AMD releases a 4k capable card that is designed to run at 95C in 2013 and everybody loses their shit.
Nvidia releases a card that is incapable of a sync compute and unintentionally runs at 104c and nobody gives a shit.
AMD does fuck up a lot, but it seems that they are always under the microscope when other companies do it just as much.
>>54633682
>AMD does fuck up a lot, but it seems that they are always under the microscope when other companies do it just as much.
Huh? Even the press is hiting out at Nvidia's bullshit marketing
https://youtu.be/bNCfn4y8dBw
Already sold my fury x on reddit and going to replace it with the 1080. The dumbass that bought it said some bullshit about Nvidia being evil and I just said haha yeah. AMD supporters are so fucking stupid. Can't wait to finally have functional drivers.
>>54633794
>drivers
Holy fuck dont get me started on AMD drivers
crimson is BROKEN
it doesnt work on any of my laptops either.
Desktop drivers are just as bad beta's unusable (so that doom 4 out of the picture)
Last stable driver was released 3 fucking months ago, are they fucking serious?
>>54633794
>>54633825
Drivers work fine on my r7 260x.
I don't buy shit on launch day though, fucking waste of money.
>>54633825
I cannot get 60hz on my 4k monitor because my Intel drivers don't support custom resolutions.
The 860M GPUs were incompatible with directx for over 3 months.
Tell me how nvidia and Intel's drivers are perfect :^)
>>54633859
>I don't buy shit on launch day though, fucking waste of money.
If you are in the EU, chances are that you will save money by buying it on launch day.
>>54634061
Ausfag though, not eurofag.
I can buy a game a year after it was released much cheaper, get better drivers, and better performance thanks to updates.
Both AMD and Nvidia have had shit launch drivers for me. Went from hd 5770, to a gtx 670, to an r7 260x, so I've experienced both. Extremely disappointed by nvidia drivers on w10.
>>54633859
thats how you know they shit posting. they probably both own 970's
>>54634326
Honestly, both companies are mostly fine. On w10 amd drivers did work better though, but that might have improved since then.
I've used both and haven't had any major issues with either of them.
>>54634345
pick a side or go back to rebb!t you fagget.
>>54634111
>I can buy a game a year after it was released much cheaper, get better drivers, and better performance thanks to updates.
WHY would you want to wait a year?
If the game had any PvP MP it would be almost dead
Sorry but your fucking retarded
>>54634397
Not every game is call of duty retard. If multiplayer lasts a year it's shit anyway
>>54633794
>>54633825
>50% troll
>50% shill
>100% cancer
>>54634397
Bf4 still has great multi-player, so does gta v.
>>54634371
I pick whichever is best value for my budget. Sides are for pussies who desperately cling to one brand and refuse to do anything meaningful with their lives, instead choosing to shill on /g/
I've got a 980 Ti, but a freesync monitor. I'm looking forward to seeing what AMD releases, and I have no issues switching if it's good.
>>54633682
>unintentionally runs at 104c
Where are you getting that from?
>>54634509
AMD cant even beat the 980ti oc let alone their own fucking cards from 2013.
Polaris better be fucking good
>>54634548
What?
>>54633911
4k monitors aren't standardized, the connector to the panel may vary in quality.. I have 4:4:4 at 60hz using hdmi 2.0 but 4:2:0 has a more stable refresh rate
Also you gotta consider the clock performance that needs to stabilize 60fps.. 4k gaming laptop is gimmicky, we've yet to see the 1080 in laptops for christmas 2016 when the price is slashed in half
>>54633682
There's a play on words on VR tech and some new renderingg tech for VR, benchmarks aren't 2x titan x, clock for clock perf is lower than maxwell, finfet arch is supposed to generate less heat now the central point is suddenly the effective power consumption (as if that were a variable metric at full load)
If you look at amds tech its way different and controversial in a better way, nvidia seems to be scrambling to tie up a solution, the financial objectives are met and the data center driverless cars hype is real, the fact that hbm2 is missing part results are skewed
The perf is defined and tech follows, not the other way round, some key metrics are obviously not met (high clock, high temp, optional cooler), crammed die surface vs double from the claimed counterpart with higher load
>>54634589
>That image
Nvidia didn't want to use GDDR5 for 2 years, because it was more expensive, why would anyone want that?
