>AMD cards aged better than nvidia No, CPUs just got stro

>amdcucks on suicide watch

>>55690370
All this tells me is that AMD needs to steal nvidias driver team since their pus are just straight up better than nvidias.

File: 14691199463281.png (498KB, 1920x1080px) Image search: [Google]

498KB, 1920x1080px

>>55690449
Why do you think GCN is seeing such a performance gain on vulkan and dx12.

somehow they're still behind nvidia on weaker hardware

File: amd.jpg (138KB, 653x726px) Image search: [Google]

138KB, 653x726px

>>55690370
>>55690496
You have 10 seconds to delete this.

>>55690370
I love this line of reasoning because it doesn't actually change the fact that AMD performance improves over time.

>>55690717
It doesn't matter when it's worse than old gpus anyway.

File: Doubt.jpg (16KB, 200x303px) Image search: [Google]

16KB, 200x303px

>>55690717
Yeah because everyone with an AMD card is using the latest i7.

When you buy AMD, you won't even get close to the performance you see in benchmarks, even a haswel i5 is a bottleneck

>>55690717
>buy new cpu instead of gpu to bypass amd incompetence
amdornes everyone

>>55690717
>Be AMD fag
>buy the latest $300 i7 every 2 years instead of a $300 GPU every 4 years
No wonder they're so poor

>>55690761
>>55690761
I got a 4770k in 2013 that I won't need to upgrade. Where was it shown that a 2600k is worse than a 6700k like in OP's pic? It isn't in the link.

File: 1467512034793.jpg (75KB, 642x582px) Image search: [Google]

75KB, 642x582px

>>55691007
>he thinks an i7 2600k is just as good as an i7 6700k

File: unity.jpg (149KB, 1234x637px) Image search: [Google]

149KB, 1234x637px

>>55691066
When it comes to games, yeah pretty much. 6700k is better but you can close the gap quite significantly with OC.

>>55691087
>showing an nvidia only benchmatk
I guarantee the result will be very different with an AMD GPU.

>>55691361
Well I was kinda hoping someone would actually post something similar with an AMD GPU which is what I was asking for. I think Digitalfoundry also did something similar with a 290x but I can't find it. I don't remember the results being that off.

>>55691385
Yeah just had a look, they're all either 970, 980 ti, or Titan x ffs.

>>55691385
https://youtu.be/xhuC8Tf9i3I

>>55691087
This game is very GPU intensive, that Titan X is already at max GPU utilization, not even an OC GTX1070 can maintain consistent 60FPS at 1920x1080 max settings on this game.

This is a poor example and even Digital Foundry mentions it.

>>55691705
You're free to look at the other results. I just had the pic off hand from another thread. The gap is small except for a few outliers.

File: maxresdefault (4).jpg (336KB, 1920x1080px) Image search: [Google]

336KB, 1920x1080px

>>55691705
There's other results too with the Titan x showing much better results.

That being said, it should be common knowledge that GPU bound games, like bf4 for example, will be affected less by cpu than CPU bound titles like GTA v or Fallout 4

I know this is off topic slightly but what kind of performance can we be expecting from vega? Amd released the 480 before the 1060 which gave nvidia some time to refine the card to make it beat the 480. Now that the 1070 and 1080 are already out, should we be expecting faster performance due to amd's ability to change up clock speeds etc to beat the higher end nvidia cards?

This is just my opinion but I really can't see amd suddenly pulling out a couple of cards which are faster than the 1070/80 like nvidia did with the 1060. Not unless it's 3xx series levels of heat and power consumption.

File: 2500k.png (46KB, 637x693px) Image search: [Google]

46KB, 637x693px

I think OP is just pulling shit out of his ass to be honest family

https://www.computerbase.de/2016-06/radeon-rx-480-test/7/#diagramm-anno-2205-1920-1080-intel-core-i5-2500k

>>55691804
GTX 1080 at best.

