You'r move amd
ThE 1060 IS TWICE AS EVERYTHING HOLY SHIY
>graph starts at 0.8
>>55354227
>Believing anything you see from a website that was proven wrong on almost everything they posted about the RX 480
Top kek
i don't expect it to be insanely faster, so why make the bench make it seem like it?
also, let's not hope that nvidia continues this founders edition shit and places it at 300 dollars.
>>55354269
Effectively the graph starts at 1.0 as the 480 is used as the baseline but it'd look weird if they didn't provide some gradients below that.
where is DX12 performance ? ^)
Cripple cards compute, bit bus and ram then brag about efficiency. RX 480 24 MH/s GTX 1080 1 MH/s
>>55354296
Are you an idiot?
The graph is done totally improperly and in a fashion to make it look like the 1060 is twice or more powerful than the RX 480 because the green bars are twice as long or more than the red bars.
At a GLANCE, it seems like the 1060 is blowing the 480 away. When you look at the small numbers, nVidia is only claiming 15% increase in normal games.
>>55354227
If they can sell it for $250, this will be really really bad for AMD. Like company-ending bad.
>>55354368
>15% increase in normal games
15% increase for the same price is huge
>>55354227
What about the 1070?
>>55354389
>same price
You're talking about Nvidia
>>55354368
>The graph is done totally improperly
Okay. Let's say that they start the graph from zero. They'd have more than half of the graph providing no useful information. That would be both a waste of time (for both the graphic artist and the audience) and a waste of space.
You sir, are an idiot. You are defending a trivial objection that demonstrates a lack of understanding and sophistication.
>>55354389
It's nVidia, the price will be like 30% more for 15% more performance.
>>55354285
>300$ in the usa
>599€ in europe
t-thanks...
>>55354417
>>55354400
Calling it right now, it will be $250, the same price 480 is going for
>>55354368
>nVidia is only claiming 15% increase in normal games.
Oh, and I forgot to get to this as well. "Only?" 15% faster is a pretty big margin. It is a big enough margin to not suggest getting a 480 if you already have a 970/290(x)/390(x) as it wouldn't be much of an upgrade.
>>55354227
This is exactly what I expected, about 15% faster than a 480 while being vastly more efficient.
AMD will compete by being cheaper, while Nvidia offers the superior product.
>>55354227
Hopefully it isn't a pathetic overclocker like GP104
>>55354406
>They'd have more than half of the graph providing no useful information
Except that's wrong, the bars would give a far more realistic look of the performance difference, which is only about 15% but looks like a 100% right now
>>55354406
I see you've never written scientific reports or papers.
Your ass would get laughed at by teachers, professors, and the general scientific community.
Business leader and marketers would suck your cock though.
>>55354450
>pathetic overclocker like gp104
Still preferable to nonexistant overclocking on fiji and the 480
>>55354461
>Except that's wrong
Most of the graph, from 0 to 1.0 would be red and green bars. The last bit to 1.4 to 1.5, or less than ONE-THIRD OF THE GRAPH IF IT STARTED AT ZERO shows the important information.
You sir, prove beyond question to be an unthinking retard and determined to stay an unthinking retard even after the obvious is explained to you.
>>55354436
100% (which the graph implies) vs 15% (claimed by the numbers)
is a large discrepancy.
I'm not critiquing the card, I'm criticizing the sneaky Jews in nVidia's marketing department.
>>55354431
the 6gb will be 299$
>>55354470
>I see you've never written scientific reports or papers.
Thanks for the classic ad hominem. Now do you have an valid argument to present?
>inb4furtherIcannotintoscale
>>55354431
That's the launch price. I'll be amazed if the 1060 launch price is below $300
>>55354406
They might as well have just given the numbers, instead of making a shit Fox News style graph like that.
>>55354490
I see someone majors in Women's Studies.
>>55354431
Pricing has already leaked out
3gb: $250
6gb $300
Not the same
>>55354493
>100% (which the graph implies)
The graph doesn't imply 100% improvements unless you fail at understanding baseline comparisons.
>>55354490
Just stop will you, we get that you don't get it
>>55354406
>Okay. Let's say that they start the graph from zero. They'd have more than half of the graph providing no useful information. That would be both a waste of time (for both the graphic artist and the audience) and a waste of space.
