[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vp / vr / w / wg / wsg / wsr / x / y ] [Home]
4chanarchives logo
3DMark DirectX12 benchmark
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.

You are currently reading a thread in /g/ - Technology

Thread replies: 220
Thread images: 39
m-muh async compute!

>Today, Futuremark released its new Time Spy benchmark for the 3DMark suite, which is loaded with DirectX 12 features to put graphics cards to the next-gen test.

>Time Spy isn’t Futuremark’s first DirectX 12 benchmark; that honor goes to the API Overhead Feature Test released in March 2015. But the Overhead Feature Test merely measured draw call performance in DirectX 11 versus DirectX 12 (or the now-defunct Mantle). Time Spy is a full-blown benchmark composed of several subtests that factor into an overall combined score. It’s the DX12 equivalent of Fire Strike, basically.

>And it lets you test with asynchronous compute both enabled and disabled, which leads to some interesting insights. More on those in a bit.

>Better yet: Time Spy’s coming to all versions of 3DMark, including the Basic Edition that’s free for personal, noncommercial use. (You can download it on Futuremark’s site or by clicking the Download demo button on 3DMark’s Steam page.) Time to start your engines, folks—at least if you have Windows 10.

>The paid-for Advanced Edition of 3DMark (which offers additional features and benchmarks) is rising in price from $25 to $30 thanks to Time Spy’s inclusion, however. If you already own 3DMark Advanced Edition, you can purchase Time Spy separately for $10. But if you buy before July 23 you’ll be able to upgrade for $5 or buy the complete Advanced Edition for $10.
>>
Forgot link
http://www.pcworld.com/article/3095301/components-graphics/3dmark-time-spy-tested-we-pit-radeon-vs-geforce-in-this-major-new-dx12-benchmark.html
>>
>>55574850
So can someone tell me who won this gen???
>>
>>55574962
Too early to tell. Vega isn't out yet and there's not enough DX12 titles to judge.
>>
>>55574962
Lel, 480 is so far behind 1070, I think even the 1060 will beat it
>>
>>55574850
hahahahah
fucking REKT
>>
File: 1465827021569.png (661 KB, 717x474) Image search: [Google]
1465827021569.png
661 KB, 717x474
>>55575011
>A $200 card is behind a $500 card

You

Dont

tell


me
>>
>>55575067

you mean a $300 card is behind a $380 card.

480 is being price gouged to shit because of the defective launch model and low yields, 1070/1080 are being price gouged to shit because of excessive demand which will normalize in a few weeks.
>>
>>55575067
Yeah I mean AMD is just budget shit, only reason to buy one these days is if you literally don't have enough money for any Nvidia
>>
>RX 480 beats OC'd 970 and 980
working as intended if you ask me
>>
((futuremark))

is futuremark still relevant nowadays? I remember the old one 3dmark03 05 06 and vantage, I like to show my scoring on some forum lel.
>>
>>55574850
>RX 480 beating GTX 980
>async doesn't matter guys!

also
>synthetic benchmarks from companies that take money from Nvidia
>>
>>55575096
lol@AMDrone
>>
>>55575097
got any proof of that?
post it here or your mom will die :)
>>
>>55575089
And whats wrong with that.

980 performance for $200 seems good
>>
>>55575097
>companies that take money from Nvidia
Lel, the usual cry of the fanboy. You got any proof to back that up or is it just denial and wishful thinking?
>>
>>55575125
two years ago maybe
>>
File: file.png (38 KB, 1569x223) Image search: [Google]
file.png
38 KB, 1569x223
oh god lord my sides

>nvidia supports async through the drivers, rather than on the hardware itself
>>
File: 1467212482874.gif (2 MB, 167x200) Image search: [Google]
1467212482874.gif
2 MB, 167x200
>>55575088
https://www.amazon.com/s/ref=nb_sb_noss_2?url=search-alias%3Daps&field-keywords=geforce+gtx+1070&sprefix=Geforce+%2Caps%2C278

What else?

Maybe the 1080 is a $100 card somewhere in the internet?
>>
>>55574850
>If you already own 3DMark Advanced Edition, you can purchase Time Spy separately for $10.
The fuck kind of jewing is this.
>>
synthetic benchmarks are worthless
>>
>>55575136
What games even require something more than a 970.


