GTX 1060 vs RX 480

Get an nvidia card. Don't play yourself. AMD is low quality pajeet shit.

480 is 100 less where I live

File: 480_launch.jpg (155KB, 490x554px) Image search: [Google]

155KB, 490x554px

>>55844930
Lol, There is no "Nvidia vs. AMD", that ended a long time ago.

>>55844930
480 is a cuck card that fries mobos, go with nvidia

File: 1469234333835.png (667KB, 3882x2171px) Image search: [Google]

667KB, 3882x2171px

>>55844930
next question pls

>>55844930
What's your cpu? 480 only pulls ahead if you're using something beasty like i7 6700k

File: ZEXVOtW.jpg (838KB, 1080x2609px) Image search: [Google]

838KB, 1080x2609px

>>55844930
Get the 1060 because the 480 DX12 is a meme

>>55844930
>RX 480 outperformed the 1060 on the Doom Vulkan patch
Doom Vulkan uses specific AMD shader optimizations. Nor was async enabled for Nvidia cards. In essence, it is a game specifically tailored to use on an AMD card and not at all beyond the basics of Vulkan for AMD cards. If this was gameworks AMDrones would be saying that IDsoft "gimped" Nvidia cards.

How about GTX 960 vs R9 380? Not everybody is made of money you know.

>>55845101
>>>55844930 (OP)
>cuck card

maybe you should give that meme a rest

RX 480 (4GB): lower performance, lower price
GTX 1060: higher performance, higher price

Don't consider the 8GB version of the RX 480, is not really worth it.

File: Wew.png (2MB, 1920x1080px) Image search: [Google]

2MB, 1920x1080px

1060 works better with budget hardware in both dx11 and dx12 so based on your question I'd say the 1060.

Here's a benchmark of a very popular game that isn't DOOM or any gameworks/amd optimized game. Even at the maximum overclock the 480 can get to with that 6 pin, it's still behind the 1060 stock most likely due to amd's inferior driver overhead. This test was done with an i5 3570k btw.

>>55845134
kek measuring dx12 games on dx11 path

seems legit

>>55845176
380 is so much better than 960 even with 2gb 380 still faster.

>buying 380 and gt 960
>when 1060 and 480 exist

>>55845169
>. . . and not at all beyond the basics of Vulkan for *Nvidia cards
Fixed.

>>55845176
The 960 is a glorified moba card and the 380 isn't much better but has housefire levels of power consumption. Avoid both.

RX480 if you want to play and DX12 games as
>>55845162 has listed

>>55845211
The 480 is about $60 bucks more expensive, and the 1060 about $160 more expensive. That is over double the price. And you have to actually find them for sale.
Also, on the RX 480 I don't get the neat aftermarket coolers that turn the fans off if the card is idle, which I really like.
The benchmarks I've seen show them to be quite close in terms of performace,
If I can run, for example, GTA V in high, fallout 4 on very high, witcher 3 on medium at 45+ fps I need no more power than that.

>>55845219
What should I get at max 150 then?

>>55845162
Perhaps you can clear up a confusion for me.

Direct X 12 provides developers with a low level API to bring out the full potential of either developer's gpus. Nvidia has far more resources, experience and connections with developers to bring out the full potential of their cards in comparison to AMD.

What has magically changed with DX12 to make it like that AMD will magically make things better for AMD than previously?

>>55845162
Lmao did they seriously advertise these games to convince me to buy amd? Half these games are utter trash or not even full games like halo 5 forge editor.

No wonder nvidia gets all the sales. They at least advertise good games like witcher 3 or (based on opinion) fallout 4.

>>55844930
wow thread is full of shills, its easy..
If you want to play dx11 games on highsted fps and can afford to buy a new garpics card in 2years buy the 1060, if you want to future proofing and good dx12/vulkan performance buy te 480 you wont need to upgrade for about 4yrs because the dx12/vulkan performance is really high.

>>55845252
What country

>>55845252
wait for the RX 470

File: [email protected] (487KB, 500x373px) Image search: [Google]

487KB, 500x373px

>>55845259
>buy te 480
>you wont need to upgrade for about 4yrs

>>55845169
Doom Shader Intrinsic on nVidia soon?
https://developer.nvidia.com/reading-between-threads-shader-intrinsics

>>55845265
>>55845270
USA actually, but I leave in at friday so I need to decide within 1 or 2 days, and but it somewhere that I can either go in person or has fast shipping that does not cost an arm and a leg.

File: value.png (43KB, 500x1210px) Image search: [Google]

43KB, 500x1210px

The 1060 is by far the best value card you can get.

>>55845308
>No async of the singularity
>No DX12
>No Vulkan
Techpowerup a shit

>>55845296
>Doom Shader Intrinsic on nVidia soon?
One can hope. With Time Spy we see a high level comparison without any vendor specific optimizing. Just pure benefit from the DX12 API. Maybe with Vulkan we can see the something more of a full-featured DX12 benefits if they give Nvidia gpus similar treatment. Then we might get our first real benchmark with something more than, if I understand the tech correctly, base 11_0.

>>55845321
>No DX12
Hitman is DX12 but they use the DX11 version because DX12 is still buggy on Nvidia gpus.
Tomb Raider is DX12.

>>55845197
Both mod are there...

>>55845321
See >>55845134

The 1060 destroys the 480 in DX12, Ashes, basically everything.

>>55845336
hitman dx12 is broken on all gpu. it does stupid shit like take 15 minutes to boot (last time i played anyway) and causes textures to become all blurry and shit. to be honest nearly every single dx12 game so far is broken in one way or another.

using this list >>55845162

>tomb raider
dx12 was broken for like 6 months before they patched it a few weeks ago. still is kinda broken though with people getting mixed results

>hitman
see above.

>gears ultimate
broken as fuck with amd and nvidia gpu being crippled

>quantum break
not even going to mention how broken this game is since everyone knows already

>ark
literally runs at 20 fps on a 1080 because of shitty optimization

that's 5 of the 10 dx12 games on that list that are currently out to buy that are broken in some way.

I think DX12 will remind people why we used abstract APIs in the first plce

File: doom_vs_nvidia.png (96KB, 790x336px) Image search: [Google]

96KB, 790x336px

>>55845329

>>55845455
>hitman dx12 is broken on all gpu.
Oh? I thought the 16.7.3 patch fixed most of the bugs.

File: upload_2016-7-29_11-40-24.png (142KB, 1171x661px) Image search: [Google]

142KB, 1171x661px

>>55845484
>>55845296
Doom still AMD optimized for Vulkan, no nVidia specific optimizations yet

>>55845406
it's also more expensive

>>55845484
Uh-huh. . . And what is your point? Are you repeating the same information that was already provided?

If the prices anything like in Norway (2500 nok for rx 480 and 2700 for 1060) then go for the 1060, since the price difference is minuscule and you probably don't want shitty P(ajeet)C

>>55845494
>amd
>drivers

>>55845308
>1060 100%
>480 116%

File: Hitman Pro 07.31.2016 - 09.19.24.01.png (159KB, 1920x1080px) Image search: [Google]

159KB, 1920x1080px

>>55845494
16.7.3?