>You sir, are an idiot. You are defending a trivial objection that demonstrates a lack of understanding and sophistication.
>Omitting useful and relevant information is useful
>Graphs need to look sophisticated, not accurate. How else is the consumer going to make a decision
>Real shills using marketing tricks in a free market is trivial
WEW LAD. WE FOUND HIM. ITS THE PROGENITOR NVIDIA FUCKBOI.
Do you own a 970, and a 480 in your rig for PhysX? Do you cry yourself to sleep knowing that you will never have an Nvidia made CPU in your PC instead of Intel or AMD?
>>55354368
fixed
>>55354504
>Now do you have an valid argument to present?
>Your ass would get laughed at by teachers, professors, and the general scientific community.
To simplify. I have an education. You, obviously do not.
>>55354511
See >>55354504
>>55354509
Neither company is going to do that. Remember the "51% utilization" of two 480s versus one 1080 graph?
>>55354227
AMD's card full RAM card is cheaper.
AMD wins.
>>55354527
Oi vey look at all that non useful information, cut that graph at 0.8 quickly
>>55354527
And absolutely nothing was gained and everything was made smaller. Why cannot idiots into scale?
>>55354516
>The graph doesn't imply 100% improvements unless you fail at understanding baseline comparisons.
It implies it unless you look at the numbers more closely.
Marketers know a lot of people will just look at the title and glance at the bars... not even looking at the number below.
Literally 400% faster
>>55354544
What was gained is that the bars now accurately give a sense of the performance difference between both cards at first glance. Even to the untrained eye.
>>55354227
I thought power efficiency doesn't matter shills?
>>55354527
Yes. Thanks.
Now when you GLANCE at the graph, it is more OBVIOUS, nVidia is claiming 15% improvement.
>>55354480
Stay salty, nigger.
I'll see your tears when non-reference PCBs come out.
>>55354529
Thank you for making it clear you do not have an argument to make as all you can do is present ad hominem fallacy after ad hominem fallacy. You even get called on it and you still think that is going make any difference.
>>55354550
>It implies it unless you look at the numbers more closely.
Let's get this straight. You believe people just look at the pretty bars and not what the x and y tables represent?
>>55354572
Oh please, it's always 'Wait for this to come out' with AMD fags, and then when that comes, it's wait for the next thing
You idiots never learn
>>55354514
>3gb: $250
lamo nividjews are in full shekel mode
>>55354544
The whole point of graphs is that you shouldn't have to look at the number to get an idea of what it says, dumb ass.
When they don't scale the graph properly, it is done to mislead you. The only way to get the proper information, is to IGNORE the bars and look at the number... in that case, they should have just made a TABLE...
but they're using their Jewish tricks to con you into believing it's twice as fast if you just glance at it.
>192 bit
>3/6gb RAM
>no dx12/vulkan support
>graph starts at 0.8
Also, the benchmarks they did must be in some dx11 nvidia optimized cherry picked result, AMD is the clear winner and will scale MUCH better into the future.
>>55354582
And those non-reference PCBs are a week or two away :O
>>55354527
but you made the top bar longer and the bottom two shorter.
Performance should be at 1.1, VR performance at 1.3, and Power Efficiency at 1.5 according to the Nvidia graph
>>55354309
Where is dx12. Oh yes still being developed.
By the time its used in enough games the 1080 will be obselete. (well half of you lot will have them by then)
It's hard not to expect the 1060 to be stronger than the 480. It will also be more bang for buck. Unless Nvidia goes full jew on pricing, they should have this in the bag.
>>55354580
You're hopeless, fuck off to /shota/ or some other non-technical board.
>>55354563
>Even to the untrained eye.
It doesn't take any training to understand the difference between 100% and 115%.
>>55354605
>The whole point of graphs is that you shouldn't have to look at the number to get an idea of what it says
You sir, just won the award for the dumbest thing said in /g/ for the day.
>>55354572
That clock scaling is fucking beautiful, unlike Pascal which needs a 300MHz overclock for 10% 3Dperformance
>>55354616
Wait I read wrong, my bad. Your graph is going by twos while Nvidia's is ones.