>>55690370
so if i have a 4770k, i'm better off with nvidia?

>>55690370
>No, CPUs just got stronger, helping alleviate the horrible driver overhead on AMDs side.

In other words they're better now than they were. As in, they're aging better. Deal with it Nvidiot.

>>55691974
>dumb amdrone

File: untitled-1 (1).png (20KB, 682x562px) Image search: [Google]

20KB, 682x562px

>>55691851
Nicely done.
Holy shit you're reaching a new level of autism.

I just wasted 5 seconds of my life to find out just how CPU agnostic this game is, in which hilariously a gtx960 is not only ahead of the 380x, but only 4 frames behind a 390.

The only one pulling shit out of his ass is you, and you're reaching in deep because you're running out of shit to post.

http://www.guru3d.com/articles_pages/anno_2205_pc_graphics_performance_benchmark_review,9.html

File: Beautiful Laughing Ladies.jpg (39KB, 540x360px) Image search: [Google]

39KB, 540x360px

>>55691974
>implying AMDumbs are rich enough to keep updating to the latest i7

>>55690370
So provide some proof of your theory. Post a comparison between the 7970 running on a 2600K and a 6700K at the same clock speed. You've been spamming these same handful of images here for literally weeks and have added precisely nothing to your shitposting in that time.

>>55691974
I can't tell if you're legit retarded or just trolling.

The point being made here is that whenever older amd gpus are being tested to see potential performance gains, it's always with the latest i7 the reviewers have so there's little change of cpu bottlenecking. In reality it's not "aging better" but cpus are becoming stronger and stronger which help out with amds major driver overhead. If you want to see this "aging better" you're gonna have to buy the latest and greatest i7 every year.

File: 2500k.png (32KB, 609x684px) Image search: [Google]

32KB, 609x684px

>>55691985
lamo, that was just the first graph that came up. There's others there, did you even look?

I'm honestly startled how it seems you've managed to get an aneurysm over something I just came across and decided to post. I honestly don't care that much.

>>55691361
>I guarantee the result will be very different with an AMD GPU.

spoiler: no significant difference

http://www.anandtech.com/show/9483/intel-skylake-review-6700k-6600k-ddr4-ddr3-ipc-6th-generation/15
http://www.anandtech.com/show/9483/intel-skylake-review-6700k-6600k-ddr4-ddr3-ipc-6th-generation/16

>>55692026
>>55692031
OP BTFO

>>55691985
You shouldn't be trusting computerbase anyway. They're unreliable.

>>55687203
>>55687267
>>55687294
>>55687398
>>55687863

>>55690370
So wait a fucking minute, an i7 really is best for gayming?

/g is so full of shit

File: Screenshot_2016-05-12-18-03-32.png (2MB, 1920x1080px) Image search: [Google]

2MB, 1920x1080px

>>55692026

File: CPU_01.png (93KB, 1299x2173px) Image search: [Google]

93KB, 1299x2173px

>>55692026
>literally posting every cpu agnostic title he can find
PretendingToBeRetarded.png

>>55692060
Only if you're runnung ayymdpoor gpu.

>>55691087
Oh fuck off, I can go and grab a screenshot that shows +25fps on that same video.

Why do you have to resort to lies and half-truths?

File: TheBigPic.jpg (604KB, 1055x1765px) Image search: [Google]

604KB, 1055x1765px

>>55692015
This is an interesting read if you're interested.

http://www.babeltechreviews.com/hd-7970-vs-gtx-680-2013-revisited/

File: 2500k 2.png (33KB, 623x694px) Image search: [Google]

33KB, 623x694px

>>55692080
There's only 4 to chose from bro. This is the one with the widest gap if it makes you happy.

Givin me a giggle how invested you are into this theory of yours.