>>54633723
There's atleast few people doing the proper review work. Most others are giving this 10/10 for no fucking reason. Are they retarded or just bought/paid shills?
>>54634652
>Are they retarded or just bought/paid shills?
Shills. They can't slam Nvidia too much or they'll be shut off from future events/opportunities to get new tech to review. If they can't get access to Nvidia's stuff then they can't get page hits.
This is the same kind of shit that happens in the sporting and video game "press".
>>54633794
>>54633825
this is how I know you're both liars.
AMD's drivers have been great for at least a year now and Nvidia has been tanking it hard.
Grow up.
>>54634111
>Extremely disappointed by nvidia drivers on w10.
They ain't better on the AMD side (I'm team red at the moment).
What I meant by my post is that graphics cards increased in price after their launch since the Euro (and mostly my nation's currency) dropped in value towards the USD, and it looks like there is gonna be another drop now like we saw back in 2014.
The GTX 970 was released in september 2014, just before the currency started to sky dive. My local prices never dropped on the 970 since release and instead they increased by 10-20€ and has since kept a steady price point. The only sales I've seen have been through different manufacturer promotions or limited campaigns such as "first ten buyers get 25% off", but the prices are pretty much the same today as they were in the end of 2014 despite the 1070/1080 being around the corner.
US consumers and other countries with stable currencies and decent import laws aren't affected as much, but some are.
>>54633682
1080 does async tho
>>54633723
>So you are charging 100$ for what?
>Design and... and... premium materials.
>What premium materials?
>Eh... Uhm... metal, aluminum? And a backplate.
Jesus Christ, I hate Nvidia and Nvidiots.
>>54634548
Dude 390 can beat 980 Ti now.
>>54639022
amd does the same shit. so you're going to pay an extra $100 for what a shorter card and less material?
>>54639374
You must be a very smart person.
>>54634589
The figures on that graph are largely inaccurate.
HBM2 draws less power than first gen HBM. Higher density HBM2 stacks draw less power. The DRAM dies and base logic are all fabbed on a smaller node.
https://www.skhynix.com/eng/product/dramHBM.jsp
Nvidia needs to be taken to task by the FTC for constantly holding press events where they just lie on stage repeatedly.
>>54633723
Why does /g/ always favor the worse cheaper alternative for gaming?
They always side with AMD and they always recommend linux.
>>54633682
No amount of whimpering is going to change the fact that it's the fastest single gpu by far, and even shits on amd cards in their most heavily favored games.
Eat shit AMD fags
>falling for nvidias marketing techniques
>>54640037
Basement dwelling poorfags and underdog mentality.
>>54639230
>390 beating 980Ti
I am all for AMD competing and forcing Nvidia to play fair, but amdshills are the worse. Stop cherrypicking AoS benchmarks. Who here even plays that game.
>>54633682
AMD
>Our engineers developed this new tech for 5 years, hope you enjoy it
NVIDIA
>It was a miracle from god
>>54640301
I have a feeling that something other than the framerates will blister.
>>54640066
Is the bottom axis FPS? What a fucking horrible graphic....
>>54636335
What do you mean AMD has shit Win10 drivers?
The only issues I've seen about AMD Win10 driver was when the new Crimson was released and that had some issues with fan controller and third party software. Meanwhile, nvidia crashes and burns with daily reports of unstable drivers
>>54635553
Yeah my laptop has been fucking up multi monitors with each driver update on my 960m, had to roll back to get it stable.
>>54634548
>amd can't beat old amd cards
What?
>>54634061
europe is dying
>>54640703
They have stabilized more now but there are still a couple of issues left.
- Mantle feature still crashes regularly when activated
- Freesync can utterly fuck up your graphics at times, had to turn it off completely when crimson was launched because it refused to work under any circumstance. Magically started to work a patch later.
- Some games still refuses to work no matter if I reinstall windows and all my drivers (beta or not).