>>55354646
I get the feeling AMD made GCN's pipeline TOO short and it's limiting their options.
>>55354580
>>55354544
>A bar chart or bar graph is a chart that presents grouped data with rectangular bars with lengths proportional to the values that they represent.
>with rectangular bars with lengths proportional to the values that they represent.
>lengths proportional to the values that they represent
https://en.wikipedia.org/wiki/Bar_chart
>>55354636
>You sir, just won the award for the dumbest thing said in /g/ for the day.
see
https://en.wikipedia.org/wiki/Bar_chart
>lengths proportional to the values that they represent
>>55354672
>Bar charts have a discrete range. Bar charts are usually scaled so that all the data can fit on the chart
Pro-tip: You might want to read ALL of your citation before you cite it.
>>55354694
See >>55354697 and enjoy your rekt.
>>55354514
>$250
Exactly like I said.
>>55354527
properly fixed
>>55354697
I don't think you know just how dumb you are.
What part of
>lengths proportional to the values that they represent
don't you understand?
Why don't you read this, first hit on my google search
http://faculty.atu.edu/mfinan/2043/section31.pdf
>>55354713
http://www.statisticshowto.com/misleading-graphs/
dumb ass
>>55354731
Thanks for the classic moving the goalposts.
But to humor you for a post:
Zero is clearly out of scale as 2/3rds of the graph is wasted space yet 0.8 is "misrepresentative" in your opinion. Then where do you suggest they rationally start the chart at? 0.5?
>>55354636
>>55354544
>>55354697
>>55354713
>>55354768
Here faggot
https://en.wikipedia.org/wiki/Misleading_graph#Truncated_graph
>>55354787
Let me make this clear. By your own citation scaling in bar graphs is a standard and accepted practice. The graph does not have to start at zero as was the initial complaint. So I ask the question again:
If 0.8 is unacceptable to you and zero is out of scale what would be an acceptable starting point on the bar scale?
>>55354768
You lost your argument, get over it.
>15% more Directx 11 performance than an RX 480 without proper texture compression and power management thanks to AMD's driver incompetence
So, in effect, you get 1 GB more VRAM for 50$ less from AMD and in a couple of weeks you'll also get equivalent Directx 11 performance while the RX 480 will utterly castrate the GTX 1060 in Directx 12.
Nvidia sure showed them. /sarcasm
>>55354820
Nice damage control
>>55354803
It's unacceptable because my citation says
>These graphs can create the impression of important change where there is relatively little change.
It would only be acceptable if the change is small.
15% change isn't small.
If nVidia was claiming, like .25% or 2% change, I could see that.
Your autism is showing if you can't grasp what is going on.
>>55354813
>You lost your argument
Your own citation proves my argument and you reject your own citation?
Thank you for proving you are the faggot with your closed-mind. You provide a citation that proves you wrong and think you have "won."
>>55354838
>It would only be acceptable if the change is small.
And I asked you where would the change be small enough in you opinion. Clearly you have not thought about where it would be clear to provide a change like the unthinking retard you are.
>15% change isn't small.
Even if they started the graph at zero the "15% change isn't small?"
I'm done extending you the benefit of the doubt that you can understand what I saying regardless of the formation of words and construction of sentences I present. That's almost as stupid as the previous line of "we shouldn't need to read what the graph actually represents when looking at a graph."
Ouch.
So wait, wasn't Nvidia supposed to be all about ignoring AMD and pretending they are not even worth their time and nobody has heard of them?
Why are they suddenly doing direct comparisons to 480? Did AMD hit a nerve there?
>>55354869
Now that would be gross misrepresentation.
>>55354872
>wasn't Nvidia supposed to be all about ignoring AMD and pretending they are not even worth their time and nobody has heard of them?
When has that ever been the case? The last few generations we've seen a pricing hierarchy based upon performance between the two companies.
>>55354889
>When has that ever been the case? The last few generations we've seen a pricing hierarchy based upon performance between the two companies.
I remember their statements around the release of 1080 pretending they are the only relevant GPU company.
>>55354872
They need to rev up the shilling because we all know that Nvidia's performance promises never turn out to be true.
By the time the GTX 1060 is available in stores (in six months or so) the RX 480 will have proper drivers and most games will run on Directx 12, making the RX 480 the 237% better buy.
