection anyway which extends throughout the card. So stop using a shitty chassis fan, what the fuck's your problem?
>>54597215
So what? I have kept a Xeon running at 80°C/100% load for days and it dinn do nuffin
>>54597215
Is 85 degrees really all that bad?
>>54601290
>>54601318
That's Intel.
This is Nvidia.
>>54599540
Kek best version of that meme I've ever seen.
>>54601337
1337 post :^)
>>54599540
All best memes come from /v/ and /tv/. I swear to god, /g/ is the most retarded board there is I don't know why I come back to it every single hour of my life.
>>54597215
There is literally nothing wrong with 85 c.
You can't make a compelling argument to say otherwise.
>>54599003
67 degrees while running a game in 1080p at 60 fps.
Which is probably exactly what the card can do. It's not really a lie, so much as it is an omittance of the temperatures you can expect in higher resolutiions.
>>54601337
That's what happens when a couple of caps go. Smells horrible
>>54601397
People like you are the reason genocide happens
>video games
Wow
>>54601397
because tech is your passion?
>>54598888
Quads confirm the housefires
Card is pretty good. That said though it is pretty much a jetstream card from nVidia themselves.
At almost $900 here in yuropoor country I can't see anyone buying this anyways.
>>54599356
And yet with 4 years of rebrands and they power hungryness.
Nvidia can barely keep up, using shillworks and reviewers like linushills
>>54601337
>installing liquid smoke for pretend housefire
>>54598833
His face looks like 'totally worth it'
>>54601489
>That's what happens when a couple of caps go. Smells horrible
it wasn't even popping capacitors, it was actually the VRMs catching fire.
don't know if it was low-quality parts or too few phases distributing the load, but there were serious design flaws on the power design of consumer Fermi cards.
>>54598890
Nice catch. These Jewish tricks must stop.
>>54598890
A better question is why it reaches 90 degrees on the pci...
>>54601408
You can when Nvidia promised temperatures in the 60's.
>>54605373
They didn't promise shit, they just showed a tech demo without saying what they did
>>54605473
>show something at a TECHNOLOGY DEMONSTRATION
>We make no guarantees that you'll see any of this performance or temperature and we're giving no legitimate proof that this GPU is based on a new node and altered software to show our GPU running at 2.1GHz, but we know you idiots, sorry, connoisseurs will buy it
>>54597215
We never waiter for NVidia GPU. All the waiting is about AMD polaris and vega
>>54599066
>probably not even 16nm
But it literally is:
http://www.techspot.com/review/1174-nvidia-geforce-gtx-1080/
>>54603591
I think its more likely a simple short circuit on the pci-e power-in.
>>54599780
>dont fall for the bait faggot.
>>54599336
>ignoring the price difference
>>54605373
You will get your temps in the 60s, but they will be just as you saw them: in games running 1080p/60fps. If you want more than that expect higher temperatures.
>>54598922
A 50,000 hour operating life at 50C is a 5000 hour operating life at 85, and the VRMs are almost always hotter than the GPU core
Like most anything in physics (engine rpm and mtbf, IC voltage and power draw, energy required for very high velocities, etc) everything tends to have a quadratic logarithm for power requirements or lifespan.
Few things are linear, like operating frequency and power draw if voltage is static
>>54598781
>>54598797
>>54598816
>>54598833
I keep wanting to root for the red team, then I'm reminded what a bunch of obnoxious faggots AMD fanboys are.
literally who cares graphics cards are obselete now that we have VR
>>54597466
dude at least get a 750ti, they're cheap as balls now and pull way less power than that card, like 60w instead of 150w
>>54607698
all fanboys are faggots. I brought my rig with an AMD GPU to a LAN party once and everyone won't shut up. Just buy what works for the money you're willing to spend
>>54597529
Amd never fucks up, they keep a pace.
They only wait for nvidia to fuck up as seen in the fermi sagas and the KRIPPLER sagas
Bring in the bants.
>>54607786
You're seriously recommending a 750ti for power draw when we know that Polaris11 will outperform a 950 at the very least, using less than that?
>>54604268
Crossfire works thru there.
>>54607900
there are no components in the traces to heat up, and the card pushes far more data to the rest of the system than Xfire does over PCIe
>>54607919
Dude i mean whats the point of this discussion we shouldnt shill any company no matter which it is.
the heat is really just a function of the tdp. how hot things get beyond that is gonna depend on fan speed and heatsink design.