>>55692081
Resorting to insults is pathetic

File: the_sound_of_fried_chicken.gif (2MB, 185x132px) Image search: [Google]

2MB, 185x132px

>>55692080
>amdrones must get either amdpoor gpu+nvidia gpu or intel cpu+amdpoor gpu if they want playable fps
>retards running ayymdpoor cpu+gpu are literal cucks

>>55692096
That isn't interesting at all. Correlation does not imply causation. I asked you to provide evidence of a test carried out on two different CPUs back to back or side to side, not one set of results from 2012 and one from 2016, with you reading into them whatever suits your argument. Not an article which has nothing to do with the subject at hand, other than you putting your own slant on its results to suit your argument.

>>55692110
You do realise your image is showing a bigger performance loss on the AMD cards than the nvidia cards, thereby proving op right, don't you?

>>55692060

in general no but it works wonders on cpu hogging games (fallout 4 comes to mind)

>>55692060
Pretty much, especially for cpu intensive games, or if you're using an AMD card.

>>55692172
They are still very playable, and you won't even notice for crisakes.
Is it really surprising an i7 is a stronger cpu to you?

>>55692172
Do you expect amdfags to be able to read graphs? The same who post that tom slide showing 960strix having 50W average draw and 480 80W.

File: 1363314622001.jpg (94KB, 563x563px) Image search: [Google]

94KB, 563x563px

So would an i5-4690 be bottlenecked by an AMD card?

I have a GTX 770 and I've been thinking of snagging an RX 480. I've never really thought about CPUs before. My 4690 doesn't seem to run into problems.

when you go on the offensive, you show your hand
you do not feel secure with the choices you've made

>>55692172
Oh sure but there's other factors like 4 threads vs 8 threads. Maybe it's not that CPUs are "better" but AMD using more cores. Like I said I just came across it. I really have no beef because I buy parts I need for work and that will almost always land me on buying a Intel CPU and an AMD GPU for other factors like rendering.

>>55692199
>muh 5 fps

You wouldn't even notice in real life

>>55691851
>>55692026
>>55692110

These graphs are deceiving. It makes you think theres no difference between a 5 year old i5 and a recent i7 when there's a huge gap especially in cpu intensive scenes by upto 30 fps.

https://youtu.be/9lCq8Qk9wjY

This video shows this. It's almost the same when the cpu isn't being stressed but when they go to the road with all the cpu heavy effects then the i5 dips below 60 fps whereas the i7 is at like 90 fps. Don't forget this game is very optimized for all types of cpu too.

The people at computerbase are probably too stupid to test cpu heavy scenes in games which would show a big difference. Most cpu perform roughly the same in most games anyway unless, like I said, it's cpu intensive scenes. Even an 8350 performs just as well as an i5 in games that aren't taxing on cpu.

>>55692200
The point is amd depends on cpu performance whild nvidia doesn't. People upgrading their gpu are getting better performance with new nvidia card than amd.

>>55692215
Do you think you will ever get rid of bottlenecks?

You have an older locked cpu.

Of course it won't get the same frame rates as an 6700k or whatever.

>>55692172
Oh and the main thing I'm questioning is that AMD hasn't improved their drivers. OP provides no proof of that and claims it's just CPUs being the only factor. AMD for sure has more driver overhead than Nvidia.

>>55692200
If AMD is supposed to be the "budget" brand, why are they so underwhelming on budget hardware?

>>55692240
For what? >5%
And I get to pay more to nvidia for that honor?
So a newer cpu performs better than an old one?

Is that the whole point of this thread?

>>55692259
Shut up nvidiot.

>>55692163
Literally a Google Search away, people have done tests with nvidia and AMD cards on weak as shit CPUs.