>inb4 4k and 144hz meme
>>
>>55575159
Its not "synthetic" shit head its running a game engine.
>>
>>55575164
are you bling or just poor?
>still on 1080p AND 60Hz
pathetic loser
>>
>>55575088
>$300. Where the fuck do you people keep getting this number from?I can find it here in the EU for less than that which means that you can certainly find it in the US too.And prices are rising cause its sold out almost everywhere,which is impressive considering how much stock AMD had.
>>
>>55575164
5K 29:10 Ultrawide @ 144Hz
>>
>>55575157
3dmark isn't a charity and their benchmarks aren't made for free

if you don't like it why don't you go use the FOSS dx12 benchmark?

right after you get done fellating RMS that is
>>
We're excited to announce that 3DMark Time Spy - our new DirectX 12 benchmark test - has been released on Steam. Time Spy is available now as a free update for all Windows editions of 3DMark, including the Steam demo.


Developed with input from AMD, Intel, Microsoft, NVIDIA, and the other members of our Benchmark Development Program, 3DMark Time Spy is one of the first DirectX 12 apps to be built "the right way" from the ground up to fully realize the performance gains that the new API offers.


DirectX 12, introduced with Windows 10, is a low-level graphics API that reduces processor overhead. With less overhead and better utilization of modern GPU hardware, a DirectX 12 game engine can draw more objects, textures and effects to the screen.


With its pure DirectX 12 engine, which supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading, 3DMark Time Spy is the ideal benchmark for testing the DirectX 12 performance of the latest graphics cards.
>>
>>55575106
found the inelite and chinklong.

>>55575127
it is true 3d mark vantage is made for ati/amd back then( I remember sapphire sponsorship were there lol), after rx480 showing up and hd 7000 series beating nv 700 series suddenly some ((futuremark)) making amd look bad well pure ((coincidence))
>>
File: 1467991146152.gif (3 MB, 200x150) Image search: [Google]
1467991146152.gif
3 MB, 200x150
>>55575181
>>55575203
Get a life, hit the weights, take a shower.

Get a clue
>>
>>55575203
There isnt even a 4k 144hz monitor on the market you cuck
>>
>>55575208
>benchmarks aren't made for free
The whole point is if you bought it first. It's not a separate product, they're including it in the original suite. However if you already bought the suite they're saying "fuck you, you can't have it".

3DM is a benchmark. You'll use it once, maybe twice a year unless you are running some benchmark review site. Almost nobody buys it and just uses the free version.

Charging for this to people who already bought the suite is really fucking scummy.
>>
>>55575224
I bench 140 kg and do cycling
shaved, perfect haircut, gf, car, apartment and job I enjoy
nice projection though, amdpoor
>>
>>55575201

the RX 480 is more than $500 currently because AMD hasn't shipped more since launch day.

https://www.amazon.com/MSI-Radeon-RX-480-8G/dp/B01GX5Z4EM

Meanwhile the price gouging on the 1070 is inching closer to MSRP, there are even a few on amazon for less than goyim edition pricing: https://www.amazon.com/MSI-GeForce-GTX-1070-AERO/dp/B01GUAJMRU
>>
>>55574850
>If you already own 3DMark Advanced Edition, you can purchase Time Spy separately for $10.
>>55575212
>Time Spy is available now as a free update for all Windows editions of 3DMark, including the Steam demo.
Which is it? Is it a free update for a $10 add-on?
>>
>>55575276
Prove all with timestampos
>>
File: pilki.jpg (20 KB, 300x399) Image search: [Google]
pilki.jpg
20 KB, 300x399
>>55575142
>Vulkan, a graphics API, enabled
>on a benchmark which explicitly uses DirectX 12, a graphics API
>>
>>55575304
This is from the Bethesda forums, for DOOM
>>
>>55575011
it's faster than 980 which is where nvidia is positioning 1060.
>>
>>55574850
Artificial benchmarks are irrelevant. Show me more test cases like DOOM.
>>
File: Timespy.png (235 KB, 1187x1068) Image search: [Google]
Timespy.png
235 KB, 1187x1068
wew lad.
>>
File: 1468037710976.jpg (24 KB, 400x400) Image search: [Google]
1468037710976.jpg
24 KB, 400x400
I have a 750ti.

Should i upgrade to 1070 or wait for the new cards?