1060 is better, but do you really want to pay an extra $100 for 5 extra FPS?

File: temp.jpg (335KB, 1671x1351px) Image search: [Google]

335KB, 1671x1351px

>>55845562
Did I stutter?

>>55845606
you should have been more clear about it being a driver and not a patch. saying patch sounds like its a game patch.

it's not a driver problem anyway. dx12 mode is still broken. i just tried booting in dx12 and it didn't boot at all and ended up crashing steam.

>>55845647
>you should have been more clear about it being a driver and not a patch. saying patch sounds like its a game patch.
I'll try to be more pedantic in the future for autists on /g/.

>>55845647
https://www.youtube.com/watch?v=OdZkR8_-LH4

So in my country both cards cost essentially the same (279 for the 1060, cheapest 8gb RX 480 is 269). At the same price the 1060 is a no brainer right?

>>55844930
Idk
The 480 has Async
Is cheaper

But the 1060 is faster in dx11 and runs much cooler
With better drivers.

The 480.

>>55845647
>>55845664

>amd sponsors game
>doesn't even boot up

I'd rather get crippled by tessellation but actually have the ability to play the damn game instead of not being able to play at all desu.

>>55845673
>The 480 has Async
The 1060/1070/1080 has Async. That's not a benefit to the 480. You could have claimed that the 480 gains about 10% with async enabled which is greater than the 5% with the 1060 but that's a different thing.

>>55845665
Yes, the 1060 is way faster

>>55845531
>1060 120%
>480 116%

Fixed that for you.

>>55845569
>>55845501

There's a $10 difference, for way better performance with the 1060

>>55845691
Let me clarify

The 1060/70/80 have NO hardware support for Async.
It's using preemption based NOT hardware.

As almost all dx12 games disable Async when they detect nvidia GPUs.

-1080 owner

>>55845661
>talking about hitman running in dx12
>retard starts spurting numbers about a patch
>the assumption is the game has been patched based on the context of the discussion
>retard starts talking about unrelated driver updates
>blames other people for not understanding his random numbers

leave that basement, dave. that musty recycled air is intoxicating you to the point you can't even talk straight.

>>55845728
Nah, the price that Nvidia is shilling is a meme. They are only going to make a small number available for that price, that's why that shit is sold out everywhere, low stock. Once that shit gets hyped up, the after market versions will release with super marked up prices. You'll be paying like $100 more for around 5 FPS more.

>>55845756
>The 1060/70/80 have NO hardware support for Async.
>It's using preemption based NOT hardware.
[citation needed]
Anandtech's review of the 1080 and 1070 indicates that there is hardware Async.

>>55845759
>
>talking about hitman running in dx12
Thank you for making it clear that you failed to grasp the context of the post you responded to. Might I suggest you use the backlinks before responding to a post? You might have noticed that the topic was about using Hitman DX12 for benchmarking AMD and Nvidia cards.

>>55844930
What resolution do you plan to play on?

RX 480 much better and future proof, unlike all Nvidia cards

>>55845774
Newegg.

Also in aus you can get evga 1060 for 20 dollars cheaper than reference 480.

>>55844930
it all depends on the games you are using.
Are you playing loads of GTA5, The Witcher 3 or Fallout4?
Then GTX is better choice.

Are you more interested in other games than AAA budget games?
rx is better.

are you only playing old games with casually playing something new now and then?
doesn't matter.

Are you playing

>>55845776
i made the original post you retard. i was talking about the game being broken for everyone (implying it was a developer fuck up) whereas the post i responded to said that it was buggy only on nvidia gpu (implying it was an nvidia driver problem). that means the post you originally responded to was mine, and that in turn means the context of the situation was about the game being broken and not gpu drivers.

go back to school you utter spastic.

>>55845835
nice meme, I'm sure it's more future proof when the 1060 beats it on everything, even in most dx12 tests, laughing at you pajeets when they claimed muh efficiency and it draws lots of power, especially the sapphire nitro+ which was supposed to be the only good AIB even more, the 1060 consumes like a 100 less watt than it

I haven't seen any aftermarket coolers for the RX 480 yet, when are they supposed to come out? How much more expensive can I expect them to be?
The GTX 1060 on the other hand launched much more recently and already has multiple aftermarket coolers such as MSI that is selling for $280, cheaper than the initial $300 price...

>>55845900
>i made the original post you retard.
Which post would that be? Would that be >>55845321 or >>55845308
which I responded to with >>55845336

Perhaps this is a rhetorical question because neither of those posts mentions anything about Hitman which you seem foolishly hung up on.

>>55845906
nothing to do with DX12 to be honest.
r9 290x is was also better for the future and a year or so later started to outperform 780Ti

>>55844930
If you're in the US, 1060 is about 25% more expensive for 8% more performance. Get the 480 unless you want Shadowplay.

If you're in the EU, they're the same price. Get the 1060, it's more powerful.

The 1060 will destroy any game at 1080p and 60 fps.

>>55845308
Looks like it just barely beats a 480 8 gig in perf/dollar. A 480 4 gig beats both in value.

>>55845855
the 1060 is at least $50

File: 1462814863759.jpg (93KB, 620x670px) Image search: [Google]

93KB, 620x670px

>>55845909
how someone can be this retarded is beyond me. keep posting though, it's fucking comedy gold.

>>55845254
AMD has async compute engines... meaning that they can get around 12% gain from async compute (a new way of submitting work to the GPU available in DX12) whereas Nvidia gets around 6% on Pascal.

Usually AMD is behind in performance, but with async compute they can get to around even, but not past Nvidia.

For some reason AMD fanboys think the gains will reach 50% in the future.

>>55845946
>how someone can be this retarded is beyond me.
I have to agree. It is quite entertaining when an idiot claims he had the original post when he failed to use the provided backlinks in the first place. It's a pretty simple error (and common) error to make but childish to try and claim some moral superiority when the evidence is plainly available in the thread.

>>55844930
Nvidia. The budget doesn't matter.

>>55845959
>AMD has async compute engines
Uhhhhh, yeah? Where did anyone claim to the otherwise?

>>55845835
That's bullshit, the 1060 is far more future proof because games have much better Nvidia support

>>55845946
>That picture

Kek I bet he's a nigger too

>>55845965
Neck yourself you subhuman

File: 1430397168041.gif (1MB, 400x368px) Image search: [Google]

1MB, 400x368px

>>55845965
>mfw he actually believes this
>p-provide muh proofz
>he can't even provide his own proof

the city of amd everyone

File: 249-1060.png (103KB, 978x324px) Image search: [Google]

103KB, 978x324px

>>55845774
There are plenty of $249 1060s

>>55845776
Nvidia is incapable of running Async because they lack a hardware scheduler.

Nvidia has been fucing with benchmarks to try to trick ppl.
Like time spy.

File: 1469915157237.jpg (56KB, 480x720px) Image search: [Google]

56KB, 480x720px

>>55844930
1060 if you want to upgrade next year, else 480

>>55845992
But are they in stock?

>>55845907
You can actually get aftermarket cooler 1060s for $249

At this point i'm kind of waiting for AMD's higher end cards to trickle out.

