I have jumped ship from AMD to NVIDIA. No regrets. Have

You did good. Fuck AMD and their sinking ship.

Yep. I did too. Welcome.

File: Nha2q9n.png (138KB, 1155x687px) Image search: [Google]

138KB, 1155x687px

Nice. I recently did the exact same. Though you probably could've saved $22.98 by buying at a Best Buy or free shipping from Newegg.

>>55809820
Always do the opposite of what /g/ says.

>>55809820
You fell for it and bought a memecard. Congrats!

>>55809820
why the fuck did you spend $460 on a fucking 70 series card?
what fucking videogame requires you to get that over what you currently have?

>tfw never had to resort to AMD
Feels good.

>>55809820
Yes, good for you. I also joined team green and will never look back. #betterdeadthanred

>>55809820
yeah that card is a beast. i got one but i kinda didn't need it since i always end up playing rocket league

>that price

retard decision

>>55809820
Great decision. AMD will be dead by the end of the year. Nvidia is the future of PC gaming.

>>55809870
This. I shill the rx480 just to trick retards.

>>55809820
Welcome to the winning team my friend.

File: 1467444171390.png (96KB, 748x710px) Image search: [Google]

96KB, 748x710px

>>55809820
>mfw AMD CPU with NVIDIA GPU

>>55810005
the worst of both worlds

>>55809884
>1070
>memecard
>not the most valuable card this gen

>>55810005
Best of both anon, I did the same.
Gtx 960+fx6300

File: 1462148509394.jpg (589KB, 1152x2048px) Image search: [Google]

589KB, 1152x2048px

This entire generation is a meme. Everything is sold out and people are happily spending 50-100 dollars over MSRP like good goyim.

Yes, good goyim. AMD doesnt belong on our good consumerist technology board :^).
Remember to upgrade next year when DX12 hits, GTX 1170 with Async Support* and a great 7.5gb RAM, only $550 MSRP.

>>55809864
>coupling a 1070 with a 1080p60 screen

Nice mem

Yea I bought an EVGA 980ti Classified when they first came out. $750. Boosts up to 1490mhz out of the box and can hit +150mhz to core and +850mhz to memory at +50mv. Pushing a single 32" 1440p panel at 60hz and am very happy with it. The new gtx 10xx series of cards seem HIDEOUS from an aesthetics perspective. I mean I have a non windows R5 case so I really don't care too much about looks but holy shit. Has anyone seen the 1080 classy? Fucking garbage.

I'll probably upgrade to the 1180 or 1080ti if they come out if only for the proper async. Until DX12 is mainstream I'm more than happy crushing any and all DX11 games on ultra 60fps 1440p.

>>55809992
I shill the rx480 because it's good and NVIDIA needs competition

>>55810326
Morally justified, but still faggotry.
The 480 is good but not great.

File: 1060vs480.png (667KB, 3882x2171px) Image search: [Google]

667KB, 3882x2171px

>>55810326
>rx 480
>good

TOP KEK

File: 1070.png (34KB, 731x276px) Image search: [Google]

34KB, 731x276px

>>55810103
>>55810083

i hope it keeps you up at night knowing i spent $430 on a 1070 supporting mein jewish overlords.

>>55809820
I have the SC version, it's a beast

File: mpv-shot0041.jpg (240KB, 1920x1080px) Image search: [Google]

240KB, 1920x1080px

>/g/ is filled to the brim with rabid Nvidia fanboys
>OP thinks he's a special snowflake for buying Nvidia

??

>>55810326
I agree with you that Nvidia (and Intel for that matter) needs competition, but shilling for a company like a retard does no one any favors.

Playing devil's advocate to keep things in perspective? Helpful.
Spouting useless things like "nvidia cannot into async" when there are facts to the otherwise? Not helpful.

>>55810401
>Spouting useless things like "nvidia cannot into async" when there are facts to the otherwise? Not helpful.
But they cant
Nvidia already got sued for their 3.5gb scandal, they're not invincible you know

File: 1337476427042.png (11KB, 211x246px) Image search: [Google]

11KB, 211x246px

>>55809820
>1070
>$459.99

>>55810401
>Spouting useless things like "nvidia cannot into async" when there are facts to the otherwise? Not helpful.
Do you really think that Nvidia can do true async?