>So i asked for a friend of mine to lend me its GTX 770 in order to benchmark, it was not the best decision since the GTX 770 is far more powerful than my HD 7850, but with MSI Afterburner it was easy to see when my GPU is getting bottlenecked, so thats why i choose saints row IV, project cars and dying light and Star swarm, since they some of the games that my GPU usage are always far away from 99% which means i am bottlenecked by something.

https://www.techpowerup.com/forums/threads/amd-high-overhead-drivers-on-dx11.209726/

>>55692259
>If AMD is supposed to be the "budget" brand
they're not tho. look at last wave of cards

they don't compete on the ultra-high end market like the gtx 980ti, but they were still good in the upper bracket of tier of cards. the r9 390 competed very well against the gtx 970. i don't think it's right to consider cards in the 300-400 dollar range as budget.

>>55692259
underwhelming?
Each and every graph posted still has fps above 60fps, and below 144fps.
Are you so sure multitudes of players are going to notice the 8fps or whatever difference?

>>55692259
>buy a budget cpu, get budget cpu performance
Rocket medicine, folks

File: 1456616499579.jpg (158KB, 720x480px) Image search: [Google]

158KB, 720x480px

>>55692317
>buy a budget cpu
>buy budget nvidia gpu
>get advertised performance

>buy budget cpu
>buy budget AMD gpu
>too bad, should have gotten an i7 extreme Edition goy, what are you, poor?

>>55692026
>>55692110
>tfw on 2500k
i actually overclocked it at last after five years of usage

>>55692354
Yes Goy, buy this Nvidia card for only 20% more!

The 2% difference in frame rates will make your gaming experience unbearable!

Buy Nvidia, The way your meant to be played!

>>55692399
>no arguments

>>55692354
dumb nvidiot

>>55690761
Bullshit I get 5500 in firestrike on my AMD card same as most anyone else. My results are aftually in the top 20% for legit benches and i use an 8 year old cpu.

File: Screenshot_2016-07-22-11-44-25_com.android.chrome_1469151884125.jpg (77KB, 719x696px) Image search: [Google]

77KB, 719x696px

>>55692399
It's sad that redditfags are smarter than AMDumbs...

>>55692406
Literally nothing wrong with AMD gpu's.
They are essentially equal in performance.
It's why you shills have to resort to making a 8fps difference a huge deal.

OH MY GOD 72fps vs 79fps AMD IS DEAD

You realize you sound ridiculous, right?

>>55692433
OH MY GOD -7% IN SOME RANDOM GRAPH NOBODY WILL NOTICE

>>55692432
>gpu only benchmark
>has nothing to do with cpu

>>55692238
Every time I press pause the i5 has a faster fps.

>>55692437
79-72 != 8 you retard

>>55692478
Actually every time I press pause there is a difference of 1-2 fps between the i5 and i7 so not much difference in the Intel scores. The FX is shit though.

holy fuck amdcucks have been literally blown the fuck out
also im going to pretty much die of laughter when amd plummets again after so many plebs bought stock @ $5.00

>>55692503
>retarded nvidiot

>>55692433
it's even more sad just how much AMD fucked over the nano/fury launches 2bh

>>55692512
They could have at least dropped it by 50mhz and undervolted to make it more appealing from a power consumption/thermal standpoint.

File: never amd.png (351KB, 1270x708px) Image search: [Google]

351KB, 1270x708px

>>55692296


>8 fps

>>55692531
DELETE

>>55692490
Yes, point out the glaring error and ignore my question:

Do you really think 5fps or whatever is going to make game play vastly improved?

I don't think anyone's going to notice a difference outside these charts.

And anyone that would notice something like that already has the best equipment possible.

Only thing I have learned is:

i7 > i5
old cpu < new cpu

>>55692531
when will we get some charts from a consumers point of view

>>55692548
>start at 8
>drop to 7
>now at 5
No, the -69fps difference won't matter at all!!!

>>55692531

Only a retard would buy a 6300, this proves nothing.

>>55692522
They're really just making weird choices all around quite honestly. It's as if they just don't understand what the fuck people want, and they've talked themselves out of trying to appeal to normfags.