>>54607925
. .
O
>>54597215
Because of the way the universe works, computers must generate a lot of heat. The only way around it is reversible computation, but this requires that you never delete anything.
>overclocks what's basically a die-shrunk Maxwell
>"wow, why is this so hot now?"
Every time
>ITT: /v/ kiddies arguing brand loyalty
There's a containment board for you bunch of retards, go there and stay go
>>54607955
Heat is a function of power drawn over time.
TDP is a cooling recommendation for designers.
To keep temps below a desired point you require x amount of cooling capacity for every y amount of power used. The fraction depends on desired load temps. In my experience a properly set up 250w cooler will keep a 160w sustained draw running at 50-55C with acceptable fan speed, but the chip's area (and thus power density) is a major factor in this. A small chip that maxes out at 160w will probably get hotter than a larger chip pulling the same, using the same cooler.
>>54603269
>the bright orange glow under the pcb
>>54601569
Did the driver menaged to jump out ?
>>54607942
why are we here?
>graphics cards get hot under load
OH REALLY
if only we could whack a big chunk of metal on it and a few fans to disperse that heat.
>>54601337
looks like the cpu cooler shorted it
>>54607739
Get a load of this genius
>>54607839
>Amd never fucks up,
AMD rebranded 3 generations in a row, it was a pretty massive fuck up that lost them over half their gaymer market share (which is the only segment of the GPU market they even have considerable share in).
>>54599233
How is changing the architecture and doing more with less just a rebranding?
>>54598756
>Components can survive temperatures like that short term, but not long term
>(citation needed)
>>54599106
>Keeps comparing Nvidia new gen to AMD old gen
Great work goy
>>54610489
what's the term? metal degradation? higher temps increase it's pace
I think it means anything at all only in workstations when they run these things 100% for 50k hours
>>54610506
>new king meme
>no brainer buy
Seriously, the heck is wrong with modern young men?
>>54610489
Have you ever been to school?
>>54599003
>Running at 75-85 degrees is not pleasant to have in your room at all
Total heat energy is what heats your room up, not temperature. As an example, if you had watercooled your 7970 crossfire cards and they ran at 50C, they would still produce more heat than a single 7970 running at 80C.
>>54610551
>>54610590
I'm not seeing any citations.
>>54610614
son, physics 9th grade
is it too early for you? didn't study it yet?
What's the point of this thread again?
>>54610627
So what's the MTBF of a 28nm silicon GPU running continuously at 85C? What about a VRM?
>>54610671
Amd (always massive disappointment) drones clutching at straws because that's all they have at this point. As if someone buys reference cards. Fucking idiots
>>54607672
Long before your stupid gaming graphics card dies on account of heat, it'll be too old and slow to be worth running at 100% anyway.
If you care that much about longevity, underclock it, or watercool it. Part of the reason that gaming cards are clocked so high is that they aren't expected to be running hot 24/7.
>>54599070
1080 doesn't hold 60C temps.
>>54610679
Surely it's more dangerous if your temperature constantly fluctuates. Not supporting nvidia but maintaining an 80 is better than a 70 which alternates by +- 10.?
>>54601408
>literally nothing wrong
enjoy your throttling cuck
>>54608454
Just to suffer.
>>54598656
basic electronic theory states for every doubling of temperature the lifetime of the component decrease by a factor of 4. More heat means breaky breaky in the middle of your marathon minecraft sessions.
>>54610994
Cringe, the shirt.
>>54611005
cringe, everything about the guy wearing it
>>54611036
Hey, I think this guy has good things going on. There is some girl out there who will have sex with him. Its probably a guy though.
>>54611055
i can definitely see the appeal of the fat-yet-not-fat-nochin body he's rocking, and that unshowered greasy hair matching that of an italian mobster
I cool all my stuff down to at least 37 c.
I figure if the human body likes that then it is good for my cpu and gpu.
>>54599272
Yes, actually, it can.
http://www.tomshardware.com/reviews/nvidia-geforce-gtx-1080-pascal,4572-11.html
>As Nvidia showed off the GeForce GTX 1080 at its press day in Austin, Texas, we kept asking ourselves how the card repeatedly showed up in demos at temperatures under 65 °C. In order to circumvent its thermal limit, we tried setting the fan to a 100% duty cycle, and that ended up being the simple solution.