>>55810360
You know what they say about a fool and his money
here's your (you), use it well.

>>55810103
>nVidia is expensive

yea yea, maybe in the states, as far as I remember AMD has always been more expensive in eastern europe since 9600GT times. They sell only about 1 AMD gpu for every 10 NVDA gpus sold here (biggest retailer stats). And thats with them having the inferior product. Cant imagine the jewing they would do if they cards were actually better lol.

>>55810401
Nvidia literally took off the hardware required for Async and tried to force it through software so they can get that "muh PPW!!!".

>>55810409
>But they cant
Source?
I'll provide one that states to the otherwise since I doubt you will (can) provide a citation.

> In fact, next to the addition of GDDR5X, I’d consider the changes to work scheduling to be the other great change to the overall Pascal core architecture. With Pascal, NVIDIA has significantly improved their ability to allocate and balance workloads, which in turn has ramifications in several difference scenarios. But for the AnandTech audience the greatest significance is going to be in what it means for work concurrency when using asynchronous compute.
http://www.anandtech.com/show/10325/the-nvidia-geforce-gtx-1080-and-1070-founders-edition-review/9

>>55810471
See >>55810481

>>55810456

thanks for the (you)

i upgraded from a 970 and it was a great purchase.

>>55810425
It's the FTW version with RGB lighting

File: 1467619811325.png (531KB, 893x3075px) Image search: [Google]

531KB, 893x3075px

>>55810481

>>55810508
Your source seems to be a circular argument, ie. not valid.
>If both AMD and nVidia are running the same code then Pascal would either gain a tiny bit or even lose performance.
His conclusion is based upon his expectations of the results of how Nvidia's async HE EXPECTS to function rather than any understanding or explanation of how it actually does function.

Also, just for the same of completeness, he seems to think that parallelism is the only valid form of async compute is parallelism which is complete hogwash. Async involves both parallelism _and_ concurrency. There's a rather good discussion of this in another Anandtech article, most notably with verbose explanations in the comments.
http://www.anandtech.com/show/10486/futuremark-releases-3dmark-time-spy-directx12-benchmark

Finally, that does not support the claim that nvidia cannot do async or that hardware async hardware was removed. At best it tries to pigeonhole async as being valid so long as it is AMD's way of doing async which no developer or designer or reviewer has claimed or agreed with. In other words, what you posted was a "no true scottsman" fallacy.

In fact, it actually helps prove that nvidia _can_ do async but does it in a different, and arguably less efficient manner likely because less cores are left idle thanks to a highly efficient scheduler for Pascal.

>>55810580
>he seems to think that parallelism is the only valid form of async
And he's right since it's objectively better.

Don't worry, senpai. When the 1100 series comes around to parallelism, you'll be one board.

>>55810341
>chart doesn't say which label is faster

ok jim

>>55810614
>And he's right since it's objectively better.
One method is better than the other and that makes the other method NOT async computing?

> Async compute itself is a catch-all term – there are lots of things you can do with asynchronous work submission/execution
You're committing the same "No true scottsman" if you are claiming concurrency is not async but parallelism is.

>>55810614
>Don't worry, senpai. When the 1100 series comes around to parallelism, you'll be one board.
I dont waste money on a new GPU every year because Im not a good little Goy.

Sorry.

>>55810614
>mudslime
>right
nope

>>55810425
There is no 1070 below 450USD right now. But at least that shit stays in stock.

1080 are impossible to buy unless you follow retailers on twitter and get email notifications.

>>55810639
Hint: Nvidia is faster

>>55810711
There's actually a bunch of them lower: http://www.nowinstock.net/computers/videocards/nvidia/gtx1070/

>>55810005
>>55810057

lol amd cpu's are trash pleb tier

>no proper async, Vulkan, DX12
>still has the faster cards (GTX1080)

dealwithit.wav

>>55810005
Unironically works better than AMD GPU with AMD CPU, because AMD still hasn't multi threaded their driver. Even tho they had years to do it.

I think someone at AMD claimed it was impossible, even tho nvidia did it. I don't even know what the fuck is wrong with them.