>>55692548
If that 8, 7, 5, or -69 fps Is The difference between maintaining a smooth 60fps or not then yes I'm sure a lot of people would be bothered.

File: eienlol.jpg (87KB, 346x347px) Image search: [Google]

87KB, 346x347px

>>55692568
>nvidia works better than amd even with amd cpus
CAN'T MAKE THIS SHIT UP

What is it with AMD and their 980 comparisons. Wouldn't it benefit you to just go the Nvidia route and say "this shit is 300x faster than our previous line" instead of just getting fucked through all these benchmarks?

>>55692531
How is it that AMD gpus suck so fucking much with cpus that aren't 600 dollars?

According to OPs logic shouldnt old Nvidea cards also age better?

>>55692565
The graph OP pictured has many combinations if you will look closely. I am also talking about all the other graphs posted in this entire thread.

I am really unconvinced this will make any real world difference. You still have not really convinced me such a insignificant figure is worth all the hate. It's like there is a YUGE performance hit from all the shilling. In reality, its really nothing.

So:

i7 > i5
old cpu < new cpu

really surprising to you?

>>55692568
6300 was actually smashing value as long as you didn't use radeon graphics

>>55692592
The 980 is THE price to performance card of the last gen

>>55692031
lol based. op kys deceitful fag

>>55692635
>970 loses 13fps going from 6700k to 9590
>390 loses 28 fps going from 6700k to 9590

No matter how you look at it, that's fucked up. It's more than double the performance loss on AMDs side compared to nvidia.

>>55692601
They work fine nvidiot.

>>55692591
Are you so blinded you didn't notice a $200 card was damn near as good as Nvidia's $650 card six months ago?

A 6300? It's old, and it's shit. Is this surprising?

>>55692601
The shilling is real holy fuck

File: Screenshot_2016-07-22-12-10-27_com.android.chrome_1469153459127.jpg (61KB, 720x523px) Image search: [Google]

61KB, 720x523px

It's sad because AMD graphics cards are actually pretty good, but they're held back by horrible drivers.

>>55692354
I do have an i7 extreme edition tho :^) I can get intel dealer pricing's. I paid $100 for a 3930k back in 2011/12

>>55692669
Why do you always pick the shit shit shit AMD processors?

Look and tell me what the difference is between the i5's

>>55691087
>you can close the gap quite significantly with OC.
>implying you can't OC the 6700k

>>55692724
Why is radeon so shit with shit processors?

>>55692724
It's still 7fps, and when you're already under 60fps, it becomes easier to notice.

Mins would tell the true story, unfortunately that graph only covers average fps.

>>55692743
Buy a shit processor, get shit performance.
You poorfags always like to cheap out on the cpu anyway, always recommending buying a better gpu. You have to have both for a good system. Is all you do gayme on your computer anyway?

>>55692743
Why is shit so shitty?

If the nitro+ is a blunder im gonna get a 1060 lads. So will you

File: 6lvSGoE.jpg (230KB, 1920x1080px) Image search: [Google]

230KB, 1920x1080px

AMD fx processors aren't that bad...

>>55692773
Anyone that would care about 7 fps is going to buy a k processor, or an i7.

>>55692775
People buying i7 are getting gtx1080 not your shitty radeon. Stay mad amdrone.

OK here's the deal. I've just been running Star Swarm on my AMD setup (R9 290 and i7 4770K).

Using DX11 the frame rate was up and down like a hookers knickers and a couple of the CPU cores were being overloaded causing bottlenecking.

When I switched to Mantle the load was nice and even over the cores and frame rates were ultra smooth with a straight line and only very slight dips on occasion.

So my question is this. If AMD GPU's can't perform well in DX11 and gets bottlenecked by certain games that rely on the CPU then that's a weak ass GPU. If you are paying for a GPU to take the load off the system it should be doing so and if you are paying around the same price for a GPU and one is just offloading a lot of the work to the CPU in either DX11 or DX12 then it makes more sense to buy the GPU that is doing most of the work.

