what is nvidia going to do when they are RIP in a few months?
can nvidia even compete with THIS?
Welp, pack the bags nvidia users, it's over.
i don't get it why it would be better
>>52970139
wait, what? in the fine print it says they have 4790k with ddr4?
>>52970174
yes, they used an intel built to promote amd gpu.
>>52970134
They are going to get payed by me when i buy a new gpu.
I use Loonix, so i can't buy amd gpu's.
>>52970150
Literally same works for Nvidia. >>52970139
Almost same. Probably better - NV historically used less power.
Retarded bait, that's coming from someone who owns neither(no desktop atm).
>>52970156
It'll objectively be better than Maxwell.
Which isn't a big deal since anyone with half a brain should be comparing it to Pascal, which we know jack shit about.
Do you guys think this level of small form factor and power efficiency introduced by the double whammy of finfet and HBM2 will make multi GPU configs way more attractive? With this tech coming out and VR needing the GPu grunt, could we possibly see smaller, more manageable dual GPU cards right through the range from mid tier upwards? Maybe we can finally get decent software support for a glorious multi GPU future.
>>52970215
>NV historically used less power
>/g/, rewriting history since 2012
I honestly do think AMD has Nvidia's number this year, as evidenced by this video:
https://www.youtube.com/watch?v=nw-QA1BanJw
>>52970229
Yes it will make it more attractive because of less power consumption and thermal issues.
The same issues stemming from a lack of profiles in new games will remain though
>>52970232
You're telling me there are no more 2kW all-AMD rigs? And where low-power matters(dedicated laptop GPUs)AMD is non-existant.
>>52970261
I'm telling you that GTX 6xx is the first generation where nVidia used less power than AMD.
So yeah, unless you move the goalposts, you're flat out wrong.
why are 4chan x displaying random black bars?
>>52970244
I remain optimistic. If multi GPU becomes more popular and Moore's law continues to stutter (heh sutter, get it) developer will be forced to cater for multi GPU and perhaps even develop for it as the primary rendering solution.
>>52970156
From what I gather AMD snagged the smaller process node over Nvidia (14nm vs 16nm) and they also have a year's worth of experience working with HBM. Furthermore, their HBM2 Polaris cards are slated to launch on schedule, whereas Pascal has been revealed to have been faked and that the first Pascal chips will use GDDR5X and not HBM2 to push out faster.
This all means that AMD will have their cards out first.
>>52970301
Guess I finaly have a reason to pick amd over nvidia this year
Nvidia is in for asshurt kek
http://videocardz.com/58237/amds-project-f-is-232mm2-discrete-gpu-made-in-14lpp-process
AMD doesn't even have a high end GPU this year, 232mm2? Enjoy your Pitcairn.
But keep on lying and FUDing if it makes you feel better
The only company that is RIP is AMD, they're gonna announce another massive loss next results day
>>52970233
>Believing AMD shills lies
TOP KEK
>>52970339
The biggest liar is Nvidia dumbass
They lied about the 970's specs and they lied again when showing Maxell gpu's as supposed Pascal gpu's in their last conference.
Keep eating that Nvidia turd though
>>52970346
You're the one thats eating Polaris turd lies
April 5th is GTC 2016 Opening Keynote :^)
>>52970368
Except I actualy just showed you specific points at which Nivdia was lying.
All you can do is yell "HURR AMD LYING KEK" like a fucking 12 year old kid.
Show me where they lied then
STAY MAD, AMDPOORFAGS POLARISTURD
>>52970377
I can't wait for them to show more gpu's that aren't actualy Pascal!
>>52970386
STAY MAD, AMDPOORFAGS
YOU CAN'T EVEN COMPETE, 232mm2 GPU MID RANGE GPU POLATURD
>>52970407
Calm down kid, if you want to know I actualy have a 780ti in my rig right now.
That doesn't mean I blindly follow any type of brand. Unlike idiots like you who start foaming at the mouth anytime anyone shows some valid criticism on their brand of choice.
>>52970419
Shut up AMDPOORFAG, you don't own any Nvidia products
>>52970437
Kid, don't talk about money, you probably still leech of your parents in a dank basement.
Come back when you can whipe your own ass idiot
>>52970447
AMDPOORFAG, you don't even earn anything but mooch off your parents basement
>>52970134
Probably release the 4 year old tech they have haven't been forced to release yet to counter amd. Then research new tech because they are only 3 years ahead.
>>52970458
I mooch of a basement?
Nice English there Pajeet
>>52970139
>compared using FRAPS
Is AMD so poor that they can't buy a dedicated recorder?