>In our standard setup with an ambient temperature of 22 °C, we measured 68 to 69 °C. In another room at 20 °C, and using a less demanding Full HD workload, we finally replicated Nvidia's 65 °C demo. The noise that's created isn't bad when you're in a large hall full of journalists and EDM blaring over the speakers. But it's much more apparent in an office.
>>54611195
The link you provided is talking about STOCK CLOCKS.
ie. it can't even keep to STOCK CLOCKS at 65c unless the fan is at 100% duty...
The 2144mhz demo was water cooled.
>>54611325
The fan curve on the 1080 tops at 55%, and it only starts to ramp up when close to 60ºC.
>>54601408
Are you also a macfag?
the reference cooler is absolutely terrible. they need to improve the fan curves too.
custom coolers will be much better.
What will AMD come up with to compete with and when?
>>54611352
Anything above 35% sounds like a vacuum on a blower
>>54597215
>Not using Intel integrated GPU
>>54611468
Tell me about it.
https://www.youtube.com/watch?v=u5YJsMaT_AE
>>54611325
How about you look at the fucking VERY NEXT PARAGRAPH you idiot.
>Focus on the overclocking run's orange line. At the beginning of our test, we registered a remarkable 2126MHz that finally stabilized at 2088MHz. Running at 1920x1080 not only saves 34W of power that isn't converted to heat, but it also facilitates GPU Boost rates above the 2.1GHz level.
>ORANGE LINE
>UNDER 70 DEGREES AT 2126-2088 MHz
Christ I don't know if you're trolling or legitimately retarded.
>>54610934
which website?
>>54611546
its in the image.
>>54611452
next week
>>54611530
>blower fan
>100%
https://www.youtube.com/watch?v=4exmfHshVmY
>>54597215
>2016
>still not using IMS PCB in video cards
fucking idiots
>>54608439
he's fine
everyone's wearing fire retardant suits, he has a helmet on, it wasn't a lot of gasoline, and they had extinguishers on it immediately.
Biggest danger was probably smoke inhalation/suffocation
>>54598756
No.
The problem of "heat stress" occurs during heating and cooling cycles. If you keep it at 100c all the time, it will last longer than if it's subjected to rapid cycling.
>>54598843
>>54598901
what flick is this from?
>>54599356
>what is 600/700 series
>what is 700/800 series
>>54598911
kino as in cinema
>>54598724
It's the cycle of waiting for a new generation of hardware to come out and then waiting for the prices to drop.
>>54601397
I find this with everything i enjoy.
Nihilism is a bitch
>>54601343
:)
>>54612182
Im a fuckwhit but isnt this in some form on watercooled and some air cooler backplates
>>54601569
Topkek, what a fucking retard.
>>54598888
>104 C
>nope this isn't and won't be a problem at all ever
>>54597215
>torture loop temperatures
that doesn't really mean anything
>>54598724
a synthetic benchmark with not particular use
pic is an AMD one in a vain attempt to avoid the chill train going
>>54614677
no. It is a printed circuit board made of metal
>>54615014
Does this tech exist for multilayer PCBs though? And wouldn't this need heatsinks on both sides to work with a graphics card? Its a neat idea but probably not feasible with current standards
>>54614822
that's the VRAM, right? the GPU itself is pretty cool
Is this going to be Fermi all over again?
who /1070/ here
>>54615084
this desu
>>54615014
Whats the advantage of this over traditional backplates?
>>54615254
they keep using the same shitty stock cooler from the late 2000's
fucking why
>>54597215
Mine 2 280x worked fine at 78-82 24/7 back then in the 2013.
>>54615084
>Does this tech exist for multilayer PCBs though?
yes (with 2 layers, or more)
>And wouldn't this need heatsinks on both sides to work with a graphics card?
not really. It will only replace the classic PCB, as an additional cooling method
>>54615357
Wait, now I'm even more lost. How the fuck is the hear supposed to escape the aluminum like this? And for multilayer, I meant 3+ layers, this just appears to be for a two sided PCB.
>>54615307
>Whats the advantage of this over traditional backplates?
it is thermal conductive
>>54610994
And what's the baseline temperature in "basic electronic theory"? 0K? I'm good then.
why do people compare a watercooled card to the 1080?
>>54615510
Crossfire scales poorly, SLI is way better. I am so disappointed in my crossfire experiences that I will never use a twin GPU rig again.
>>54599412
Yeah on minecraft