With this DX12 shitstorm all around i dont want my new card to be obsolete in 1 year
>>
File: 1417056231352.jpg (19 KB, 280x400) Image search: [Google]
1417056231352.jpg
19 KB, 280x400
>>55575276
Uh huh
>>
>>55575088
>announced as $200 ~ $250 card
>its being price gouged
>can't look anywhere but ebay and amazon for non-retailers
also
>$380
so the rx480 is price gouged to $300, but nvidias "msrp" stands correctly without being able to find retailers selling at that price?
you fucking shills are so easy to spot. its always one over priced rx480 you found on ebay or amazon by some unverified seller, vs the msrp of a card that hasn't dropped below $50 above msrp as of yet.
>>
>amdboys call Futuremark, a neutral third party, shills for Nvidia
>while they themselves keep quoting fucking Ashes of the Singularity
TOP KEX
>>
>>55575404

are you retarded? the 480 is being sold for $500 (which is more than price gouged 1070s), not $300.
>>
>>55575284
Who the actual fuck buys tech from amazon?Just buy it from newegg or microcenter if you live in the states.And oh yea the 1070 is finally reaching its advertised price a decade after its release,still dosnt change the original point that the 480 is a 200-240 dollar card while1070 is a full $140 (at mrsp) more expensive.The 480 isnt meant to compete with it,that would be like the 380 competing with a 980
>>
>>55574850
>I care about Futuremark synthetic benchmark crap
>>
>>55575419
>found it on amazon from someone who bought a bunch and are selling them over priced
goodjob, compare scalper prices to retail.
$200~$250 is still the price, just because you can't find it doesn't mean its untrue, I literally saw them at a local Fry's yesterday for $240.
>>
>>55575427
>Just buy it from newegg or microcenter if you live in the states.

480 is perpetually out of stock on newegg and microcenter is only in like one city, i wouldn't buy this AMD trash card anyway.

>And oh yea the 1070 is finally reaching its advertised price a decade after its release,

you mean 3 weeks after launch, which is pretty short when you compare it to AMD's paper launches (like the fury x being $1000 for all of Q3 '15 despite being worse than the 980ti across the board).

>,still dosnt change the original point that the 480 is a 200-240 dollar card

the 480 is a $350-$500 card with the performance of a $200 card.
>>
>>55575404
Havent you fucking heard, the rx480 costs $600 and nobody buys a card for that money, even the 1070 is cheaper oh my god.
Now pack your stuff and buy the 1070 or 1080 and support nvidia. AMD should be bankrupt and dead so they would stop shitting all over the drivers with their async much feature.
>>
>>55575487
>the 480 is a $350-$500 card with the performance of a $200 card.

Good goy

https://www.amazon.com/s/ref=nb_sb_noss_2?url=search-alias%3Daps&field-keywords=980
>>
>>55575276
Let's see a pic of that chest then, Rocky
Don't forget the timestamp
>>
>>55575501

?

the 480 is slower than a 970.
>>
In germany its in stock and costs the normal price
>>
>>55575521
XD
>>
File: 924764714.jpg (32 KB, 512x512) Image search: [Google]
924764714.jpg
32 KB, 512x512
>>55575303
>>55575394
>buttblasted amdkeks
>>
>>55575546
>Damage control
>>
>>55575546
>I pay a fuckton more for 50% faster shit
>THAT MEANS I AM SUPERIOR

Gooooood gooooy
>>
>>55575212
>Developed with input from AMD, Intel, Microsoft, NVIDIA, and the other members of our Benchmark Development Program

>>55575221
Where is your god now fanboy?
>>
File: mark2.jpg (130 KB, 1203x1229) Image search: [Google]
mark2.jpg
130 KB, 1203x1229
>>55575360
Is that OC even stable, lad?
>>
>>55575256
So, how many years of free full-featured DLC with your purchase would you be satisfied with?

I mean, jeez, you can still use the basic version of the new test for free.
>>
>55575011

So the 480 really does beat the 980?
I'm for sure buying this card now.
>>
>>55575287
There's a free version, and an advanced version with more settings.
>>
>>55575487
Are you brain dead?I guess we should have compared the 1070 to the 1080 cause it was being sold at the 1080 mrsp.The Rx 480 is a 200-240 dollar card,to be fair I dont know of other major tech retailers in the states but here in europe you can still find the 8gb for €250-260 whic isnt a huge diference from launch prices.
>>
File: Time Spy.png (225 KB, 1266x266) Image search: [Google]
Time Spy.png
225 KB, 1266x266
1600mhz core/8000mhz vram
>>
>>55575587
Yeah, it wants to keep voltage at 1.05 so it drops to 2100 or 2088 after a minute or so depending on the program.
>>
File: wat.jpg (17 KB, 246x227) Image search: [Google]
wat.jpg
17 KB, 246x227
Wait, you can use DX12 benchmark in Windows 7?
>>
>>55575689
>system requirements
>Win 10 64bit
>>
>>55575698