I can't speak for every AMD user or every nvidia user for that part either. But based on my own experiences I've seen AMD behave odd on functioning systems and nvidia behave really badly on systems with no or low maintenance.
I don't wanna give nvidia bad or good credit, but I think most of their issues is due to their users being, simply expressed, idiots when it comes to computers.
>>54633682
You don't judge the winners OP, as in war.
>>54633682
I doubt they called this card a GTX 1080
>>54633682
It's only 104c because of their janky ass OEM cooler. Once the board partners get here we'll see what this baby can do.
>>54633682
>runs at 104c
Running Furmark. It doesn't reach those temps when running a game.
>>54633723
The faces of disappointment when the get called on their bullshit is gold
>>54633825
Dude, I played Doom pre-alpha and beta for weeks with my 380, I ran it at locked 60FPS on ultra with 1440p, how the fuck are the drivers broken? Please explain?
>>54634652
They flew them to Texas for an all expenses paid vacation on some ranch or some shit.
It's ok guys AMD has new drivers today to fix DOOM. Sorry for the inconvenience again.
>>54640573
AoS is the only game that supports async compute and has a good implementation of dx12, it's not biased to favor AMD as nvidiots think (the good performance of gtx1080 proves my point), so it's a good metric to evaluate how future dx12 games will run on actual hardware. Now that nvidia released an architecture somewhat capable of doing so be assured that they will allow and push developers to implement async compute in future games, in order to gain a few more fps on pascal and also cuck maxwell/kepler users.
So yes, in a few months we will have 390s cucking 980tis and 7970s (kek) raping 970s.
It's the same story all over again
>>54643022
I love that the amd no drivers meme has ended. They're offering much better support for their current arch.
>>54642920
My card doent go over 75C in Furmark with the fan locked to 850rpm. The power draw for the 1080 is only ~30w higher than what I have.
>>54643709
What card do you have? I think Furmark makes Maxwell go nutty or something and draw a shit ton of power.
>>54643652
>I love that the amd no drivers meme has ended.
I love how I keep hearing this year after year and all the kiddies think they know shit because they lived through two seconds of AMD being shit at drivers but now they're suddenly so much better.
News flash: people have been complaining about ATi drivers for over 15 years now, probably over 20, and every few years all you retards think that their driver problems are in the past: they're not. ATi/AMD still sucks shit at drivers.
(Ironically posted with a Fury Nano, which I am pretty happy with, but I am under no delusion that it has good drivers.)
>>54643789
Just know it's modified to hit pci+6 pin overdraw limits (about 162w on my machine) strapped at an accelero tt2, no chips on the ram or vrms
That's besides the point. Hitting silicon thermal limits tells me Nvidia only designed their cooler to handle typical scenarios, and not full load uses. Pretty shameful for a "premium" device.
>>54643886
Not sure what you're on about. It's better than it has ever been and right now that's even better than Nvidia which is currently underwhelming as of late.
>>54634529
His ass. Pascal is no hotter than Maxwell or Kepler which were both very good with thermals, especially compared to Fermi. Everyone's just making a meme out of the stock fan speeds which don't ramp up until the card hits 83 degrees just like the past 3 generations of cards.
>>54633682
104° is the max operating temperature.
>>54644190
Sure, at 1000 hours lifespan to the major power components.
I dont know how good you are at math tripnigger, but that's 41 days of operation before component failure.
>>54644134
This. I gave away my gtx 650 after a bunch of w10 issues, upgraded to an r7 260x instead and I'm happier.
Nvidia drivers were pretty good when I first got the 650, but now it's just meh
>>54640066
82C housefires
>>54633825
I'm an AMD fanboy but I have to admit crimson is cancer.
There's a weird underclocking bug that they STILL have not fixed. I'm using old drivers for this reason.
>>54642996
These are the 12 year olds who shitposts on here. Everyone knows competition is good for the consumer, unless you're a blind sheep.
>>54643009
Yup the guy from PC Gamer skipped out on his kid's birthday to go to the press event
>>54644289
With what card? 390 here and I've had zero issues with drivers over the past year or so.