>>52970468
Pretty much
Also notice they used an outdated driver for Nvidia, pure desperation there
>>52970468
Wouldn't they just log the performance with FRAPS ? Recording would have an effect on performance, bigger than logging framerate.
That said, I'm willing to bet they use FRAPS as a way to say "we did it using 3rd party tools, honest !".
>>52970437
Stay mad kiddo
>>52970496
was it outdated for the time they were benchmarking this test?
>>52970500
Didn't think about that. Might as well have used Steam's frame indicator.
>>52970542
It's an EA game though.
And it probably doesn't offer anything to log performance over the course of a benchmark.
>AMD
I dunno why people argue about which card is what, just get what you think looks nice. If AMD has the best card available, get their shit. If Nvidia, do the same.
>>52970205
4790k is ddr3 only, it cant use ddr4
>>52970139
>GTX 950 on medium preset at 1080p consumed 140W
wat. maybe the whole system.
>>52970509
>3GB vram
JUST
>16GB DDR4 2600
>4790K
>>52970605
Intel has a history of underreporting its electricity usage while AMD the opposite. This doesn't become evidence unless under load.
>>52970658
You mean Nvidia. And no they don't, Anandtech and numerous other sources have tested the tdp W of GPU's and besides the 950 only has a single PCIe connector.
AMD is going full retard now.
>>52970134
>>52970139
>AMD doubling down on gaymer shit
No one important cares.
The reason why AMD is losing marketshare is because no one in production-related jobs respects your GPU anymore.
>>52970301
AMD will have 14LPP which is effectively 16nm. Get your facts straight you fucking 12 year old.
>>52970648
Holy fucking shit AMD is retarded, also
>same performance as a 950
>Even AMD uses Intel CPUs
>>52970643
when that card was released I don't think 8gb framebuffers were really that popular
>>52971129
because their current cpu's are very old and cause bottlenecks
>>52970787
>a goddamned 5750 from the stone age
>trying to use ray tracing
>>52970605
It is the whole system usage in both cases.
Did none of you fucking read articles concerning this? They're pretty important.
>>52971129
You'd have to be retarded to use AMD processors. Not even AMD is that stupid.
Especially when your GPUs are actually as fast or faster than the competition, you're not going to sit around ruining your benchmarks with a shitty CPU.
will this mean we can finally have high performance single slot cards again?
This will be an exciting year.
>>52971486
Maybe in 2 years.
>>52971486
No, because games (or shitty ports) have never been so unoptimized as they are today.
Fury x2 when??
>>52971598
When nitrogen cooling appears
>>52971605
>>52971605
That is just plain retarded, you'd have to buy new tanks to replace old ones everyday
>>52971646
but it's cheaper! and goes faster!
Didn't read the thread. So what you've concluded? Who is retard?
>>52971701
>Who is retard?
People who conclude which device is best before it's been released and benchmarked.
![=_utf-8_B_Njg3cHgtVHJvbGxpbmcxMDEuZ2lm_=-763212[1].gif](https://i.imgur.com/c6wsBZw.gif "=_utf-8_B_Njg3cHgtVHJvbGxpbmcxMDEuZ2lm_=-763212[1].gif")
>this thread
Though, I am genuinely exited how AMD/nVidia update their architectures aside from HBM.
>>52970139
AMD is known for inflating the value of their GPUs. This test was likely performed in an AMD lab under ideal conditions.
>>52971762
https://www.youtube.com/watch?v=9jlwvuqzhjg
>>52971757
Nvidia is going to be adding new Antialiasing modes and an improved lighting engine, probably more VR hardware goodies too. AMD is going to add a geometry crusher so they will finally be able to do tessellation.
>>52971808
Staged as fuck
>>52971845
>tessellation
Yet another useless gimmick.
>>52971856
sure, as useless as hbao.
which will become obsolete with vxgi by the way.
cant wait for amd to cry just like with hbao.
>>52971856
It's not useless per se, but developers are lazy and use it for retarded things. Adaptive tessellation is only supposed to intensify around object edges, since throwing polys at front-facing surfaces is retarded, and also dynamically scale with distance. As far as I'm aware, Gran Turismo 6 is the only game that does this properly, and it ran on a fucking PS3.
>subhuman trash can't even get more actual performance so now everything is about performance/watt
>>52971878
Hardly vxgi is very intrusive when it comes to rendering compared to HBAO.
I am sure nvidia companies like cd project will give it a shot but I can't think anyone else will. HBAO can be implemented fairly quickly so developers put it into their engines as a pc "goodie". On the other hand vxgi won't run on this generation of consoles and neither on the next (probably AMD as well). Thereby making heavy changes to how the engine calculates light is not very economical if only a part of pc users can use it anyway.