Fugg
>>
>>55575689
No, blame Microsoft. This is like what Halo was for Vista
>>
>>55574850
So for 250$ I get 80% of the performance of a card that's 200$ more expensive?
>>
>$400 > $800
they actually did it, the absolute madmen
>>
>>55575756
That feel when Halo 3 never came on PC and I still don't know how the story ends.
>>
>>55575777
Too bad you can't play a benchmark
>>
>>55575791

What are you? Casual?
>>
>>55575777
>$800
>>
>>55575791
too bad you can't play DOOM on your geforce gimpworks card
>>
>>55575781
Next year for Halo 3 anniversary as a Play Anywhere title
>>
>>55575807
wut
>>
File: amd dx12.png (2 MB, 1450x3400) Image search: [Google]
amd dx12.png
2 MB, 1450x3400
>mfw dx12
>>
>>55575777
Now show me the microstutter graph.
>>
>>55575676
How the hell did you get it to go that high tho? I get artifacts up the ass and driver crash when I try your numbers.
>>
>>55575835
stutter, you say?

http://www.overclock.net/t/1605618/nvidia-update-gtx-1080-1070-high-dpc-latency-and-stuttering
>>
>>55575672
Still worse than a stock 390
>>
>>55575835
I think Polaris was supposed to really improve CF, and considering how well the 480 scales that appears to be true, so maybe microstutter isn't as horrendous as it used to.
>>
>>55575864
Where's one to compare?
>>
>>55575888
polaris wasn't intended for traditional CF, they want to push multiadapter in DX12/Vulkan

>>55575835
you're right, FPS only means so much when you exclude stutter
https://www.youtube.com/watch?v=uxnFv8cnjZA
>>
>>55574850
>tries to shill on amd
>forgets to put the correct async enabled chart
>gets smoked
love how on every single charts they do have the overall score and on async enable its only the "graphics" score..
guess physics passing through compute pipeline regressing nvidia isnt something to show
kek
>>
>>55575777
3dmark has always had extremely good multi-gpu scaling

actual games are never anywhere close
>>
>>55575860
Its not like you can do anything to these cards, they're voltage locked. MSI gaming 1070 using evga precision x 16 pascal version, max power target 126%, 92 c temp target set to priority. Fan curve set around 70% at 55c. clock offset +245, mem offset +600. Rest is lottery.
>>
>>55575936
DX12 multi-adapter has allowed up to 95% scaling so far. Modern APIs actually allow for proper parallel computing as opposed to legacy APIs like DX11.
>>
File: 1392959249130.png (163 KB, 554x439) Image search: [Google]
1392959249130.png
163 KB, 554x439
Why the fuck are the AMD fans whining in this thread over these benchmarks? It doesn't show anything unexpected. AMD cards benefit heavily from async compute, 9xx Nvidia cards don't benefit at all. 1070 trashes everything AMD has to over with and without. Everything is as it should be.

I pity people who bought a 980 though.
>>
>>55575984
>1070 trashes everything AMD has to over with and without
i would fucking hope a $450 brand new card would be beating old cards and a $200 card.
>>
>>55576008
Considering old R9 200 cards thrashed Maxwell, this is indeed a surprise
>>
>>55576023
a 370 thrashes a 780 ti now, the gimping is real.
>>
>>55575984
so now synthetics are back on the table?
2 weeks ago they were out
kek
>>
>>55576032
It's not gimping, it's a heavily AMD optimized title leveraging the hardware

But I will say that Kepler is aging like milk, and it's pretty sad. But actively being gimped? No, there's no evidence for that
>>
>>55576062
DOOM is now a synthetic benchmark? Total War is now a synthetic benchmark?

Time to stop posting.
>>
File: doom_1920_o_vs_v.png (837 KB, 1096x629) Image search: [Google]
doom_1920_o_vs_v.png
837 KB, 1096x629
>>55576032
Only with Vulkan on, thankfully the game still plays well with OpenGL.

But yeah, Kepler cannot into next gen APIs.
>>
>>55576122
>OpenGL
Give it a rest man it isn't 1992 anymore.
>>
>>55576133
I'm just a poor gimped 780 Ti user trying to get by until Vega comes out.
>>
>Sponsored by Galax
>>
>>55576085
i highly doubt being 12% to 10% ahead from 480 "thrasing"
>>
File: 1353276243149.gif (483 KB, 141x141) Image search: [Google]
1353276243149.gif
483 KB, 141x141
>>55576145
>until Vega comes out.
I'm with you in that boat senpai.
>>
>>55576122
>>55576145
With how much the 780ti tanks in Vulkan, could it be a driver issue? Could they patch it? I don't think Nvidia has released an actual Vulkan driver yet have they?
>>
>>55575145
holy shit that gif is terrifying
>>
File: mlg tyrone.jpg (252 KB, 996x1501) Image search: [Google]
mlg tyrone.jpg
252 KB, 996x1501
>>55575756
And this is why DX12 will be the new DX10. See you in DX13.
>>
>Partnering a fucking RX 480 with a shit processor

http://www.3dmark.com/spy/6892

The only RX 480 I could find.