>>55810614
>>55810508
>>55810471
>>55810445
I know this is pointless to ask but now that you've been provided proof that Nvidia does async, "true" async that just isn't the same approach as AMD uses, can you give up the misconception that "Nvidia can't do async?"

>>55809820
no
>>>/v/

>>55810997
It only does it on Pascal for it's crappier method. Prior to that it didn't and thus the Nvidia can't do AC was born. We don't say it cannot do AC anymore. We just say it uses a crappier implementation and Nvidia has enough cout to 'encourage' devs to use their implementation. I am not saying Futuremark were paid off since AMD signed off on it. But it would have been nice to see a path for both architectures offered instead of the one option that puts Pascal in a good (but not favorable) light.

>>55811105
/v/ rarely talk about GPU's and the technology behind them. Go there now and look through the catalogue. GPU's are technology and we discuss the technology behind them. Fuck off with this back to /v/ meme.

>>55811131
>It only does it on Pascal for it's crappier method
Debatable quality aside, are you acceding the point that Pascal does async?

>>55811131
>Prior to that it didn't
Maxwell was capable of async.
>From a technical perspective, NVIDIA has slowly evolved their work queue execution abilities over time. Consumer Kepler (GK10x) could only handle a single work queue, while Big Kepler (GK110/GK210) added HyperQ, which introduced a 32 queue setup, but one that could only be used with pure compute workloads. For HPC users this was a big deal, but for consumer use cases there was no support for mixing HyperQ compute queues with a graphics queue.
>Moving to Maxwell, Maxwell 1 was a repeat of Big Kepler, offering HyperQ without any way to mix it with graphics. It was only with Maxwell 2 that NVIDIA finally gained the ability to mix compute queues with graphics mode, allowing for the single graphics queue to be joined with up to 31 compute queues, for a total of 32 queues.
>This from a technical perspective is all that you need to offer a basic level of asynchronous compute support: expose multiple queues so that asynchronous jobs can be submitted.

>We just say
"We?" Are you including others like >>55810409
>But [Nvidia] can't [into async].
And
>>55810471
>Nvidia literally took off the hardware required for Async
These guys don't say that Nvidia is using "a crappier method." They say Pascal cannot do it.

>>55811145
>we discuss the technology behind them
Wait, so the technology behind modern GPUs is just corporate shilling and buzzwords?
Amazing...

File: async vs premption.webm (2MB, 960x540px) Image search: [Google]

2MB, 960x540px

>>55810508

Enjoy shit image quality, low res. shadows and shit DX12/Vulkan performance

>>55811347
wow you sound a bit mad there.
calm down and get a job.
you'll move up some day.

>>55811361
>giving me shit for being an informed consumer and not wanting to get screwed by novidya

Intel CPU and AMD gpu masterrace familia

>>55810508
If Time Spy had coded specifically to use AMD's optimized ACEs, which would favor one vendor (AMD) over another (Nvidia) they would invalidate their reliability to be provide a neutral benchmark. That would, in effect, be using code to enhance AMD's results.

Oh, and just to clean up technical jargon. Concurrency is parallelism. It is static parallelism.

>>55810898
R&D on nasa budget

>>55811347
You mean enjoy better performance than any AMD card available

>>55810898
>I don't even know what the fuck is wrong with them.

AMD is totally incompetent at this point. I mean they just released a video card that couldn't even meet the simple power limits of a motherboard. They are retards.

>>55811425
No. There should be two code paths upon detection of what the GPU currently being used is. If Nvidia use concurrent. If AMD use parallel. The days when you only had one benchmark path (DX11) are over. DX12 does not work like that anymore. It can use different paths depending on which you implement.

>>55811510
>There should be two code paths upon detection of what the GPU currently being used is. If Nvidia use concurrent. If AMD use parallel.
How are you going to write two code paths for the same thing? Concurrency is parallelism. What you are asking for is a different code path that is optimized for AMD rather than a path that is vendor neutral.

I have a 970 and perfectly happy with it. I will only upgrade when a card comes out at a reasonable price that can get 60 fps at 4K.

>>55811534
>What you are asking for is a different code path that is optimized for AMD

This is exactly what AMDtards want. They don't give a shit about objectivism, they just want to try to make AMD look good at any cost, including lying cheating and stealing.