>>55845990
>Neck yourself you subhuman
Why would I do that when I've got a live one here in this thread that has abandoned trying to present anything even roughly similar to a substantive argument? I expect it is because you are an immature twit who has to have the last word.

Let's test this out. My next response to your post will be "It almost hurts to be right so often." You can prove me wrong by not responding to this post or the next one as there will be no point in responding except for the immature need to have the last word.

Depends on whether you like islam green or communist red more.

>>55845925
>If you're in the US, 1060 is about 25% more expensive for 8% more performance.

Nope, see >>55845308

The 1060 at $249 is by far the best value card you can get.

>>55845992
None of those are in stock. You purposely cropped off the "back order."

>>55846002
>Nvidia is incapable of running Async because they lack a hardware scheduler.
Repeating the same claim is not providing a source. Am I to understand that you have nothing but resorting to repetition? At least you didn't resort to personal attacks.

>>55846021
Sign up for notifications and get them when they come back in.

The 1060 sells out so quickly because it's an amazing card at an amazing price.

>>55846014
Why would I do that when I've got a live one here in this thread (>>55846014) that has abandoned trying to present anything even roughly similar to a substantive argument? I expect it is because you are an immature twit who has to have the last word.

Let's test this out. My next response to your post will be "It almost hurts to be right so often." You can prove me wrong by not responding to this post or the next one as there will be no point in responding except for the immature need to have the last word.

>>55846025
Don't even bother trying to argue logic with these AMDtards, they aren't interested in the real world, only spin.

>>55846037
It almost hurts to be right so often.

>>55846034
See: >>55845774

File: 1466450460155.jpg (26KB, 251x242px) Image search: [Google]

26KB, 251x242px

>>55846046
>It almost hurts to be right so often.

>>55846002
You don't know how they implement it, all you know is that they get some gains from async compute, but not as much as AMD.

>>55846045
Intellectual honesty behooves me to keep my mind open to new evidence. In addition to this, civility requires me to provide someone making a claim the opportunity to provide such evidence.

>>55846014
Lmao it's this guy. I remember him from a couple of months ago when he said the exact same shit. He got blown the fuck out for trying to be clever.

>>55845974
No but a guy did post asking "why is AMD better for DX12" so anon was just explaining why.

>>55846037
The autism... baka

>>55846008
Where?

>>55846047
Bro they come back in stock all the time

Check: http://www.nowinstock.net/computers/videocards/nvidia/gtx1060/

>>55846046
>>55846051

It almost hurts to be right so often.

>>55846060
>why is AMD better
So you mislinked?

>>55845930
>gtx 1060
>$399

>rx 480
>$419

https://au.pcpartpicker.com/products/video-card/#c=373,370&sort=a8&page=1

>>55846062
New Egg

>>55845193
Hm ok, so 480 is much much better performance / price, but 1080 sits at the top end at best.

480 seems to be about half the price for at best a 20% drop in performance.

In many games they are equal and in some AMD is better.

File: The_Simpsons-Jeff_Albertson.png (69KB, 250x369px) Image search: [Google]

69KB, 250x369px

>>55845965
>>55846014
>>55846057

Leave.

>>55846025
Async is disabled in dx12 games.
Any fps improvement is due to a more efficient preemption algorithm

1080 owner

>>55846088
Nope, the 1060 has the best price performance of any card available, see >>55845308

>>55846066
>$399

File: jewvidia.png (78KB, 1014x713px) Image search: [Google]

78KB, 1014x713px

480 of course

The 970 has already gotten "legacy" status by nvidia: http://www.nvidia.com/page/legacy.html

Which means no more performance improvements. Nvidia -> buying into planned obsolescence. Seeing how the 1060 can't dx12/vulkan you can expect the same thing.

If you are not committed to Freesync and it is cheaper than the 480 buy the 1060. If you can find the 480 cheaper by around $30 buy it.

>>55846060
Wait a second. Disregard >>55846074
The question was not "Why is DX12 better for AMD." The question was "How does DX12 turn around Nvidia's dominance?" Even your post demonstrates that, at best, AMD would be competitive but not better than Nvidia. This does not seem any different than the past few years or at least enough of a difference to suggest AMD over an equivalent Nvidia product like say suggesting the 480 over the 1060 for any other reason than the previous one, i.e. if it is faster in the games you play.

File: youre-retarded2.jpg (126KB, 489x400px) Image search: [Google]

126KB, 489x400px

>>55846091

>>55846102
>$249

will AMDrones commit seppuku when proper hw async comes with Volta ?

>>55846091
Someone sounds distinctly anally aggravated but not from disagreement but from my continued posting in this thread even if it is not a direct response to the original, and demonstrably foolish, premise.

I'm honored to have somehow evoked such an emotional response for simply being civil and intellectually honest.

>>55846103
That's not true at all, I get performance improvements and driver updates to my 970 all the time.

>>55846127
>not in stock
Keep trying. You won't find a 1060 in stock for less than $300.

>>55846096
>Async is disabled in dx12 games.
Yes. Unless the developer enables it. And as far as I am aware every DX12 game currently has it enabled.

But I digress. This has nothing to do with the so far baseless claim that Nvidia does not have hardware async support.

The real question is : how can /g/ be so autistic to keep recommending nvidia knowing the 780/780 TI history?

>>55845193
>GTX970
>61% GPU
>3414MB VRAM used.

It's throttling GPU usage because it is limiting itself from using the last 0.5GB.
Even then it's faster than RX480.

They said the 4GB was fixed, it wasn't.
A broken card is faster than AMDs fastest card.

>>55846140
Nvida even said in a tweet that Async was disabled by default

Holy Fuck
You gaymers still at it? Fuck you all to /v/ alread

>>55846130
I'm not even the same person you retard. Get out of here with your shitty talk like you're some superior human.

>m-muh big words

>>55846142
>780/780 TI history?
What history would that be? Would this history be in some sort of comparison with the 7970 which was rebranded for the last two generations and effectively still not legacy hardware like the 780/780ti because Nvidia is actually capable of designing, manufacturing and launching new hardware on the same node for every level of the market?

>>55846103
what does previous-generation mean to you?

File: asynchronous-shaders-scheduling-dx12-amd.jpg (124KB, 960x540px) Image search: [Google]

124KB, 960x540px

>>55846116

http://www.overclock.net/t/1606224/various-futuremarks-time-spy-directx-12-benchmark-compromised-less-compute-parallelism-than-doom-aots-also

https://www.computerbase.de/2016-05/geforce-gtx-1080-test/11/

http://ext3h.makegames.de/DX12_Compute.html

http://www.eteknix.com/pascal-gtx-1080-async-compute-explored/

Nvidia doesn't have async shaders at all.

>>55846165
>Nvida even said in a tweet that Async was disabled by default
In the driver and needs dev support? Just like how DX12 disables async unless dev supports it, aka "enables" it?

>>55846164
gpu utilization is the 3rd number. the first number is fan speed.

>>55846103
>amdrone talking about driver support of old cards
that's rich

>>55846199
Makes sense.

Still mad that the memory usage wasn't fixed.