Meanwhile the highest GTX 970 score I could find.

http://www.3dmark.com/spy/8419
>>
>>55576201
Supposedly it was added back in 356.39 for all the currently supported cards. If it's failing this hard on Kepler its probably because the hardware is shit at it and not just a driver issue.

With as well known as this problem is if Nvidia planned on saying something or had a fix they would probably have announced it by now.
>>
>>55576201
they released a game ready today
but the fun part is
>no games was listed
>vr fixes
>no word about async
>>
>>55575089
This.

Nvidia is so much higher quality than AMD
>>
>>55575125
>980 performance

Top kek, a 480 can't even beat a 970
>>
>>55576333
you must be new.
>>
>>55576333
Just look at the dx12 benchmarks anon
>>
>>55576318
indeed you need special quality to be this low on the barrel bottom lol
https://forums.geforce.com/default/board/172/geforce-1000-series/
>>
>>55576230

FX is a shit CPU, more news at eleven.
>>
>>55575011
>Lel, 480 is so far behind 1070,

It has 70% of its performance, while costing 50-60% of its price. That means the 480 is ahead in price/performance.
>>
>>55575088
>you mean a $300 card is behind a $380 card.

If you go by price gouging then it is a $300 card versus a $500 one.
>>
>>55574850
>Free for freeware users
>Raises the price of non-free version
>People who already paid money need to pay more

That's so nice of you, 3DMark
>>
File: neo g.jpg (19 KB, 745x478) Image search: [Google]
neo g.jpg
19 KB, 745x478
>>55574850
Fucking delet dis NAO
>>
File: 1465233349307.jpg (4 MB, 10000x10000) Image search: [Google]
1465233349307.jpg
4 MB, 10000x10000
>>55574850
>480 dominating 980
>based 390X giving its customers all kinds of extra love
(not shown)
>Fury X absolutely fucking up 1070, and likely 1080, as well.
Who did you think you were shilling for, again?
>>
>>55574850
where 980ti
>>
>>55574962
Playstation Neo, baby!
>>
>>55576410
60% performance while overclocked.
There's a price premium on higher performance, news at 11.
>>
>>55576481
(not him, but...) Then why are you even comparing the two?
>>
File: timespy-3.png (76 KB, 690x996) Image search: [Google]
timespy-3.png
76 KB, 690x996
>>55576467
Fury X is still behind a stock 1070 kiddo even in its best case element. I considered getting one but seeing it 40% behind in some games is putrid.
>>
File: tmp_18786L235T4K1455635660.png (55 KB, 996x624) Image search: [Google]
tmp_18786L235T4K1455635660.png
55 KB, 996x624
>>55576547
Yfw dun goofed
>>
Just did a test on a R9 290 OC'd to 1075/1500 on an i7 4770 @ 4.2Ghz and 16GB DDR3 using 16.7.2 drivers in Windows 10.

I get around 4000 give or take.
>>
>>55576580
I can link any number of vids comparing vulkan doom and those numbers aren't even close to what its actually doing. Looks like shills to me.
>>
>>55576673
Looks like you haven't read about the doom update a few days ago.
>>
Excel charts: The Thread.
>>
Amd is garbage, the only reason they "Age like wine" is because they have shit drivers for the first year.
>>
>>55576723
It's been no secret that AMD drivers have been holding them back. Only recently have they been stepping up their optimization game have they started to show their true power.

AMD was always going to benefit way more than Nvidia with the new API's, partly due to the better implementation of async along with less driver overhead
>>
File: my graph is better than yours.jpg (95 KB, 653x587) Image search: [Google]
my graph is better than yours.jpg
95 KB, 653x587
>>55576461
>>
AdoredTV AMD shill Doom analysis video is up.