Does that make sense to anyone?

With all these AMD vs nVidia threads, I feel like I'm the only person in the world who bought an R9 Nano.

Pretty happy with it actually. For whatever reason it's quiet as fuck in my computer and I got it at a rather nice discount, brand new ($325 thanks to Jet + Amex cashback stuff, back during the price drop to $400 beginning of this year).

>>55692814
Or they'll look at review sites, see the 390 getting slightly better performance than the 970, take it home, and enjoy -8fps...

File: gta.png (3MB, 1920x1080px) Image search: [Google]

3MB, 1920x1080px

>>55692478
ok..

>>55692834
Yeah sorry I got bored of waiting to get to that part.

>>55692820
I'm not mad, it's amusing to see you autists fling shit for 7fps.

Again:

i5<i7
old cpu<new cpu

Is that the point of this thread?

>>55692807
Yep! I certainly will. I am waiting for the fucking benchmarks. I hope they come out before my Nitro gets shipped. I don't want to pay for return postage. Actually I may get the 1070 instead if that's the case and just deal with the extra moola it will cost. I am tired of this shit.

>>55692830
It's like you refuse to admit that if you buy a cheap processor, you won't get the fps on the review charts, because they use the best cpu.
Is that hardly surprising? Are you superman and can notice 7fps?

>>55692834
>none of the cores go past 80%
>very different framerates
odd... might be gpu bound.

I wish IN Elite was here.

>>55692878
Please read and reply to this.
>>55692823

As an AMD owner I am trying to get my head why anyone would buy a GPU that is weaker than the other at offloading work AWAY from the CPU than the other at the same price. I know a bit about this async compute and other stuff. But basically it looks to me like AMD is just relying more on the CPU to do it's work which in my eyes is fucking terrible value.

>>55692531
>1080p
>no AA or AO
>cards getting > 140fps
what are you trying to make it cpu bottlenecked?

>>55692823
>>55692919
OK it looks like I was being too simplistic. Just had a quick look at what DX11 does with threading. Hmm. Looks like AMD tied their hands by betting the farm on multi threading. Why can't they just make GPU's the same as Nvidia do??? lol

Oh well. Fuck this I'm going with whichever works best and costs the best.

>>55692968
But AMD drivers only utilise a single thread, that's a massive part of the problem

>>55691997
>AMD is for poors
>When it clearly outperforms Nvidia when paired with a high end processor

:^)

>>55693003
I read it's why they are trying to abandon it because they can't write drivers for shit. Basically they want DX12/Vulkan so they can fire their driver team.

AMD can't into software.

>>55693026
If AMD can't into software and drivers, how the fuck did they build Mantle?

>>55693003
That's false though about the single thread. They just have more overhead. In DX11. DX12 or whatever changes this even more
You might be confused with AMD's processors getting lower single threaded performance.

>>55693043
A bunch of monkeys with typewriters locked in a room.

>>55692531
Cheapest i7-6700k I see is $344
Cheapest FX-6300 - $99
Cheapest 980ti - $370
RX 480 is sold out, but I think $250 or so is fair.

AMD + AMD = $5.63 / avg frame
AMD + Nvidia = $4.62 / avg frame
Intel + AMD = $5.43 / avg frame
Intel + Nvidia = $5.16 / avg frame

Seems like AMD CPU + Nvidia card is best bang for your buck. If you have a 60hz monitor, the extra frames won't help much, but the better min framerate would, and I assume you could bump some setting on the non-amd/amd setups. This also discounts the fact that AM3+ mobos are probably cheaper than LGA 1151 at the moment. Also wonder how an i3 or an i5 would stack up here.

>>55692432
Lol I get normal 100% performance in firestrike with Gtx 980 and stock i5 750.
Doesn't translate exactly like that to games tho.