>>55846215
it was fixed.

https://www.youtube.com/watch?v=p6MDUjFiUR4

Help me understand senpais
https://developer.nvidia.com/dx12-dos-and-donts

Do’s

Prefer a tasks graph architecture for parallel draw submission
This way you may achieve sufficient parallelism in terms of draw submission whilst making sure that resource and command queue dependencies get respected

Consider a ‘Master Render Thread’ for work submission with a couple of ‘Worker Threads’ for command list recording, resource creation and PSO ‘Pipeline Stata Object’ (PSO) compilation
The idea is to get the worker threads generate command lists and for the master thread to pick those up and submit them
Expect to maintain separate render paths for each IHV minimum
The app has to replace driver reasoning about how to most efficiently drive the underlying hardware

Don’ts

Don’t rely on the driver to parallelize any Direct3D12 works in driver threads
On DX11 the driver does farm off asynchronous tasks to driver worker threads where possible – this doesn’t happen anymore under DX12
While the total cost of work submission in DX12 has been reduced, the amount of work measured on the application’s thread may be larger due to the loss of driver threading. The more efficiently one can use parallel hardware cores of the CPU to submit work in parallel, the more benefit in terms of draw call submission performance can be expected.

File: 20160731_022707-1.jpg (3MB, 2236x2236px) Image search: [Google]

3MB, 2236x2236px

>>55846197
Dude just stop.
We all know nvidia can't do Async.

I didn't buy a 1080 for dx12 support

>>55846226
>don't use anything that could boost amds performance

>>55846223
Hmm, looks like I'm gonna install MSI Afterburner and check for myself.
Last time it ran like poop if I exceeded the 3.5GB mark.

>>55846164
>A broken card is faster than AMDs fastest card.

That about sums up AMD at this point. They can't even compete with Nvidia's 2 year old broken card.

>>55846215
>>55846223

it was fixed* but the 970 uses less vram in this situation because it has the least vram out of the 3. the other cards have 6/8 gb so the game will probably cache/render in more assets on the cards with more vram. in the same video as the screenshot (480 vs 970 vs 1060) there are points where the game does utilize more than 3.5gb vram on the 970.

>>55846231
>We all know
And yet no one, of all these "we" you allege to speak for, has provided a source. The closest we've come is a tweet. A redundant tweet about DX12's low level API removing driver dependency.

That's the best you can do when asked to provide a source? And you expect "us" (see how anyone can claim "we" don't are right?) to take you seriously?

>>55846226
>Don’t rely on the driver to parallelize any Direct3D12 works in driver threads
>On DX11 the driver does farm off asynchronous tasks to driver worker threads where possible – this doesn’t happen anymore under DX12

Does this mean some form of async was already being done in DX11 path through drivers that's why benefits from DX12 async is small or nonexistent?
And now devs will be the one who will program async?

>>55846252
>(see how anyone can claim "we" [...] are right?
Fixed

>>55846246
RX480 vs 1080
days later
RX480 vs 1070
Week later
RX480 vs 1060
Now
RX480 vs 970

But then again, some old ass AMD cards still work stellar and sometimes even better than the newer RX480.

They REALLY should've gone 8-pin

>>55846252
>>55846193
kys shill

>>55846262
Don't old ass nvidia cards like the 780 ti also compete with the 480 in dx12/opengl?

>GTX 1060
>Legacy in less than 2 years

>>55846275
780?
No,
780Ti, now that is a card that still kicks on.
It's sometimes faster than my 970 in some games.

>>55846275
https://www.computerbase.de/2016-07/doom-vulkan-benchmarks-amd-nvidia/

>rx480 almost as fast as a 980ti in vulkan

no they don't

>>55846269
I saw the links. I stopped reading at the first one because it demonstrates illogical at best. Time Spy is demonstrable vendor neutral. Only AMDrones would believe that failing to use code that would benefit only one card is somehow being vendor neutral.

Would you like to convince me that the rest of the links are not similarly ludicrous?

>>55846231
The 10 series does async just fine.

http://www.anandtech.com/show/10325/the-nvidia-geforce-gtx-1080-and-1070-founders-edition-review/9

In fact the 10 series is much better than anything AMD has in DX12.

Nvidia is superior to AMD in DX12...and everything.

>>55846301
Timespy doesn't use Async.

>>55846299
Yeah the funny thing is that Time Spy was actually developed with AMD's input. They literally helped to make it.

Yet Nvidia cards destroy AMD cards in Time Spy, which shows clearly that Nvidia is better in DX12.

>>55846290
It's a shame that people believe Nvidia gimps older cards

>>55846292
>vulkan
>low level API
>only kepler is underperforming

Thanks Bethesda!

>>55846301
anand is nvidia shill central

>>55846318
sauce, uranus

File: Screenshot_2016-07-25-21-11-43-1.png (303KB, 1003x1315px) Image search: [Google]

303KB, 1003x1315px

>>55846252
You are offering no evidence to the contrary.

>>55846280
Volta comes in may 2017

File: Screenshot_2016-07-25-21-11-46-1.png (283KB, 967x1374px) Image search: [Google]

283KB, 967x1374px

>>55846333
It's simple DirectX 11 optimizations don't work in DirectX 12

>>55846301
Dude. I wouldn't bother. They seem to disregard inconvenient facts. I already brought up the Anandtech article.

I find it hilarious that one of the citations that I'm supposed to "kys" over actually proves the AMDrone wrong (in German so the translation is questionable) as
>Pascal benefits actually Async Compute
They confuse the fact that Maxwell is not Pascall and Maxwell was slower in Async.

>>55846333
Shifting the burden of proof? It is not anyone's obligation to disprove a claim.

>>55846333
Oh, and I provided that Anandtech's review of the 1080/1070 is evidence to the contrary.

Are there any benchmarks of any of the new cards in creative suite? Or does no-one benchmark those anymore.

>>55846362
It's all video games now

>>55846352
Link or ur a a little bitch

File: Summit.jpg (41KB, 678x381px) Image search: [Google]

41KB, 678x381px

>>55846339
>2017
That's HPC Volta senpai, not consumer.
http://www.anandtech.com/show/8727/nvidia-ibm-supercomputers

>>55846379
>Link or ur a a little bitch
Link to the post or link to the article?

For the former see >>55845776
For the latter another anon already provided it in >>55846301

Now that I've fulfilled your attempt to shift the burden of proof can you finally meet my valid request to provide a citation that states Pascal does not have hardware DX12 Async support?

>>55846301
Top tier anand detailed review.

>>55846396
Just read that article it has absolutely no proof that Pascal can do async at all.

It just shows a bunch of dumb slides from Nvidia themselves

File: sad.jpg (7KB, 228x221px) Image search: [Google]

7KB, 228x221px

>mfw one more week left until I get my Nitro+

the wait is almost over

I don't understand why anyone is buying an amd 480 card when the 490 is coming out by the end of the year which is in a few months?

>>55846256
>nvidia already doing async and makes no big deal out of it.
>AMD catches up years later and treat it as the 2nd coming of Christ.

AMD that desperate?