http://www.youtube.com/watch?v=lXijkS_qIzA
>>
>>55576673
>>55576696
Here is Digital Foundry's
https://www.youtube.com/watch?v=ZCHmV3c7H1Q
>Nvidia Shills Starting To Sweat Hard
>>
>>55576911
>AMD shill

just because you have a preference doesn't make you a 'shill' ya dumb cunt
>>
Amd poorfags need to give it up, When you buy Nvidia it just feels like a premium Item, Amd feels like a cheap knockoff
>>
>>55576949
It's because AMD cards use the cheapest components possible. They're really only built to survive benchmarks by their shills, and generally fail within 6-12 months. Nvidia cards cost more, but pretty much work forever.
>>
>>55576911
>Dead wrong for months about Polaris performance.
>Finally one game shows numbers that aren't embarrassing for a new 2304cu card on 14nm.
>Declares victory
>Begrudgingly acknowledges Polaris is still shit in caveat
>>
>>55576949
>feels like a premium
yeah, only feels like. not showing anything.
>>
File: Vulkan.png (83 KB, 1095x280) Image search: [Google]
Vulkan.png
83 KB, 1095x280
>>55576928
>Nvidia Shills Starting To Sweat Hard
Fury X is still 30 fps behind 1080, pulls even with 980 ti and the lower tier 1070. Scores are on stock speeds.
I'm sweating because AMD is actually doing what it should be instead of the usual 25% deficit.
BTW I'd have bought a 490 if they'd actually released one and it wasn't shit.
>>
>>55577019
Nvidia cards are better than AMD ones in every way, that's the real truth. You do pay more for quality.
>>
>>55576949
>>55577000
Screencapped, thanks.

You're both(?) really bad at your jobs
>>
>>55577190
You have an iPhone, don't you?
>>
>>55577000
>[AMD] really only built to survive benchmarks by their shills

That's funny because it wasn't AMD that implemented "GPU ""Boost""" which causes GPUs to thermal throttle after 10 minutes of use.

NVIDIA cards are designed to perform well in 1-5 minute benchmarks but collapse under any gaming session longer than 10 minutes when reaching their target temperature.

It's really pathetic how desperate they look when you understand the shady shit they pull all the time.
>>
File: 8MmvGDr.jpg (119 KB, 648x674) Image search: [Google]
8MmvGDr.jpg
119 KB, 648x674
>>55577227

Have some evidence to back that up.
>>
>>55577227
Bro your arguments aren't going to make your poor fag card perform better
>>
>>55577227
Poo in the loo, Pajeet. It's well known that AMD cards are unreliable. Nobody will touch a used AMD card because they know it's most likely months away from going up in smoke. That's why big partner companies like EVGA, Palit and Zotac refuse to work with them, as they'd be forced to implement AMD's planned obselescence measures.
>>
>>55576929
I watch all his vids. He is a shill. But I am an AMD fanboy so I can call him that without prejudice.
>>
>>55577107
Yeah I meant as in they were saying it loses to a Fury X even with Vulkan.
The Fury X is $420 (Newegg PowerColor)
and the GTX 1070 is $420 (Amazon MSI AERO)

aka pleb tier

So if you don't care about thermals / tdp, the FuryX is right at the 1070.

I'm probably sounding like a AMDRONE now.

In reality I'm just excited there's going to be some actual competition between the two now.
>>
File: aKCAW5x.jpg (105 KB, 796x805) Image search: [Google]
aKCAW5x.jpg
105 KB, 796x805
>>55577286
DELET
>>
>>55577324
>if you don't care about thermals.
Fury X uses an air, bro. Thermals are a non issue
>>
>>55577397
*aio

Ironic
>>
File: 1468521406053.jpg (120 KB, 548x629) Image search: [Google]
1468521406053.jpg
120 KB, 548x629
>960 significantly faster than the 780 Ti

How peculiar.
>>
>>55577397
>>55577413
Fury (non-X) uses air
>>
>>55577418
Seems like a driver issue, we'll see if it gets fixed
>>
>>55577426
Yeah, but he said fury x
>>
>>55577445
I was implying he might have been confused
>>
>>55577461
Meh it has plenty of other issues. Knowing what I do now I'd buy a 1070 just because the fans on the msi at least are dead silent up to about 80%.
>>
>>55577445
>>55577461
yeah sorry i wasnt thinking right about the thermals,tdp is still way higher
>>
>>55576834
lewd
>>
>>55577418
Are you telling me I should not get the MSI GTX970 4 GB?
>>
>>55577442
Doubtful that it's a driver issue, given that the original Titan, which is the exact same chip as the 780 and 780 Ti, is so much higher. Seems more like the 3GB Keplers shitting themselves when they run out of VRAM, whereas the 3GB AMD cards are managing better. Nvidia won't be doing any driver work on Kepler either way. They've given it one update in almost two years, and only then because people were screaming about "gimping" when they performed like absolute shit in Project Cars and The Witcher 3.
>>
>>55577509
Same on my fury X.
Plus, it's downright goddamn sexy
>>
>>55577418
You think that's bad?
A fucking 370 (7850) is faster than the 780ti. There has to be something wrong.
>>
File: 1437102295502.png (41 KB, 802x799) Image search: [Google]
1437102295502.png
41 KB, 802x799
>>55575011
>>55575933
>>55576122
DELETE THIS
>>
File: 2814802.png (395 KB, 923x684) Image search: [Google]
2814802.png
395 KB, 923x684
>>55577942
>>
>>55577286
If you buy the reference cooler you should literally suit yourself. I get steady 2ghz on firestrike stress test with the g1.
>>
>>55577942
The 780 ti lose around 50% of it's performance on Vulkan, ofc something is wrong.
>>
>>55577107
sorry bro but fury x is 25% faster than the 1070 not 0. https://www.techpowerup.com/224095/doom-with-vulkan-renderer-significantly-faster-on-amd-gpus
also I noticed you cannot calculate
now go and cry like a little bitch fanboi
>>
>>55578071
The question is though if the problem is with the Kepler architecture or if its a driver issue?