>>55354908
>I remember their statements around the release of 1080 pretending they are the only relevant GPU company.
Source? I suspect your memory may be a bit faulty. Even when Nvidia released the Titan and the Titan X, uncontested fastest GPUs of their generation, there wasn't any implication of "AMD isn't even worth their time."
Ouch that must hurt
>>55354940
The problem is exactly that I have no source but remember something like that so I am a bit surprised with direct comparison to 15% marketshare guys.
>>55354943
>AMD performance is literally nothing
How can AMD even compete? How can they keep their business running? When will they finally sell the Radeon Group to Intel?
>>55354943
>>55354976
le hivemind
>>55354963
Given the history of pricing and gpu releases I'd say your memory is faulty. Perhaps you are remembering what some fanboi here posted in regards to Nvidia and misattributing it to Nvidia?
>>55354991
Pricing isn't really about marketing statements. While I am sure they are 100% aware of AMD I could see their marketing team pretending they are the only ones to guys who wouldn't even otherwise know of AMD.
>>55355007
>Pricing isn't really about marketing statements.
Am I understanding this correctly. It isn't what they actually do that matters as much as what you kinda remember them saying that matters?
I'm done taking you seriously if that is how you sincerely think.
>>55354582
They don't realise they're a living meme when they post this shit.
>>55354976
That's a fantastic graph according to this idiot
>>55354636
>>55354227
>1.1
>mutch
kek
>>55354908
>I remember their statements around the release of 1080 pretending they are the only relevant GPU company.
well thear not wrong lolol
Here AMD fanboys, a card that actualy performs like a 980 while being much more efficient than a 480
Basically Nvidia wins on all fronts, and AMD is the budget option once again
PAPERLAUNCH
>>55355660
considering that's 10% faster on stock clocks and the 480 only overclocks 1.5%, that is indeed a lot.
>>55354514
>3gb $250
jew lad
How shit do you think it's gonna be in DX12 /g/?
>>55355945
>hitman or ASOTS
Every single time
You do realize nobody actualy plays those games?
>>55355945
So the 390 is faster than Polaris even in DX12?
>>55355981
Not by much but nobody was expecting a $200 card to outperform a $300+ one. Also we get better energy efficiency with this new card so even people with 430W PSU will be able to use it.
>>55355963
>N-n-no one plays those games anyway baka
Alright. Whatever you say.
>>55356030
Do you play Hitman or ASOTS?
I don't
>>55355963
>>55356030
I don't think his argument is that no one plays those games but that those are niche games that aren't representative of the wider DX12 gaming industry.
>>55355981
Of course. It has more stream processors and all that. The architectural improvements puts them much closer than just looking at the raw core count would suggest but it still can't overcome that big a discrepancy. Anyone with more than two brain cells was saying that before launch.
>>55355945
http://techreport.com/news/30335/here-an-early-look-at-dx12-inside-the-second-benchmark-data
Worse.
It's full kek
>>55355963
>>55356043
The thing is, the humiliation is not so bad right now. It's only like 15 DX12 games out there. But as time goes on more DX12 titles will be released and nvidia will get shit FPS across all cards including pascal since they never implemented full hardware async compute and more salt will be poured on nvidia's wounds.
Hope you really really hate things like VR and all the new games coming out in 2016 and beyond.
>>55356082
The only games where AMD was able to keep up with Nvidia in DX12 were two AMD sponsored games.
What a fucking surprise
>>55356043
I play AotS, as do around 78,000 others. Don't know what this asots thing you keep referring to is.
>>55356056
>Hitman is now a niche game
Well, alright.
AMDKEKS ON SUICIDE WATCH
>>55354227
>He thinks the 1060 won't become an irrelevant gpu within a year of its released
If you thing the 1060 will be better than the 480 in the long term, you clearly know nothing about gpu's and their drivers.
480 is going to get massive driver spport for years to come. Coupled with dx12 optimization, whatever lead the 1060 may have initially will drop off significantly in time. This is what happened in the past, and will what happen again here.
Only a literal retard/nividia shill will prefer the 1060 to the 480 at this stage.
>>55356091
Get ready fuckboi, AMD is about to rape Nvidia when more DX12 games come out.