>>55846424
Thank you for proving me right when I previously posted:
>I wouldn't bother. They seem to disregard inconvenient facts.

Oh, and thank you for meeting my expectation that you still cannot provide a citation.

I see no further point in responding.

File: 1070-async.png (17KB, 650x300px) Image search: [Google]

17KB, 650x300px

>>55846424
Uhh clearly you didn't get far enough.

Pic related, the 1070 doing async compute.

>>55846454
>I see no further point in responding
Good.

>>55846463
Dude. The AMDrone's in this thread fail to adequately read their own citations and you expect them to read anyone else's?

>>55846463
Here is a thought. What if async compute doesn't magically gives you free 30% gain and the only reason why AMD cards got such an insane gain in DOOM and Ashes of Benchmarks is:
a) Because their drivers are absolutely garbage in DX11 so AMD actually tried for once
b) Because those two games are simply heavily shilled for AMD same way goymworks put Nvidia 20 fps ahead of AMD counterparts. That would also explain why Hitman episode 2 and Rise of Tomb Raider actually favours Nvidia in performance gain. Async in reality is 5-10% performance gain and nothing else.

>>55845498
>Nvidia has no drivers
kek

>>55846463
Nvidiamark uses preemption retard. Async gets deactivated at driver level.

File: 1455622256183.png (812KB, 600x600px) Image search: [Google]

812KB, 600x600px

>>55846483
>AMD actually tried for once

>>55846463
you forgot to show taht 1070s CPU performance is dropping as GPU performance is growing.

>>55846483
>goymworks put Nvidia 20 fps ahead of AMD counterparts.
One big difference. "goymworks" (strong indicator of being an irrational AMDrone) is not incorporated in DX11 (or 10, 9, etc.) so there is no reason to believe AMD will magically gain 20% or more with DX12 or any other low level API unless the developers have optimized only for AMD gpus.

>>55846483
>Doom
>"The way it's meant to be played"
>AMD Shilled

>>55846492
>Disregards inconvenient fact that proves he didn't disregard the article provided

>>55846492
What's the fuss?
Async benefits AMD so they use it.
Async costs nvidia some little performance so they disable it.

Why do you care so much? What's the point? Just enjoy your fucking card and stop being autistic on an anonymous image board.

>>55846516
>Async costs nvidia some little performance
On Maxwell. On Pascal as well? [citation needed]

>>55846516
>Unified shaders benefit AMD so they use it.
>Unified shaders costs nVidia some little performance so they disable it.

>>55844930
There are no Vulcan titles worth a damn.

>>55846524
Sorry. On maxwell.
I forgot about that.

>>55846530
Does it make you happy when nvidia customers lose performance? Is that what gets you off?
It gets your dick hard, boi?

File: Screenshot_2016-07-31-03-18-44-1.png (225KB, 1059x1288px) Image search: [Google]

225KB, 1059x1288px

>>55846463
Time spy has doesn't use Async you stupid mongoloid.

I already said that:
NOT A SINGLE DX12 GAME ENABLES ASYNC WITH NVIDIA CARDS.

Frame per second gains are the new preemption software and not async.

>>55845256
That is literally every major release for the next few months. Big titles aren't going to suddenly go back to DX11 afterwards.

is this /v/?

>>55846545
nVidia is always staggering progress with every AMDs tech they are reluctant to use it.

GDDR5
Unified Shaders
Async Shaders

Anything that benefits everyone, nVidia is reluctant to use it, they only like proprietary shit, I don't give 2 fucks weather nVidia users loose performance or not.

Their company should use something that is open and available to all.

>>55846545
>Sorry. On maxwell.
>I forgot about that.
That's one the more hilarious things about AMDrones continuing to repeat this myth. They cite the very tweet that demonstrates how Nvidia dropped Async support in drivers, because it slows Maxwell 1 and 2 cards, and does nothing for the Async hardware support on Pascal. That this somehow proves that Pascal doesn't have async support even though every test with async enabled for Pascal shows a clear improvement. They even cite http://ext3h.makegames.de/DX12_Compute.html which explains and shows the SAME THING!

>>55846096
>I can spout memewords without any comprehension
You don't know that. You can't look into the driver code.

File: temp.jpg (261KB, 1673x1225px) Image search: [Google]

261KB, 1673x1225px

>>55846550
https://www.computerbase.de/2016-05/geforce-gtx-1080-test/11/

The funny part is this was linked as support that Pascal does not support Async!

>>55846165
For Maxwell

>>55846574
>Implying AMD doesn't do the same

>>55846096
Earth is flat, the planet being round is a meme.

Property and home owner

>>55846584
Async is disabled in games.
Preemption is being used on a driver level.

Nvidia is only cherry picking dx12 aspects that benefit them.
ASYNC IS noT one of them

>>55846599
Name one open and non-proprietary software/hardware feature that AMD tried to deny nVidia?

>>55846129
No, since the power usage will skyrocket and you'll lose "muh efficiencee".
The reason why Kepler became so much more efficient in comparison to Fermi was because dropping hardware solutions for stuff like that in favour of software.

>>55846616
TresFX
And Async ha ha

File: AMD.png (1MB, 1075x778px) Image search: [Google]

1MB, 1075x778px

>>55846584
>You don't know that.
He does know that because he read the AMD memo and AMD never lies or is wrong.

>>55845722
>1060
>$245
Yeah not in every country. The RX 480 is 80€ - 100€ cheaper.

>>55846616
TressFX.

Wasn't even open source, Nvidia got fucked hard by AMD in the last tomb raider.

>>55844930
How long do you need it for? If you're going to upgrade in a year or two anyway get a 1060 because right now it fares better in the majority of games.

If you want something that'll last the next 4 years get a 480. It'll be continually supported and will more than likely be universally outperforming in the 1060 in 2 years.

Then again if you really did need something to last a long time I'd recommend saving for longer and seeing how the RX 490 vs GTX 1070 fight goes and see how AMD actually fares in all the upcoming DX12 titles.

>>55846627
>>55846622
TressFX is MIT licensed.

File: ac_980ti_vs_fury_x.png (7KB, 798x752px) Image search: [Google]

7KB, 798x752px

>>55846575
Problem with nVidia is that it never got better with Async after 700 series.

Here is Async test some dude did, lower is better.
as you can see nVidia is faster with 32blocks but anything over 32 is shit, while AMD is always equally shit.

>>55844930
Get a 1080

>>55846675
>Problem with nVidia is that it never got better with Async after 700 series.
Your picture is a comparison of Maxwell 2 not Pascal. It's already been noted that Maxwell 2 does not benefit from Async. What is the purpose of your post? Repeating a commonly known point that does not extend to the claim that Pascal does not support async?

>>55846675
Expand on that pic.
>Interest intensifies

>>55846684
I think it is safe to assume he cannot afford to get a 1080 considering how he is coming from a 950 and didn't include any option of buying the 1080.

Try and get Sapphire's RX 480 Nitro+. Benchmarks on the 8gb version have it on par with the 1060 and it keeps up with overclocked 1060s when you overclock it. 8 pin means no fried mobos. Get the 4gb version. I doubt there'll be a significant performance delta.

>>55845134
uhh which one is which in this graph?