You did good.

The BEST AMD fanboys can say is that MAYBE AMD's hardware async compute will matter some time in the future.

>>55811534
It cannot be vendor neutral since it favors light load AC and does not allow the AMD architecture to use it's full power. AMD can handle AC+graphics in parallel. Nvidia cannot. How can you test the full potential of each if one is not allowed to perform at it's fullest? I am not saying it should be at the expense of gimping Nvidia's performance in any way. Both GPU architecture's should be allowed to show off their full potential. A 'blanket' DX12 'benchmark' is worthless. Of course Nvidia fans are happy because it skews the results to make their GPU's not look so bad. I don't care if AMD signed off on this. They are dumb shits for doing so.

>>55811570
Just like GameWorks? Skewing performance in Nvidias favour and Nvidia blackmailing reviewers into using GWG's to make themselves look best ever?
Face it, the benchmark SHOULD use HEAVY asynchronous compute+graphics.
If Nvidia are too lazy to implement that, its their own fault. Nobody is stopping them from bringing true improvements/innovation.

You're just upset that GameWorks isn't going to be able to stop AMD in the future.

>>55811657
>It cannot be vendor neutral since it favors light load AC and does not allow the AMD architecture to use it's full power.
Would using its full power require low level optimizations that would specifically benefit one card over another?

>How can you test the full potential of each if one is not allowed to perform at it's fullest?
You don't. That's left to each game developer to work with either vendor to get out the best of either card. Thus why we see Doom getting such good results for AMD cards. It takes full advantage of the shader optimizations AMD provides but nothing at all that Nvidia provides.

That leaves but one choice. A high level approach that highlights the basics both cards are capable of without any unfair optimizations.

>>55811763
>That leaves but one choice *if you want to have a reliable apples-to-apples comparison.
Fixed.

File: CofDH0wVMAEt88R.jpg (167KB, 1200x990px) Image search: [Google]

167KB, 1200x990px

>>55809820
Been waiting for over a week constantly checking sites for the OC Nitro+ 8GB rx 480.

Feeling like I should just use the igpu from my 4790k to tide me over until Vega... probably September time. Or maybe I could wait until August 1. I was hoping to get the 480 by August 1 at least so I could have it ready by my birthday the next day.

>>55811817
If you can wait for Vega, it'll provide a few options:
1. Vega could be good, could answer Nvidia on the high end segment.
2. Vega could be meh. Though with the release of Vega, Polaris MAY go down in price slightly. Even if not, there would be stock again at least.
3. If Vega isn't any good, consider Nvidia if that's something you're willing to do.

>>55811817
I'm with you on this struggle. I fell like bitting the bullet and buying a Fury for $350 right now. The GTX 1070 is too overpriced, and I cannot get a GTX 1060 for a reasonable price either.

>>55811861
>If Vega isn't any good, consider Nvidia if that's something you're willing to do.
DX12 hype bit me too hard to consider that. It's either get the 480 or igpu -> 490 for me.

>>55811903
Fair enough. I don't blame you, especially not considering I've got a Fury X on a custom loop so don't plan to change until Navi (or maybe go crossfire.)
In that case, unless you really can't bear the iGPU, you'd definitely get more value out of waiting because its always worth paying out for a higher grade card if the price increase isn't drastic.

>>55811760
You're just spewing more bullshit. Time Spy was literally created with AMD's input. It's a balanced benchmark but you're mad because it shows AMD cards are just plain bad, even in DX12.

>>55811903
Nvidia cards are literally better than AMD at DX12. See Time Spy, the new DX12 benchmark, where it clearly shows this.

File: 3LY8DmC.gif (808KB, 328x268px) Image search: [Google]

808KB, 328x268px

>>55811938
I've seen those charts and benchmarks and I'm aware of the 1060/1070 performance. The other thing nagging at me with this choice is the async thing shown in >>55811278 and this feeling that AMD may have more advanced stuff for the not-near future. I may get the 480 soon and then use that until some 590, for example. The performance difference between the 480 and 1060 is negligible to me considering I've been playing an 8850M from my laptop for the past two years, and nothing better than the R9 280X. 1060's 6GB is also too little for me.