>>54644289
Fury X - no issues
>>54644289
R7 260x here. Zero issues.
What card?
>>54637012
it doesn't, it still sucks ass in DX12, it tries to emulate it however in drivers
>>54644289
390x and windows 8.1, no issues at all
>>54644289
LMAO they still haven't fixed it? It's almost been a year.
In the meantime use clock blocker. It'll increase your power consumption a fuck ton though. That's what you get for buying amd. Ghetto fixes everywhere.
>>54633682
Because Nvidia has good marketing.
AMD is killing the GPU marketshare on products overall after intel since they have the consoles. That can give them enough money till 2018 to make their stacked GPUs.
After that, it's ogre for Nvidia.
>>54644767
Still waiting for the new line up of amd gpus, im currently using a 380 and want to upgrade, but I want to see which gpu is better espicially with new dx12 games.
>>54644816
I bought an R9 390x to hold me over till the Radeon Pro Duo came out, it disappointed me so I'm waiting till the series after the 400 (might be 500 series since they're rebranding the 200 series again for laptops) to buy the high end off those, may even buy a Pro Duo in between that time.
Still AMD with their scalable GPU plans is going to absolutely wreck Nvidia.
>tfw AMD has just been sandbagging for yeras
>>54642996
DOOM ALPHA
AMD - Great
Nvidia - OK
BETA
AMD - Good
Nvidia - Good
Release
AMD - Ok-ish
Nvidia - Better then good, but not great.
Don't ask me how the fuck that happened, but it did.
>>54645026
I can tell you why
>GAMEWORKS
>>54645026
Nvidia actually work on improving their drivers before releases of major games.
AMD are always a step behind.
>>54645057
Well when your entire R&D consists of one programmer, Rumpelstiltskin spinning straw into copper wiring, and a goat, you'd be behind too.
>>54645057
yea, it's kinda sad, but the reason they came up with Mantle (which in turn transformed into DX12 and Vulkan) was so that they didn't have to tailor software for software, the games would have direct access to the hardware and thus you wouldn't need to release as much drivers.
That's what they wanted but you know, DOOM isn't in Vulkan, it's in OGL 4.5
>>54645026
Kyle is always right.
>>54644765
>installing an application to clock block you
>>54644289
really? Works great with xFire 7950s
>>54645057
I understand how Nvidia got better, but why did AMD performance get worse? They don't add a bunch of new shit into a game before release. They optimize performance.
>>54639544
read the vertical axis label, dummy.
it's about total memory power under max load for a system with a reasonable bus width (e.g., 4096b for HBM).
HBM chops back IO power dramatically, but it does nothing to make DRAM cell accesses any more frugal.
> Row Power = reading a DRAM array row into sense amps and latches
> Column Power = selecting and shunting out a subset of bits in the active row
when you're pushing 1 TB/s or whatever, that's a LOT of internal DRAM accesses.
>>54645306
It didn't got worse, the beta ran at Medium settings capped at 60FPS, literally bruteforced its way trough the beta, once settings started to get heavy the AMD shitty OpenGL implementation became more obvious.
AMD is know for having bad opengl implementation to the point where their own extensions don't even work.
>>54645306
Not Vulkan. AMD was banking hard on the Vulkan release that ID Software promised but didn't deliver since it's still OGL 4.5
>>54645310
>I have literally no idea what I'm talking about
The DRAM dies are lower voltage, and lower current, you shit eating child. The entire stack of first gen HBM draws about 3.5w~ under load. Higher density HBM2 modules draw less power than this.
Power consumption is not increasing at all, no matter how you try to cut it.
>>54633723
did someone cough "bullshit" at 4:15?
>>54645550
What are memory controllers.
>>54633682
amd comes up with new technology, nvidia/intel follows behind them blatantly copies them, and sucks ign/kotaku/etc dick to get developers on board their stupid rip offs. the sooner you realise this the sooner you can forget about the whole issue and just buy intel/nvidia if you can afford it and amd if you cannot
>>54645026
It's like you can watch as Id is paid off by Nvidia in real-time.
>>54633723
Are they drunk ?