>>55846685
Well, pascal has async in theory, it's just hard to use because, reasons.

>>55846713
Hint: The 1060 is way faster

>>55846634
That's not true at all.

The 1060 is better now and will only get better as time goes on. Nvidia has better DX12 performance and continually updates their drivers, while AMD cards just keep getting slower through gimping over time.

File: NotSureIfSerious.jpg (31KB, 689x313px) Image search: [Google]

31KB, 689x313px

>>55846742

>>55845406
2% is not destroying something.

>>55846770
The 1060 is 12% faster than the 480, see >>55845134

>>55846758
>AMD Cards
>Getting gimped

>NVidia Cards
>Getting better over time

what is the 780/780 ti vs 290?

>>55846791
He mentioned Ashes and DX12.
Go look at the Ashes DX12 test.
2% is not destroying something.

File: 59c6ef77_computerbase.jpg (73KB, 624x469px) Image search: [Google]

73KB, 624x469px

>>55846796
Oh hey, look, proof that nvidia isn't gimping kepler in relatively new dx12 title

What do anons?

I can get zotac mini 1060 for 294 euro this week, which I know will be cheapest 1060 in my country, as everything else is 350-450 eur and at that point I would rather buy 1070 for 500 euro or wait possibly till september for rx 480 aib's who will cost God knows how much as reference prices here is 300-350. I want a card that would last me 3 years and will be pairing it with i3 6100.

File: satensmug2.jpg (157KB, 560x570px) Image search: [Google]

157KB, 560x570px

Seems async support is the new excuse for amdrones why rx480 sucks so much.

>>55846808
If the 1060 is the better deal get the 1060

>>55846823
>Nvidiots still comparing poo in loo cooler to non poo coolers

Memes aside amd reference cards are always shit and the non-reference 480s which cost like $220 are kicking ass

File: 1363739327139.png (154KB, 309x302px) Image search: [Google]

154KB, 309x302px

>nvidiots face when no async compute

>>55846758
>AMD cards get gimped drivers
Did you just try and make up your own meme?

File: 1379089675047.gif (875KB, 500x318px) Image search: [Google]

875KB, 500x318px

>>55846903
>mfw you can unlock the last 4 gigs with software but the last .5 gigs is a hardware deficiency and thus unlockable

>>55846807
>980Ti beating Fury X in DX12 AoTS
>GTX 960 beating the Fury X in anything
>GTX 960 beating the 390 in anything
>GTX 770 beating the Fury X in anything
>GTX 770 beating the 390 in anything

>>55846947
Dude, what???

>>55846947
>>980Ti beating Fury X in DX12 AoTS
By .1 of a fps? Not statistically significant and easily disregarded as within the margin of error.

File: 20160731014731_1.jpg (2MB, 3840x2160px) Image search: [Google]

2MB, 3840x2160px

>>55844930

i just bought an RX480 and it's pretty amazing

i play GTA5 at native 4k with everything on high (not ultra) and i get 45 FPS. pic related, a screenshot i took yesterday

and the big advantage for me: i can buy a second RX480 when i have some money to spare, something you won't be able to do with the 1060

wattman is pretty cool too. tl;dr don't listen to paid nvidia shills on /g/, don't waste your money on a more expensive, less futureproof card, and don't give it to a company that fucks its customers in the ass without lube (referring to the 3.5gb debacle)

http://www.bitsandchips.it/7-software/7300-test-nuovi-driver-crimson-16-7-3-boost-in-rise-of-the-tomb-raider-dx12

>>55846973
>something that will allow me to play some of the newer games on at least 60fps.
You are suggesting that the OP not get the 480?

>>55846954
Bro, what?

>>55846973
>TFW because I have a Fury X, I don't have wattman
It sucks because I thought wattman was meant to be for all AMD cards, but the RX480 has its own driver download and I assume that's why my thing hasn't updated from Overdrive.
I even did a fresh install of the drivers to be sure and no effect.

>>55847004
Translate from spaghetti to human

>>55847010
I think "Bro" is confused as to why you responded to a post about how Kepler is not getting gimped by Nvidia through drivers with a post about how FuryX is faster than a lot of older or cheaper gpus.

>>55847008
Considering he's talking about playing it on 4k, I think you better check yourself again m8e

>>55847038
I don't recall the OP mentioning any specific resolution he wanted 60fps in. Would it be safe to assume that he does not want 60fps in 4k because he didn't specifically mention 4k?

File: 1463571134016.jpg (111KB, 700x700px) Image search: [Google]

111KB, 700x700px

>>55846935
AMD gave away tens of thousands of free gigabytes.

Where as nVida kicked and screamed at muh shekels over 0.5 gigs per card.
If that's not a reason to think nVidia only see you for your wallet I don't know what it'll take.

[spoiler]nVidia will start renting cards on a contract that'll cost %250 of the cards "Retail value"[/spoiler]

>>55847008
If he is getting 40fps at 4k, you should get significantly more at 1080p

>>55847065
I don't see any specification for 1080 in the OP.

>>55847063
>AMD gave away tens of thousands of free gigabytes.
Source?

>>55847071
>>55847062
If people don't mention 4k specifically then 1080 is the safest bet and what they are most likely playing at.

>>55847038
Gta 1060 is 50% faster at gta v 4k than the rx 480...

http://m.gadgets.ndtv.com/laptops/reviews/msi-geforce-gtx-1060-gaming-x-and-zotac-geforce-gtx-1060-amp-edition-review-862713

>>55847062
No, it is not safe to assume that OP has a 4k monitor, since 4k monitor users are very rare even now and 1920x1080@60 remains to be the de facto standard even today.
I use a 2560x1440@144hz monitor but even I am one of the few who do.

Unless OP states what monitor he's using, the standard assumption should always be 1920x1080.

>>55847088
And if people don't specify they want Max/High graphic options we should assume they mean 60fps on low/medium?

>>55847095
>http://m.gadgets.ndtv.com/laptops/reviews/msi-geforce-gtx-1060-gaming-x-and-zotac-geforce-gtx-1060-amp-edition-review-862713

That's 18%

>>55846096
>>55846610
You clearly don't know how applications implement async, so you should shut the fuck up on top of Mt. Stupid there.

>>55847097
I have to disagree that lack of information means we should assume one thing. But given that he admits he is a budget gamer I'd agree with you that in this situation it is safe to assume he meant 1080p.

>>55847095
>comparing non-reference cards to reference
Fuck off

File: 1464933033043.png (18KB, 500x500px) Image search: [Google]

18KB, 500x500px

>>55847083
>google flash rx 480 to 8gb

IT'S RAINING MEMORY AT AMD

http://www.gamersnexus.net/guides/2506-how-to-flash-rx-480-vbios-from-4gb-to-8gb

>>55847106
No, you should assume High/Ultra @ 1080p.

>>55847126
>You clearly don't know how applications implement async
Ignorance has never stopped AMDrones before. . .

>>55847130
But you can buy non reference 1060s cheaper than a reference 480...

>>55847122
You're right, I misread it, sorry.
But that still means on the 1060 he'd probably get 50+ fps instead of 45 like that anon with the 480.