People regretting getting the 970 because it actually has 3.5GB instead of the full 4GB also affects my perception of Nvidia cards in the negative.

>>55812037
>i want more than 6gb of vram
For what? If you are playing at higher resolutions than 1080p the 480 is garbage anyway. It is just classic amd padding out their hardware stats with things they cannot make proper use of.

>>55812037
>this feeling that AMD may have more advanced stuff for the not-near future

That's complete bullshit. AMD sadly doesn't even have the money to create anything advanced anymore. You can see this how the newest card can't even compete with Nvidia's midrange. If you want long term support, definitely go with Nvidia because their cards will be far better supported than anything by AMD.

>>55812199
but they forgot about my keplar card..... :c

File: Nitro+ 480.jpg (45KB, 800x662px) Image search: [Google]

45KB, 800x662px

>>55812161
Maybe video rendering and encoding?

>>55812199
...ok. Which of the new nvidia cards near $260 should I get that are sleek like the Nitro+?

Maybe I'll just get a 590 whenever it comes out or whatever the Nvidia counterpart would be.

>>55809820
>$459

you should have waited, its still way too overpriced for the card.

>>55810341
>a card that costs 27% more is 12% faster

Wew lad

>>55812596
Yeah, and doesn't even burn your motherboard or hit 90c on load

>>55811861
That's a whole fuckin year to wait.

Is there anything in the next couple of months at least?

I'm in a dilemma too. I want a Freesync monitor. That pretty much rules out Nvidia. But I don't want a shit monitor either. I would prefer 1440p and keep reading about shitty Acer IPS, not so hot Asus limited Freesync range and IPS QC problems. Then I look at BenQ and it's a TN panel. But AMD has nothing new to drive it and it would be stupid to buy a Fury X now.

>>55812675
Are you the kind of idiot who buys stock or "Founders editions" ?

>>55809820
>$460
>to play computer games

>>55812802
Nah, but lots of people are considering the amount of shilling this board experiences on a daily basis.

File: 1460606471153.jpg (32KB, 500x280px) Image search: [Google]

32KB, 500x280px

>>55809820
did exact opposite - no regrets
imagine that!

>>55812809
Considering ur mum is for free, I can see your point.

File: 1454378707218.png (154KB, 557x605px) Image search: [Google]

154KB, 557x605px

>>55809820
Good on ya lad.

>>55812675
That really depends on your environment. If you live in a place with good aircon or where the temps in your room rarely hit 28c in the summer then it should not be a real issues. The Powercolor barely breaks 69c under load in a 21c room. Add another 8c and then it's 76c. Big deal. If you are buying a ref 480 you are gonna water cool it anyhow. Aren't you? Well you should be. Heat is only an issue if it's hitting over 100 on a regular basis. The real consideration should be fan noise and power usage if you are poor. Again the Powercolor is pretty damned quiet under load and in quiet mode nearly matches the 1060 by 1% difference.

>>55809840

> $0.20 has been deposited into your account.

File: DPC.png (19KB, 575x486px) Image search: [Google]

19KB, 575x486px

>>55809820
FEELS BAD man
GTX 1080 lateNCY IS FUCKING RETARDED.
DX12 support gains like 4% fps
fewls bad

>>55812850
I mean it nearly matches the 1060's power usage by 1%

>>55809840
>>55809863
>>55809952
>>55809988
>>55809997

Good goy. Keep drinking the Koolaid.

>>55812863
Is that after the driver update? If so then I feel sorry for you bro. Also if using Win10 use LatencyMon as DPCLC is bugged in Win10. I am staying with AMD for now since I need to use low latency ASIO apps.

RX 480 just got a 10% increase in performance with Rise of the Tomb Raider on the latest driver. That could make things interesting with regards to the 1060 :)

>>55812878
>Also if using Win10 use LatencyMon
thanks, i didn't know

Nvidia said they found the cause, but aren't ready for a fix until after the next Major driver update.
i am 100% certain the issue is caused by the voltage and clock speed management.
as this 1080 idles 10C higher(29C) than my 780ti did(19C) max temps are the same at 32C

File: nvidia.jpg (94KB, 1152x866px) Image search: [Google]

94KB, 1152x866px

>buying a graphics card that doesn't look like this
you fucked up

>>55811278
>AMD cards are hotter, because they don't micro idle.