>>52970643
I run games that supposedly hog up vram fine even on 1440p.
Shadows of mordor for example runs beautifully all high and ultra settings 1440p.
So yeah, not really bothered
>amd
https://www.youtube.com/watch?v=8lTGk0R12gs
Tbh it's all it takes for me to switch to AMD GPUs is that I'm not required to buy a new fuckhueg PSU twice the price of what I save from switching to AMD from nVidia
http://wccftech.com/amd-polaris-gpu-die-size-232-mm-2/
AMD IS FINISHED AND BANKRUPT!!!
>232mm GPU
>2016
>290X performance in 2016
CHAPTER 11 INCOMING!!
>>52972493
290x performance in dx12 games equals 980 performance, if they can bring that to a lower price levels, Nvidia is fucked.
Not to mention, that will be their low/mid tier card, their high tier card with 14nm and hbm2 will make mince meat of the titan x
>>52972628
That IS their highend card, AMD doesn't have the money or expertise to make a big chip anymore, it'll have to settle for selling midranged GPUs as its flagship while Nvidia sells 600mm GPUs that stomp it in performance
>>52972645
So what you're saying is that you're assuming whatever makes you feel better?
>>52973151
I'm saying that which is derived from semiconductor knowledge and logic, you can only throw at hominems at me since you lack both.
>>52973334
>semiconductor knowledge and logic
ayyy lmao
>>52972645
>>52972493
>shitposting the exact same thing in every thread
That die is comparatively larger than Fiji. Its newer arch, newer PowerTune IP.
Shitposting is against the rules on all boards outside of /b/. Go back to where you belong.
>>52973354
>>52973381
>you can only throw ad hominem
Case in point
>>52970134
Probably the same thing I am going to do after I die
>You question is stupid
>>52972224
what is VXAO.
>>52970207
>Linux
>can't buy AMD GPUs
Its not there yet for all games, but situations where the open source driver beats Catalyst and is competitive with the Nvidia driver are increasing.
Polaris will be officially supported by AMD using the open source stack.
>tfw want a Fury X
>more news on X2 announced
>still want to wait for Zen
>>52970590
most likely a typo
>>52971853
not staged, one is running a 14nm gpu one is using a 28 if i remember right.
also, the amd one could be a high end gpu that is running a measly 60 on medium settings, whereas the nvidia card is stressing the fuck out to pull the same frames.
>>52972183
Those are equivalent, getting more perf/watt in the end means better performance in the ~300W power envelope high-end GPUs have had for years.
>>52973855
>>also, the amd one could be a high end gpu that is running a measly 60 on medium settings, whereas the nvidia card is stressing the fuck out to pull the same frames.
Did none of you read anything? AMD showed the used Polaris GPU, it's size is around 140mm^2
That's a Cape Verde sized GPU, ergo low-end.
>>52973855
The AMD GPU running is their entry level Polaris.
Performance is somewhere around a Radeon 7870 - 7950.
>>52970297
>>52970244
>>52970229
multi gpu will not be a viable option till vulcan and dx12 come out and we know that those api's have their shit together.
currently multipgu have to many points of failure to be more attractive than a single overkill gpu.
>>52973757
>more news on X2 announced
You do not want to open up your ass to the Crossfire "support" assrape from AMD, trust me.
>>52973886
no i watched the video that was linked, where is the reading material that tells us 140mm is the low end?
>>52973943
>In any case, the GPU RTG showed off was a small GPU. And while Raja’s hand is hardly a scientifically accurate basis for size comparisons, if I had to guess I would wager it’s a bit smaller than RTG’s 28nm Cape Verde GPU or NVIDIA’s GK107 GPU, which is to say that it’s likely smaller than 120mm2. This is clearly meant to be RTG’s low-end GPU, and given the evolving state of FinFET yields, I wouldn’t be surprised if this was the very first GPU design they got back from Global Foundries as its size makes it comparable to current high-end FinFET-based SoCs.
http://www.anandtech.com/show/9886/amd-reveals-polaris-gpu-architecture
P-please... don't buy AMD...
>>52973972
>trusting Anandtech's word
>when it's well spread knowledge they shill for AMD whenever possible
This pretty much confirms this is a large as fuck GPU(600mm^2)+ underclocked and framelimited to 200Mhz
In short, nothing to write home about, guess there's no point in waiting for Polaris when it'll be both slower and less power efficeint
>>52971129
It's only common sense when trying to market a GPU, that you remove as many bottlenecks elsewhere in the system, so that the full potential of the GPU is realised.