>>55847127
For me, its a matter of context. Like you say, he's asking about a single RX480. Nobody sane would think that a single RX480 would get 60+fps at anything set to high or higher at 4k.
If it was asking about the upcoming Titan X (pascal) 60, then sure, 4k would be the safe assumption.

But for most people on /g/, getting a 4k monitor and then asking about anything but the absolute highest just says 1080, perhaps 1440 at a push.

>>55847132
>google
Shifting the burden of proof? Yeah, no.

While I am aware that the first batch of reviewer's cards came with 8gbs that is popularly speculated as being because AMD didn't make any 4gb cards at the time and tried to effectuate it for reviewers by giving them 8gb cards with a bios lock on addressing the other 4gb. However, that doesn't serve as evidence that "tens of thousands" were "given away" in any way.

Are you even familiar with how many 4gb versions of the 480 have sold?

>>55847095
about 25% but what ever makes you sleep better.

>>55847150
>Nobody sane
This is summer time in /g/. AND we've had plenty of evidence from AMDrone posting that "sane" is loosely applicable. . .

>>55847151
I gave you a link fuckhead. Or are you a bot?

>>55847146
>But you can buy non reference 1060s cheaper than a reference 480...
>back order

>But that still means on the 1060 he'd probably get 50+ fps instead of 45 like that anon with the 480.
The other anon was pointing out how decent the 480 did while also mentioning when he has the money he can get a second one and link them for even better performance, something the 1060 can't do.

>>55847150
>But for most people on /g/, getting a 4k monitor and then asking about anything but the absolute highest just says 1080, perhaps 1440 at a push.
Agreed.

>>55847180
If you have the money for 2 480s or 1060s you have no reason not to buy a single 1070/1080 instead.

Multigpu is horrendous.

>>55847196
>I have no money for two rx480
>But next year this time I will

>Just wait for a year and get 1070

>>55847179
>I gave you a link fuckhead
You gave me a link to how an early reviewer's card came with 8gbs of memory on board that was locked off by bios. I already noted that.

Let me ask you a question. If the first few 4gb 480s provided by AMD to reviewers (not to be confused to cards sold to consumers) came with 8gbs on board means that every 4gb 480 sold came with 8gbs on board?

>>55847196
Thing is some people don't have for 2 at once, but they can afford one now, then grab a second one later on when they save more or prices drop.

>>55847180
But aren't most games going to be dx12 or vulkan? With built in multiadapter?
Which makes sli and crossfire redundant?

>>55847206
That is only applicable if you're using a cpu with no gpu and every gpu you own is dead. Otherwise put up with high settings instead of ultra for a year for fucks sake.

>>55847213
People are flashing them in the wild and getting 8gb

Keep your eyes closed, that way you won't notice the 0.5 nVidia stole from you.

http://wccftech.com/amd-rx-480-4gb-retail-cards-8gb/

https://www.reddit.com/r/Amd/comments/4rfo2h/flashing_guide_rx480_4gb_to_8gb/

>>55847032
I'm just commenting on how out of whack those benchmarks are. Even in DX11. a 960 beating a Fury X?

>>55844930
GTX 1060 because buying RX 480 because it has better async is pointless because async is still way off and by the time it's mainstream (if it ever becomes mainstream), you may aswell buy a new GPU and see what async performance is on both nvidia and amd.

RX 480 is overrated, AMD guys should just jump straight to 490.

>>55847257
>in the wild
Uh-huh.
Did I ask if some of the consumer cards were flashable or if EVERY consumer card was flashible?

>There’s no doubt that AMD and its partners will distribute 4GB Radeon RX 480 cards packing the correct amount of physical memory rather soon. After all, AMD will only take the financial hit for so long. What will be interesting to see is how and when altered
I think it is foolish to believe AMD is going to continue to give away free memory unless you can demonstrate that more than a few early 4gb models came with 8gbs of VRAM.

>>55847273
>I'm just commenting on how out of whack those benchmarks are
By noting, like the graph shows, that the 770 and 960 do not beat the Fury X or 390? Are you expecting the graph to show that the 770 or 960 beats the Fury X or 390?

>>55847286
I'm waiting for 590 myself.

>>55847281
this

>>55847273
Ashes is cpu intensive and amd suffers massively in these games in dx11, notice how much better they perform when the stress of poor drivers is lifted from the cpu with dx12?

I've got a bit more of a challenging question.

390x or 980ti? They're pretty neck and neck but the 390x I can get for about $100 less, but it also consumes about 100 more watts of power.

>>55847289
My bad, but still. nVida stile 0.5 from you, and AMD gave away many with an additional 4.

>>55847344
nvidia stole 0.5gb from 970 owners? really? care to show me a picture of a 970 pcb with only 3.5gb worth of vram chips/modules?

>>55847344
>nVida stile 0.5 from you
Except they didn't. There is a full 4gbs of addressable VRAM it is just that Nvidia failed to indicate the last .5gbs was on a different slower bus.

>AMD gave away many with an additional 4
Assumes this was a choice that they had to make rather than an issue with sourcing 512mb chips.

>>55847028
Tomb raider get a 12% improvement with 16.7.3 drivers

File: 1461090171933.jpg (529KB, 1340x1588px) Image search: [Google]

529KB, 1340x1588px

>>55847370
lol this guy is praising amd like they voluntarily gave away 4gb 480 with 4gb deactivated. i bet he'll defend amd also selling early 390 batches as literal bios flashed 290's. like in gpu-z they actually came up as 290 instead of 390. literal jews.

>>55847289
being fair he didn't say EVERY consumer card was flashable initially. The original statement was
>AMD gave away tens of thousands of free gigabytes

>>55847358
>208 bit bus

>>55847391
>being fair he didn't say EVERY consumer card was flashable initially.
No, he said "tens of millions of gbs" were flashible. If you want to get pedantic we could calculate this to mean at least 20,000gbs or 5000 cards. Considering that the 4gb is probably going to be produced in small numbers to upsell to the 8gb version, I find it likely that even 5000 is a larger number than every 4gb version sold to date.

>>55845906
>>55845921
this is the real kicker; AMD keeps improving their old cards with drivers, nvidia stops doing performance updates a year after release.

go AMD and play for 5 years or prepare to buy a new nvidia card within 2 years

>>55847396
where's my picture?

>>55847400
>>being fair he didn't say EVERY consumer card was flashable initially.
See
>>55847063
>AMD gave away tens of thousands of free gigabytes.
Go back to learning how to read a newspaper you mandarin fuck.

>>55847403
im not a poorfag so I buy a new nvidia card every year

>>55847413
I don't understand what your issue is.

His claim was "tens of thousands of gigabytes" were given away. This involves two things. One that 4gb cards had 8gbs of memory and that the 4gb cards are flashible to 8gbs.

I have now addressed both these aspects of his claim and it seems you think I missed something. I do not see what I missed and your vague allegation of having missed something is not helping in any way. It seems safe to assume you are making a vague allegation because you cannot specify how I failed to address his questioned claim.