And further, if software does Nvidia even give a shit about fixing it?
>>
>>55578112
Nvidia aren't fixing anything to do with Kepler and anybody who says otherwise is delusional. They haven't done shit for Kepler apart from the Witcher 3 fix since Maxwell launched in 2014.

t. GTX 780 owner until two weeks ago
>>
>>55578136

>t. GTX 780 owner until two weeks ago

So which 1070 did you buy?
>>
>>55576376
Haha haha haha hah amirite? A tech support forum has tech support questions! What a fail company. Epic fail amirite? What a fail! Lol fail nvidia more like nvfaildia lololol. Every single card has to be perfect with no issues ever.


Wasn't there something about amd cards blowing up mobos? Like a lot of them? Because amd lied their asses off about their card's power draw?
>>
>>55578245
>Like a lot of them?

You tell me.

https://www.youtube.com/watch?v=KHhiW4J6Hqo
>>
>>55578245
show me that blown up mobos.
>>
>2016
>dx12
Obsolete. Vulkan is the future. Get with it, /g/. Even NuDoom hopped on that shit quicker than a 40 year old virgin on a cheap whore.
>>
>>55578107
>Just relinking the same euroshit SS with some randos blurb over it
Thanks for no info kid
>>
File: timespy-3.png (270 KB, 602x869) Image search: [Google]
timespy-3.png
270 KB, 602x869
>1080 has almost the same gains as the 480
Does it means that the 1060 will have async compute?

I think i'm shelving my idea of buying a 480 desu
>>
Nobody has yet to answer this question.

Why was both Nvidia and AMD running OpenGL 4.5 in the Doom beta and then AMD's gets deprecated to 4.3 on the games release and is STILL 4.3 even with the Vulkan update?

I don't get the whole 'it's AMD's shitty driver' thing. If it was working fine in the beta why change it?
>>
>>55578665
>If it was working fine in the beta why change it?
Because it was working entirely too fine and needed "fixed".
>>
Async Compute is not heavily used on this one. TO be able to make those AMD ACE, it needs to have heavy batches of load to really shine
>>
>>55575374
Depends on how badly you need to upgrade right now. If it's working fine and you can stand another 6 months then I'd wait for vega/1080ti.
Unless you can't afford 1080ti or don't like AMD, then buy a 1070 now.
>>
File: nvidia.jpg (11 KB, 259x194) Image search: [Google]
nvidia.jpg
11 KB, 259x194
>>55576580
>>
>>55577222
He also thinks garden of eden was in Louisiana and that earth is 6000 years old. He supports trump.
>>
wew, the green team is getting his ass blasted yet again, better hire more and better shills
>:^)
>>
>>55577324
only for murrica. At the rest of the world fury x is much more expensive than the 1070 at this time.
>>
File: ss (2016-07-14 at 10.22.05).jpg (114 KB, 1499x316) Image search: [Google]
ss (2016-07-14 at 10.22.05).jpg
114 KB, 1499x316
i thought this was supposed to run good on AMD?
>>
>>55581288
Cpu is bringing it down a little
>>
>>55581288
Runs better than 980.
>>55578602

If you're talking about relative performance gain being lower, that's probably due to due tessellation/preemption/low async offload.
>>
>>55581288
what res?
>>
>>55581334
He got a score, so he ran it in default settings which is rendered at 1440p
>>
>>55579856
they're actually pitting the $240 480 vs the $430 1070 against each other, they're desperate lmao.
>>
File: storere.png (113 KB, 1160x509) Image search: [Google]
storere.png
113 KB, 1160x509
>>55575284
>AMD hasn't shipped more since launch day.