File: mV9bFz7.gif (3MB, 320x180px) Image search: [Google]

3MB, 320x180px

>>55693108
So then what causes the overhead? I thought it was both AMD drivers and DX11 fighting over core 0

>>55693229
I didnt mean compared to other cards moran, i meant with respect to my own model of gpu

Fucking woosh

>>55692031
Thanks for reminding me that Nvidia is just shills and shitposters and that they aren't worth listening to

File: 800x1000px-LL-d6042000_W3.jpg (103KB, 800x450px) Image search: [Google]

103KB, 800x450px

>>55693467
>>55692665
>>55692052
>>55692031
>mfw dual core i3 6100 is as fast as quad core i5 2500
>mfw these these idiots think Intel hasn't improved CPU performance in the last 4 generations
LaughingWhores.jpg

>nvidiots and amdrones in full force
>not neutral masterrace buying card from a different company each generation to see the best of both worlds
pathethic t b h

>>55693607
>not shitposting as both to trigger the real autists and shills

>>55690717
>improves over time
>have to wait 2 years for your fans to fail and your mobo to burn down to shitpost on /g/ about how your shit card beat Nvidia's unwanted stepchildren they don't even make anymore

AMD experience 10/10

File: 1469086294876.png (17KB, 506x292px) Image search: [Google]

17KB, 506x292px

>>55693640
Sounds like this Kepler Kuck is mad

>>55693640
i have AMD graphics cards from 2008 and 2012 that never failed me

>>55693666
>muh gimping
Nice meme kid, getting a bit stale now.

http://www.babeltechreviews.com/hd-7970-vs-gtx-680-2013-revisited/3/

>>55693685
No gimping, just stagnation

File: Screenshot_2016-07-15-15-48-07_com.android.chrome_1468561701732.jpg (32KB, 720x297px) Image search: [Google]

32KB, 720x297px

>>55693702
>stagnation
Lol, okay buddy.

>>55693685
>getting a bit stale now.

Just like Nvidia's support for Kepler

>>55693716
>driver that "fixed" shitty performance after people complained

If you want to use that as proof of support I feel sorry for you. Even after that update Kekpler cards still beat it on a Gameworks title no less. Just sad. Nvidia pours some sprinkles on your plate of shit and you eat it up?

>>55693329
If this is the case have their been any reviews the show the cpu load breakdown per core when running amd gpus vs nvidia? It would show if it's the drivers leaning on a single core more vs something else.

>>55693739
Of course people had to complain for nvidia to fix it up, how else would they find out?
When's AMDs fix coming out for project cars?
What's that? Never?
Kek

>>55692031
Man if I had the money I'd buy a 5775c. It's a lateral upgrade from a 4670k, but the drop in temp, wattage, gain in igpu performance, and vtd means it'd make a great option for virtualizing windows for gaming while having plenty sufficient performance for linux. Too bad they're so expensive.

>>55693812
If I had the money I'd get an fx4350+Titan x just to fuck with people on /g/ and call them poor even though my cpu is a poorfag tier housefire

>>55693827
>Pentium 4 SLI board
>SLI dual titan x
>water cool and OC the P4
I doubt tiatns would run on shuch and old MB but it would be maximum trolling.

>>55693795
Can't fix what's broken from the ground up on the devs side due to shady practices. Only proves my point what Nvidia's intentions are. You fall into their trap and even buy into it.

>>55693879
>quad xeon 8860 motherboard
>64GB ram
>RAID 0 pcie ssd
>10tb HDD
>gt7600

>>55693893
>can't fix what's broken
Of course, AMD drivers are broken pieces of shit. That's common knowledge by now

>>55693917
Only more telling how Nvidia's cards don't last then.

File: 1451960682235.jpg (110KB, 1024x569px) Image search: [Google]

110KB, 1024x569px

>>55693930

File: 76319.png (23KB, 650x300px) Image search: [Google]

23KB, 650x300px

>>55692031
>posting proof that the driver overhead is a real issue while claiming it isn't and op is full of shit
Nice.