What now Nvidiots? Your card cant even keep it up.

Heres the thing you chuckefucks fail to completely absorb:
the nvidia gpu cannot actually RUN async, it can accept async tasks and sorts them out in drivers so it runs it using context swtiching, but it's all serial. It can TAKE async instructions, and it can context switch really fast, but all the operations are being run sequentially instead of in parallel, which is the entire point of async compute.

their suggested implementation, which futuremark used, maximizes the performance on their cards by working around the limitation, also means having the async compute method being done in a less-than-ideal way for truly parallel-capable hardware.

tl;dr: nvidia's async is "close enough" but not actually asynchronous in execution

Guys my 1080 has so much latency holy fucking shit. I'm actually on SUICIDE WATCH right now.
I should have waited for the 1080Ti and not been a beta tester for Nvidias Jewish equipment practices.
Why is the latency so damn bad. Literally unplayable. Fuuuuuccckkkkk.

File: 1af2f03c6df7ee6d2cdff670e9028ce7.1600.jpg (24KB, 500x375px) Image search: [Google]

24KB, 500x375px

>>55812320
I ended up getting the XFX R9 390 going for $269.99 with $20 mail-in rebate + Total War: Warhammer on newegg.

File: 1370465069126.gif (316KB, 269x169px) Image search: [Google]

316KB, 269x169px

Currently on the market for a new GPU to replace my trusty vapor-x 7970 and was just about set on the fury x despite its age but the measly 4gb vram put me off, just too little ram for a card that I plan to spend a few years with.

So since vega Isn't coming out till next year I guess i'll get a 1080 for now and sell it when vega launches, people seem to pay higher for used nvidia cards so it'll be all good.

>>55813058
pic and tim stamp or ur full of sht.
yes my 1080 has high latency, but games are still perfectly usable.

>>55812982
AMD cards are hotter because they use a HArdware scheduler.
even a non coding normie like me knows this.

>>55813099
This has to be the best graphics card you can currently get for your money , Even considering the 480.

The Performance , The Price , The Backplate... My god its all there.

>>55813146
Thanks. Feels great to be reassured in my decison.

>>55809820
>No Async
>No DX12
>No Vulkan
Wew enjoy it lad

>>55813108

Should add that while I haven't had any problems with my 7970, a few weeks ago I noticed one of the fans stopped working so I emailed sapphire to see If they could sell me a new fan and basically their response was "lol no our problem" as well as "get some shitty aftermarket gpu cooler and frankenstein your card".

Ebay does sell proper replacement fans but they're fucking expensive, costing 10% of what the card Is worth. Card runs fine with only one fan with temperatures under load at about 75ºC but it seems kinda wasteful to have only one of the two fans working.

>>55813256
You forgot
>no freesync

Threads like this are the reason /g/ needs post IDs.

I'm not in the loop for some time now, but what's going on? Why the hate on AMD.

Couple of years ago the hate was concentrated at NVIDIA.

>>55809820
The 1070 is pretty good and the temps are way better than I expected.

>>55813287
How much are you looking to spend ? Imo you have a bunch of options depending on what kind of purchase you wanna make. If you want something that will just werk for games go with the 1080 , its a beast in games.

If you want something a little more professional / creator go for a dual amd card setup. I would get two 390s which would cost you about the same thing as a single 1080 but you got that extra gpu power and also heat. You could also invest in the Pro Duo which is about $1500 but is effectively two liquid cooled nanos so instead of 8bg of shared vram you get 4.

Apparently a beast for 3d rendering and video editing , I would wait for some more benchmarks before buying such a card though.

>>55813353

I'm more fond of having a "it just works" setup, so highest end card possible that'll give me anything over 60fps for as many years possible. Never considered dual gpu solutions since they don't work on all games and OCing Is not an option since I value stability(i leave my computer on 24/7).

Only thing I do with my computer that Isn't gayming or browsing Is some autocad every now and then but my 7970 seems to hold up alright so I guess any newer high end card would be even better.

As I said, I'm currently Interested In the 1080 since It's the current nvidia flagship and nvidia cards seem to hold their price better over time so I'd have to put less money Into the newer card if I actually go forth with my plan to get vega when It comes out.