You say AMD are retarded for doing this, but it's quite the contrary. AMD are intelligent for using such a processor, as nobody is naive enough to get caught up on the fact that AMD used Intel parts for benchmarking, as opposed to their own 3 year old parts.
>>52971429
amd had to bite the bullet on everything they show for a while, no high end gamer wants an amd cpu, and to best show of their gpus intel is better.
>>52971486
you want the gpu to be over cooled rather than under, so i would rather have dual slot even if it was a weak as fuck card.
>>52974008
Wow the level of butthurt here is unreal.
You have to be fuming mad and retarded to think Anandtech is ever favorable to AMD. Even when AMD sponsors Anandtech to get advertisements on the front page they go out of their way to skew their benchmarks against AMD.
>>52970787
But more and more software is utilising OpenCL/GL now, which AMD cards have always been better with. After Effects has also had OpenGL support on CC for ages.
>>52973734
The open-source xorg drivers (radeon, radeonsi) are very IPC dependent: with Sandy HD6850, 6870 and R9 270 has nice performance, but with Piledriver, not so great.
Nvidia blob on the other hand is nowhere near as dependent on IPC and Nvidia have massively better gl libraries (this has been true for the last 12 years) compared to either mesa or AMD's closed libraries.
>>52974018
>and to best show of their gpus intel is better
Especially with their dogshit CPU-limited driver.
>>52974008
Shill.
>>52974049
>AMD
>Better with OpenGL
This is bullshit, on both Windows and Linux (catalyst or mesa).
>>52971845
they already tessalate, however nvidia wants to push sub pixel tessellation to fuck with amd and give their old gen cards a hard time so people upgrade.
>>52971856
well think of it this way, anything round could easily eat up 10-20000 pollys to make it look real fucking round, the closer you are the more detail it gets the further away the less, you see this in deus ex hr, however it has such a low lod distance that you literally see it happen in real time, ideally it would be far enough away you don't.
you can also add high levels of detail close up on top of a displacement map that has a good effect too.
>>52971997
tessellation gets abused way to much.
it should add detail on top of a displacement map, but usually it just makes a box a solid color wireframe.
>>52970407
Welcome back, Nvidiot. Been a while since I've seen you about.
>>52974049
are there any video editors that will be accelerated with a 280x?
my old 5770 would accelerate video encoding, but my 280x just doesn't work.
>>52974129
Anything on Creative Cloud has OpenCL support now, even After Effects. I use a 270X at work and AE runs great.
>>52974303
So? You wrote that
>OpenGL, which AMD cards have always been better with,
which definitely is not true, quite the opposite. Nvidia had and still has massively better OpenGL performance.
Who would be stupid enough to buy AMD when GameWorks exists?
>>52970215
>what is Fermi
>>52974409
To be honest most features added by gameworks is useless shit that often doesn't add to the look of the game; hair werks is a hog and manages to look worse than traditional hair, tessellation looks and works like a 3D modeling student's first attempt, and Physx does look cool but zero games actually make serious use of it other than add wow particle effects that blind you and make it harder to actually play.
>>52974457
And don't forget FX 5900, the original housefires GPU.
>>52970327
Lurk moar you faggot
24/7 amd vs. nvidia
don't you faggots ever get tired of this?
and the retarded ass assumptions here always
>they will BTFO them
>no they will BTFO them!
they will just continue bringing new shit until the end of days
there's nothing like a monopoly in this world anymore, too strongly regulated
both companies will survive the war
deal with it
>>52970437
Gimped as fuck
>>52971856
>Nvidia shillworks 64x tesselation
>>IT'S NEEDED GOYS, NOT OUR FAULT AMD CANNOT INTO IT
>AMD solves tesselation issues
>>IT'S A GIMMICK YOU IDIOTS
Fucking shills...
>>52974107
That may be, but they're far for unusable with an AMD CPU. I have an R9 280 and a FX 6300 and my scores in benchmarks and framerates in many games are still better than my old 650ti and in benchmarks like Unigine Valley at least my scores are very similar to what users with 960s report.
Some stuff like tessellation and games that use GL 4 features like Bioshock Infinie are still slow but it plays most games fine and handles 2160p desktop perfectly.
>>52973757
>>more news on X2 announced
yeah, the news is that Scott wasson trolled the everliving fuck out of the rumor mill.
>>52974748
>nvidia
>the feel when I'm currently heating my apartment with my 380x and the new gen will take that away from me
feels cold man
>>52974494
The noise.
THE NOISE HOLY CRAP