Yet the Micro center by me has them every 4 days or so.
I mean why the fuck do you guys make shit up when there are people on line who knows if you're lying or not?

Link:
http://www.microcenter.com/product/466374/Radeon_RX_480_8GB_GDDR5_PCIe_Graphics_Card
>>
>>55581319
yeah, I was looking at graphics score though

>>55581320
>If you're talking about relative performance gain being lower, that's probably due to due tessellation/preemption/low async offload.
I just ran the test and got a result. if there a DX11/DX12 option? I was hoping to beat/close to 970s and 480s, but the one posted here beat me by 100 pts

>>55581334
no idea lol
>>
>>55575164
>not being poor is a meme
>>
>>55575164
Not all games even get 144 FPS at 1080p, yet alone at 4K
>>
>>55581960
Even if a game gets something like 90-100fps on a 144hz monitor it will still be way smoother than standard 60fps
>>
File: TimeSpy.png (179 KB, 1193x234) Image search: [Google]
TimeSpy.png
179 KB, 1193x234
>>55574850
Neat, I like the references to other benchmarks.
>>
>>55582698
And games
>>
>>55574850
>DX12
>not vulkan

end your life
>>
>>55576077
It's getting worse performance as time goes on. Not just worse relative to everyone else, worse compared to the past. Games that used to play just fine now sometimes dip below 60.
>>
>>55578391
Seriously is this board retarded? I want to believe that these are all just summerfags but there's less than 10 posts mentioning vulkan.
>>
You know what annoys me about this benchmark?
>It doesn't recognize 16.7.2 drivers as valid
>It runs @ 1440p which is not really relevant to 1080 gamers which are still the majority.
>>
>>55582744
https://www.youtube.com/watch?v=DvCsT_dk-1E
>>
>>55574850
>Galax sponsored
>therefore nvidia sponsored

It's clear that its made to make amd cards look bad, it's not even using much async compute from the description
>>
>>55576410

price/performance is overrated. cheaper cards always have an advantage. there's tons of cards that people will give away for free that allow you to play games. even if that free card got 1 fps, that would give it an infinite price/performance ratio.
>>
>>55576410
Yes anon, it always has been like this and will always be like this.
>>
>>55577442
kek
waiting nvidia to give async via drivers
waiting since august 2015
"we will update via drivers soon™"
>>
>>55578665
opengl 4.5 is almost at mantle levels...
https://www.opengl.org/registry/doc/glspec45.core.pdf
it is literally build around gcn 4.0 look at the specs and then look at how nvidia does their pipeline and then amd..
>>
File: AchievementGet.png (12 KB, 940x74) Image search: [Google]
AchievementGet.png
12 KB, 940x74
Okay, now I can finally go to sleep.
>>
>>55575088
Alert extreme shill - MSRP nvidia claims

Alert
>>
>>55581320
so we need to dump tesselation which is the strength of nvidia and favor gazillion loads of async which favors amd? very unbiased yes
>>
I don't care about the amd vs nvidia fanboyism going on.

So, lemme get this right, if I bought 3DMark for the $25 or whatever it was (or $5 on sale), then now I'm missing content.
but, right now, if I go and download the free version, then I will have the TimeSpy, which I don't have.
Or, I can spend $5 to upgrade, or $10 if I don't have it, to get the TimeSpy complete with options or something.

this fucking thing should have been a free upgrade to people who already bought it the damn thing. not this confusing release that they have now.
>>
>>55574850
Why do they even make benchmarks? None of that shit matters in games.

If I wanted to stress my gpu, I'd just use heaven.
>>
>>55577311

>It's well known that AMD cards are unreliable

[citation needed]

>That's why big partner companies like EVGA, Palit and Zotac refuse to work with them

I take it you don't know that zotac and sapphire are sister companies. Of course, you might know and you are just being a shill.
>>
>>55578665
>If it was working fine in the beta why change it?
Because Nvidia paid them loadsamoney to do so.
>>
>>55584828
You can just download the free TimeSpy version for paid 3DMark. You just won't be able to run custom benchmarks or disable the demo, etc.
Thread replies: 220
Thread images: 39

banner
banner
[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vp / vr / w / wg / wsg / wsr / x / y] [Home]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
If a post contains personal/copyrighted/illegal content you can contact me at [email protected] with that post and thread number and it will be removed as soon as possible.
DMCA Content Takedown via dmca.com
All images are hosted on imgur.com, send takedown notices to them.
This is a 4chan archive - all of the content originated from them. If you need IP information for a Poster - you need to contact them. This website shows only archived content.