>>55809820
>I bought a video card 2 minutes ago
>No regrets yet
Good for you, OP

I want to buy a rx480...but I won't overclock it or anything else. Will reference be enough?

>>55813887
If you are going for the 480 there is literally no reason not to get the 390 instead.

They both cost about the same and appart from the 480 being slightly more recent the 390 performs better in all aspects , disregarding heat and power draw , but thats not a problem if you dont plan on OC ing the fuck out of it.

>$460
>$379 (((MSRP)))

>>55814031
Well I saw benchmarks and the 480 was better? But like I said, I want to buy the reference model, because I am sick of waiting for the nitro. Would I see huge drawbacks?

>>55812875
le evil joo meme
>>>/pol/

>>55814031
Yeah, a lot of power draws are shown atfull whack and that's what's putting me off on rx 480, vs 1060. I can't quite make sense of how overblown concerns of the 480 are.

>>55814337
Not him but I don't see it as literally meaning jew but a slightly ignorant shorthand for corporation piss taking. granted pol may literally think of it as a jew conspiracy.

>>55814276
The RX 480 partner cards will give better performance than ref. Grab a Powercolor or Nitro+ OC. With Trixx 3.0 it should hopefully be able to verclock past 1400Mhz. At around 1420Mhz it wipes the floor with the GTX 1060. The only downside is the extra power and heat required. Over on overclockers.co.uk forums one guy was able to overclock his Nitro+ OC to 1400Mhz with just Wattman alone. He got lucky on the silicon lottery I guess but that means his card should overclock like a beast when Trixx 3.0 is released or the new Afterburner if it gets there first.

>>55810457
>eastern europe
who cares about slavs you subhuman?

>>55815008
Mh....ok thanks! But like I said I don´t want to overclock it.

I am really confused, I see benchmarks where the 480 is better or at least good with the 390x. And people saying 480 will benefit most from new drivers. On the other hand, buying tech when it releases is always stupid and 390x should have more power than the 480 or do I have a mistake here?

>>55815074
They are about equal. For now. RX 480 Powercolor has better power usage and quieter than ref or Nitro+

>>55815107
so rx 480 will be better than a 390x?

Well then looks like I wait either for a rx 480 nitro or I will buy the reference, either way it will be better than a 390x

>>55815130
If you are buying an XFX Nitro (Good card btw) you may as well get the Powercolor. Both are 310mm right? I have not checked.

>>55812596
what the fuck shitty country do you live in? let me guess, you're an amerifat or kangaroo fucker. the 1060 aib vs 480 aib is 9% cheaper but 12% faster.

>>55815170
you mean the length of them? if yes Nitro is 30,8cm and XFX looks like 14,3cm

I would have bought the 390x nitro, because it is only 335€ and this is only 10€ more than a rx 480 nitro.

>>55815170
Aren't they notoriously bad on returns?

kill yourself

>>55815228
>being a yurocuck

>>55815251
Who Powercolor? I don't have experience with them but did read somewhere they can be. Might just be a isolated case though. You still have rights no matter who you buy from so I would not sweat it too much.

>>55815325
Problem is I live in the UK and for peace of mind I like whoever I buy from to have UK returns centres.

I wonder how the Gigabyte g1 rx 480 will fare?

Failing that it's an EVGA or Zotac 1060.

>>55815068
Slavs like me who have to move west and steal your jobs in order to earn enough money to be able to buy decent PC.

File: t3_4v5n9f.jpg (56KB, 663x656px) Image search: [Google]

56KB, 663x656px

Give it a year and you'll regret your choice.

File: 1417290693514.jpg (54KB, 323x249px) Image search: [Google]

54KB, 323x249px

>>55809870

>>55815958
do you mean me >>55815130 ?

But this is also a AMD optimized game, the real graphical monster I want to play is ROTTR. Also on Fallout 4 the 480 is better than the 390x?

>>55816070
oh btw. I only play in 1080p

>>55816070
Check out this. Huge boost to RX 480 in RotTR with the latest AMD driver. With a mild overclock it will equal or surpass the GTX 1060 entirely.

>>55815716

Hey bro, my 6670 finally died... Wanna send me your old AMD card?