File: 76329.png (24KB, 650x300px) Image search: [Google]

24KB, 650x300px

>>55694011
>i7 6700k
>82 fps

>i7 2600k
>73 fps

Meanwhile, with Nvidia...

>>55694011
If the driver overhead is real why is the drop only in the 2600k? Shouldn't it be gradual?

>>55694199
2600k only supports ddr3 1333mhz without overclocking.

>>55690496
with the way gcn works, it scales more with a stronger processor than a weaker one.

in doom - vulkan, async is enabled on amd cards and async is used HEAVILY in doom - vulkan. the older cpu's cannot feed the ace's and cu's fast enough. you still get a boost, but not as big. slap in a 6700k and, well in doom, it turns that $200 card into a $400 one. its able to keep up with the 480 and feed it plenty.

nvidia on the other hand doesn't have async enabled. id disabled it since it gives nvidia cards a regression and they're waiting for nvidia to release a driver to reenable async for nvidia cards (nvidia lacks hardware schedulers and their cards cannot do graphics + compute at the same time on the hardware level, so they depend on software scheduling and preemption). so the only benefit nvidia is getting is the general less driver overhead. which is why nvidia gets a bigger boost with older cpus and not newer, stronger cpus. the older ones cannot keep up with the driver overhead, so switching over to dx12 frees up a lot of resources for older cpu's while the 6700k is strong enough that it doesn't matter so nvidia see's less of a boost.

thats why you'll notice the stronger the processor becomes, the less of a boost the 1060 receives, and the higher the boost the 480 starts to receive.

gcn is built to be fed, and utilize async. the more you feed it, the more powerful it becomes. give it a ton of things to do and it shine. vulkan / dx12 will always give amd a boost but the stronger the cpu, the more boost you'll get.

if you're building a pc now a simple 6100 is more than enough for a 480. if you're on a first generation i7, it be best to upgrade. regardless if its amd or nvidia.

>>55694782
Got any source for async being disabled on nvidia hardware?
A quick Google Search turned up nothing for me

>>55695124
https://community.bethesda.net/thread/54585?tstart=0
>Does DOOM support asynchronous compute when running on the Vulkan API?

>Asynchronous compute is a feature that provides additional performance gains on top of the baseline id Tech 6 Vulkan feature set.

>Currently asynchronous compute is only supported on AMD GPUs and requires DOOM Vulkan supported drivers to run. We are working with NVIDIA to enable asynchronous compute in Vulkan on NVIDIA GPUs. We hope to have an update soon.

also futuremark stated async is disabled by nvidia for maxwell and below in time spy.

>>55695299
Sweet, thanks. No idea why that wasn't coming up for me...

That's interesting, but if nvidia is going to use more cpu for async, wouldn't it be better to leave it disabled?

>>55690496
Wait, wasn't the point of a low level API to pretty much kill driver overhead?

>>55697337
That and to allow for better concurrency.
AMD's solution takes the 2nd part to the extreme, but you still need a good CPU to see the benefits.

Point is, AMD's cards here are very powerful, but currently we have 2 problems: 1) The rest of the system can't keep up so a lot of the HW remains idle and 2) The software isn't producing enough work to keep all parts of the GPU busy.

Doom on Vulkan is one of few games around that shows what the hardware can do when you fix both problems.

>>55693378
That's what I meant too genius.

>>55695124
Maxwell doesn't do async. Pascal does, in some way that gains them about half the performance gain as AMD does. But it's impossible to say what exactly the driver is doing and Nvidia isn't gonna tell us.

File: time-spy-results-async-vs-no-async.png (37KB, 1209x2006px) Image search: [Google]

37KB, 1209x2006px

>>55694782
>>55695124
I believe he's talking about Maxwell, or Pascal drivers have changed since that was written.