>>55847335
980ti is stronger than Fury X in most DX11 cases, though 390 is stronger in some DX12 games.
Its a question of do you want to spend extra money on an FPS improvement or is the FPS of a 390 good enough for you?
Might be worth getting the 390 as a more price/perf guided choice and then moving up to the 490/590/690 as necessary.
If you get a 390, I wouldn't think you'd want to upgrade until at least 590 though.

>>55847452
>that *enough* the 4gb cards are flashible to 8gbs *to meet the claim of "tens of thousands."
Fixed for clarity's sake.

File: amd.png (42KB, 653x726px) Image search: [Google]

42KB, 653x726px

>>55847385
DELETE

>>55847303
Did you read the rest of the graph? It shows that in DX11 the 770 and 960 beat both the Fury X and 390. That is what I am questioning.

>>55847400
>No, he said "tens of millions of gbs" were flashible
>>55847063
>AMD gave away tens of thousands of free gigabytes.

Thousands are vastly different from millions.

>>55847469
>Thousands are vastly different from millions.
Are you assuming that millions of 4gb 480s have sold in the few weeks it has been out? I'd like to see a factual basis for this assumption.

>>55847335
Since when is the 390X neck and neck with the 980Ti?

>>55847476
>I'd like to see a *rational basis for this assumption.
Fixed for clarity's sake.

>>55847442
im extremely proud of you. I play maybe one or two non-dota/csgo games per year and have no reason to splash cash like that

>>55847442
If said new card is constant cycle of Titans then yes, you are.

>>55847477
Since DX12?

Seriously thou, if the OP wants better performance now and doesn't mind paying extra, then get the GTX 1060. However if the OP wants better performance for future titles using DX12 and Vulkan, then get one of the Sapphire RX 480s.

Review:
http://hexus.net/tech/reviews/graphics/94969-sapphire-radeon-rx-480-nitro-4gb-8gb-oc/

>I'm looking to upgrade my GPU from a GTX 950 to something that will allow me to play some of the newer games on at least 60fps.
>I'm looking to upgrade my GPU from a GTX 950 to something that will allow me to play some of the newer games on at least 60fps.
/v/
fuck off kid
/v/
/v/
/v/
/v/
fuck of kid
/v/
/v/

File: 1463655002323.jpg (52KB, 600x604px) Image search: [Google]

52KB, 600x604px

>>55847476
>Are you assuming that millions of 4gb 480s have sold in the few weeks it has been out?

That's you

File: w3.jpg (527KB, 1920x1080px) Image search: [Google]

527KB, 1920x1080px

>B-BUT MUH HARDWAREUNBOXED!!!!1

>>55847469
nigga the most popular gpu in the world right now is the 970 and it took like a year to hit 1 million sales

>>55847540
>That's you
No it isn't. I'm glad we agree that the assumption of millions having been sold to date is ridiculous. Thank you for clearing up how the there was nothing overlooked regardless of your presented vague allegation.

>>55847547
i watched that same video. stop bullshitting. the 970 and 1060 are always between 2-10 fps higher than the 480 except in like 3 scenes.

>>55847554
The point is that the other guy said THOUSANDS, not MILLIONS like you tried to imply.

Just face it, you've got egg on your face and badly so.

File: legit.jpg (174KB, 1344x756px) Image search: [Google]

174KB, 1344x756px

>1060 is much faster. Its insane!

>>55847549
10,000 X 4 = 40,000
>tens of thousands
How can you be this retarded, oh that's right. nVidiot

File: staymad.png (3MB, 1921x1023px) Image search: [Google]

3MB, 1921x1023px

>>55847547
i can cherrypick too

>>55847465
Must be the DX11 driver for it being utter shit.
Normally the Fury X smokes a 960 in everything aside from gamesworks benchmarks.
Ashes is incredibly CPU-side since its an RTS, especially on the scale it is.
So yeah. Probably just the driver fighting with the game rather than working with it.

>>55847570
>not MILLIONS like you tried to imply.
He presented his claim in the past tense and I refuted his claim in the past tense. I did not even imply it.
>if EVERY consumer card WAS flashible?
And yet you failed to recognize this and think I was talking about the future.

The only implication is in your mind due to your failure at basic reading comprehension.

>>55847602
see
>>55847575
It's basic maths. nVidiot.

File: 1436932577791.png (744KB, 964x768px) Image search: [Google]

744KB, 964x768px

>>55847547
>both 480 and 970 at 99% gpu utilization
>1060 is at 91%

Are you seriously this retarded? Have a good look at >>55847586 for the real performance gap.

>>55847385
>It's on when amd does it

>>55847621
I'm glad you abandoned the "implied millions" idiocy.

>>55847575
Which fits within what I posted before about "at least 20,000?" Thanks for supporting what I previously posted!

Holy fuck the desperate grasping at straws in this thread is ridiculous!

File: lil.jpg (151KB, 830x661px) Image search: [Google]

151KB, 830x661px

>>55847568
>Lying on the internet
wew.

>>55847575
don't act stupid buddy. everyone here knows that the original poster claimed it as if all the sales of 4gb 480 have 8gb ram. we're talking about a single batch here which is most likely like the first 1000, which would have mostly been given to reviewers and retard redditors who pre ordered it 6 months in advance.

File: defemd.jpg (227KB, 1284x715px) Image search: [Google]

227KB, 1284x715px

>muh gayworks performance
Welp, everything comes with the price.

>>55847661
what point are you even trying to make? this video >>55847586 doesn't even use hairworks so why are you talking about it?

File: sosi.jpg (107KB, 1781x753px) Image search: [Google]

107KB, 1781x753px

>irresponsible amount of performance!

>>55847662
>don't act stupid buddy
Considering that he calls others "nvidiots" and projects his fanboyism on others I think it's safe to assume he is an AMDrone and isn't acting.

>>55847307
Well, you're going to have to wait for late 2017 then.

It doesn't really matter. Both cards are great value.

File: vreti2.jpg (94KB, 677x526px) Image search: [Google]

94KB, 677x526px

Id rather have 3% slower card than a card which will turn into pumpkin in a year.

>>55847662
Still cant run away form 3.5
>implying non existent implications
>>55847063
Was clear in what he said.

>>55847705
Don't you mean "late 2018?" Vega's not expected to be out until early 2017 and I doubt that AMD is going to buck the trend and release a new version of their high end cards in less than a year.

>>55845162
Jesus Christ, is this the lineup that should convince us to upgrade from ours i5 [email protected], gtx ~670 3gb-970 3.5 gb?
My shit is rocking steady at 1440p~@40-60 fps,everthing on max, except that one bullshit gimmick every game has that gives some leverage over other gpu camp,gimmick that even owners of gpu that was made for keep turned off.
Fuck, did graphical progress slowed down since 2006.

>>55847724
>did graphical progress slowed down since 2006.
Yes. Thank you game consoles!

>>55847710
except they both aren't. the reason many people are complaining is because amd fucked up their msrp for aib cards. over here in the UK the gtx 1060 aib starts at £229 whereas the 480 starts at £250. if you put that into a calculator the 1060 is ~9% cheaper but ~12% faster on average. who in their right mind (unless they're a deluded shill) would spend more to ultimately get less in return?