>>55809952
>resort to AMD
>being this young

>>>/out/

>>55810492
Just curious why you felt the need to make that small upgrade. Wasn't the 970 performing well enough? I have a 670 and am only now thinking of upgrading. And even then maybe not.

>>55809870
Don't install Gentoo

File: IMG_0664.png (32KB, 624x468px) Image search: [Google]

32KB, 624x468px

>>55809864
>Gay-PC

>Buying EVGA

File: 1337521399567.jpg (53KB, 622x562px) Image search: [Google]

53KB, 622x562px

>>55809820
>he fell for the meme and bought an overpriced Maxwell

https://www.youtube.com/watch?v=nDaekpMBYUA

>Debating between 480 and 1060
>Can't buy either right now anyways

Guess I''ll just keep using my 6970

>>55809820
>Have I done you proud
No you're a massive disappointment. If you wanted someone to be proud of you for buying an nvidia card you should have posted to /v/.

File: jew intensifies.gif (212KB, 501x585px) Image search: [Google]

212KB, 501x585px

>>55810036
>most valuable card

>>55810341
>averaging percentages from different games
Meaningless numbers. Literally mixing pears with apples.

>20 fps / 25 fps
>25% FASTER!!!!! OMG!!!!

>100 fps / 105 fps
>5% faster

Yes, the 1060 is faster, but not by 12%.

>>55813099
lucky you I wanted that card so bad but the thing was non existing and only found 390X which was freaking 455$, 20 less than a GTX 1070 so naturally gone team grean

File: 1468889058719.jpg (119KB, 796x805px) Image search: [Google]

119KB, 796x805px

>>55809820
>$459.99 for a $379.99 gpu

>>55820404
That's what happens when the other company doesn't provide any competing product. Remember the 5850/70 before the 470/480 launched?

>>55820404
That's what happens when you buy from a company the imposes artificial scarcity. They make even more money and people are more desperate to buy because they might not get one otherwise.

File: 1468847807266.png (1MB, 1098x1645px) Image search: [Google]

1MB, 1098x1645px

>>55810005
INTEL AMD HERE

>>55818018
Maxwell scales 1.1 per clock over Pascal.

1450 Mhz 980ti should be equal to 1600 Mhz 1080 (thats the stock speed).

With 1080 going 2000 Mhz on OC, you can assume the difference between the two is roughly 25% on OC vs OC.

1080 stock = 980 ti OC

>>55820501
>GTX 480
Oh man, those housefire memes...

>>55816204

144hz.

>>55820561
Anon, why are you a retard? Stock 1080 beats the shit out of a 980ti, even OC'd. 1070 is ~5-10% better than a stock 980ti.

File: Nvidia legacy products.png (110KB, 983x1192px) Image search: [Google]

110KB, 983x1192px

>maxwell is now officially on legacy support

Man those 980ti's and titan x's sure lasted long goy.

Yup. Enjoy having fun with your GPU instead of using it to fulfil a personal white-knighting pajeet crusade.

If I had the choice between a 480 sapphire nitro (316€) shipping sometimes in mid august or a 390x nitro (333€) which should I get? 1080p gaming, 500 watt psu, i5 6600 non k version

>>55809820
I paid 550 for a 1080 on amazon with a 10% wedding discount that I applied for way back, I chose to use it now.

i was thinking of being a rich bitch and shilling out on a 1080 and big 21:9 monitor, among other parts like a 6700k and mobo to use it with

Which companies have good fans? Is it still Asus and Gigabyte at the top? Is Zotac any good?
Also, with a 6700 would i need wangblows10 because reasons? I'd like to use win7, since im never gonna switch to linux despite telling myself I will

>>55809820
Congratulations

You got tricked by the shills

>>55821644
>shilling is the reason why AMD hasn't yet produced a product that is competitive with the 1070.
Sounds like fanboy logic to me.

>>55820783
Stock 1080 is only about ~10% better than 980ti OC @ 1450.

On some games, the difference dies down to ~5%, on other 15% or so difference.


I wouldn't say "beats the shit out of" but rather handily beats it.

>>55821850
>Stock 1080 is only about ~10% better than 980ti OC @ 1450.
Source?