>>55356107
>I play AotS, as do around 78,000 others
>78000
What a hit game
>1060
>6gb
>$300
Good goy.
>>55356112
>damage control
Lol that graph scale 1 is almost 2/3rds less then 1.4 uhh?
>>55356124
Someone check me on this but SoM is not a DX12 game?
>>55356124
>AMD w-wil win, just w-wait till next year!
Yeah yeah, you faggots said the same shit with the 300 series, then with Fiji, and then with Poolaris.
It's always next year with you guys, wait for this wait for that. When is that moment actualy coming?
>>55356178
Well, for Fiji and the 300 series, it's now. AMD cards keep getting better with time whereas NVidia cards are dependent on constant driver optimisations for each individual game, which they stop making as soon as new cards are out. If you bought a 300 series you're far better off now than if you bought a 900 series.
>>55356210
>it is now
What is now exactly?
Those cards performing well in two AMD sponsored games but getting shat on in everything else while being vastly inferior in terms of power efficiency and overclocking capabilities?
Yes that moment is now, it has been that way with AMD cards for the past 4 years though
>>55356210
>Software vs Hardware based
This is very true.
>>55356210
> AMD cards keep getting better with time
Because they are shit at launch kek
>>55356210
Hell even my R9 290 is seeing some significant gains recently. Especially in DX12. Even Rise of the Tomb Raider which has a shitty 'non-full fat' implementation of DX12 gave me a huge boost over DX11. If they had done it properly I'd clearly be getting well over 60FPS constant @ 1080. I benchmark at 59.94 FPS average on High on DX12.
>>55356255
>being this mad
>>55356278
I ain't mad, I'm laughing, it's true that AMD cards usualy suck at launch, thus there is more room for improvement through drivers
> TWICE AS EVERYTHING
Didn't read the thread, but I'm pretty sure that shit have been pointed out.
I must admit the reference RX480 is a bit crap but then again I would never buy reference anything. That sweet looking Sapphire is only a little whiles away though so we shall see how it performs with 8 pin and proper cooling.
>>55356289
I love this AMD drivers are better meme, keep it up pajeet.
>>55356333
Did you even read what i said moron?
I was actualy making fun of AMD drivers
>>55356346
You are propagating the AMD drivers improve headroom meme because you are brainwashed by pajeets.
>>55356178
Just keep waiting ;_;
>>55356369
Their launch drivers usualy suck, and since last 4 generations of their cards all have the same rebranded architectures, optimizations also benefit their older cards.
So yes, there is more room for improvement with AMD cards. However, I'd rather have a card that performs well NOW, not in 2 years.
>>55356346
>>55356369
>Trolls fighting trolls
>>55354227
Price ?
It looks interesting if it is not above the RX480 in terms of price.
>>55356141
980ti have 6...fury 4
>>55356474
HBM though
>>55356439
Expect it to be a bit more expensive just like the last few generations. If it performs a little better it gets priced a little higher.
>>55356496
>HBMeme hype
And what about the 980Ti?
>>55354976
make the graph start at 3.5
>>55356569
What about it, the 980ti simply is a more powerful card, that doesn't mean HBM isn't superior to GDDR5
>>55356590
Wasn't the point about 6GBs being insufficient not about HBM being superior?
>>55355963
>mfw clocked over 200 hours in Hitman already
>>55356604
Yeah because that's one out of two games your card can run properly
>>55354227
Why aren't you guys showing the actual graph?
>>55356619
right now I have a 970, going to get a couple of 480s though
>>55354227
I actually use my GPU for more than games, and the FLOP/$ of AMD cannot be beat. Some of the deck slides AMD had which really werent showcased include the fact that GCN4 has native mixed precision and half precision compute, unlike Nvidia which needed to redesign their CUDA cores to handle it right.
A >6Tflop capable $200 card is right up anyone's alley if they aren't cucked by CUDA.
>>55356057
kek
>>55354227
Founders Edition only $349!
>>55354976
>>55356621
>>55356057
>>55354976
>>55354943
i love you /g/
So fucking fake. The RX 480 is going to curb stomp the 1060.
>>55354514
3GB $300 Founders Edition
6GB $350 Founders Edition
