![SAPPHIRE-Radeon-RX-480-NITRO-2[1].jpg](https://i.imgur.com/vIAU7bF.jpg "SAPPHIRE-Radeon-RX-480-NITRO-2[1]")
>According to our information, some european stores already received their stock, and they are ready to sell Polaris 10-based graphcis cards. In China a company called Dataland (Chinese PowerColor brand), was already selling places on reservation lists.
Meanwhile some reviewers already have their cards for tests, so we expect to see a lot of leaks in the coming days.
http://videocardz.com/60992/sapphire-radeon-rx-480-nitro-pictured
Do you like? I mostly only care about how well it will perform but kinda like the look.
![SAPPHIRE-Radeon-RX-480-NITRO-1[1].jpg](https://i.imgur.com/bzYzd0m.jpg "SAPPHIRE-Radeon-RX-480-NITRO-1[1]")
good thing I checked the catalog. Was about to post this.
And let's just avoid discussing the source of the leaks.
>>55014231
I posted yesterday, I work at AMD. Test cards are being sent out. No one can officially release benches until the 29th but of course they will leak it.
>>55014836
I remember you. I actually believe you. Thanks for heads up mate.
>Make AYYMPOORFAGS card
>No DVI which is what the AYYMPOORFAGS use, shitty Korean cheapo monitors with DVI only
>All AYYMDPOORFAGS have buy a new DP monitor also, making it cost more than Nvidia cards
AYY
Y
Y
>>55014836
Can you tell me about OC capabilities of RX 480?
inb4
>I hope this single mid-end card beats the 3dmark Firestrike score of a 4 way sli 980Ti's for a price of 29,99 or I'm done with AMD.
>>55014231
Looks like a damn cheese grinder
>>55014791
it's reddit.
>>55014924
you don't strain your rice?
>>55014882
there will be an hdmi port, so what are you shilling about?
Why did they make the NITRO cooler look so shit? The old design was perfect.
LOOK AT THIS FUCKING SHIT.
THIS CARD IS HEADING FOR DISASTER.
LEAKED ALREADY.
BENCHMARKS AND TESTS WILL PROVE IT'S SHIT.
AMD FAGGOTS STILL POOR AND SHILLING
>>55014882
>>55015313
Why nvidiots always talk in CAPS?
>>55015313
LOOK AT THIS [spoiler]7[/spoiler]
>>55015313
you have a sad life.
>>55015313
Try harder nvidiot shill.
>>55015378
AMD goose's cooked. Just look at that curry dindu holding the poolaris during that stupid conference like it was the holy grail, the great saviour of company that's been failing miserably for the past 5 years.
200$ dorra for a rebranded 280/290. 1060 will wipe the floor with this cockamamie desperation mode 480.
>>55015572
4/10 made me reply
>rebranded
Its gcn 4.0 you stupid idiot.
Get three and a halfed
>>55015572
Dude, I own an nVidia GPU, but you're just fucking pathetic. You're letting a company act as your identifying feature.
Go masturbate or something, god damn.
>>55015662
What matter are the results. 390 came out, seemingly better than the 970, but it was late, the 480 i going to make it obsolete while not really offering any significant power gains.Nvidia shills the last shekel out of their loyal goyums but they always make sure their cards offer a hefty amount of power and win benchmarks -- that's what matter, more fps.
Who gives a fuck about tdp? Electricity is cheap.
>>55015728
The 390 against the 970 holds merit.
Nonetheless, the 480 against the 1070 has no place in this discussion. It's a 200 (~) card vs a 400 (+) one.
>>55014231
am i the only one who likes the cooler design? shame for the blue stripe though.
>>55014231
>http://videocardz.com/60992/sapphire-radeon-rx-480-nitro-pictured
>4GB models are also expected, but availability of such cards will depend on AIBs.
>4GB
>4GB
>4GB
>>55014791
Wait what? It will cost 300€?
>>55016013
Still cheaper than £400.
>>55016056
Much like the 1080 is selling for $800 the 480 will be more initially. New = pay the premium. I never expected it to be $199 to euro/gb. VAT plus other shit will pump the price upwards. US has different taxing schemes per state so you can't really compare it anyhow.
>>55016101
True but what the fuck. 200$ would be 150€ plus taxes 180€ so 230€ for the 8gb...
>>55014231
Looks like shit desu senpai, but I only care about how it performs and the price.
>>55016116
I hoped it would be maximum 250€
>>55015916
Im not sure yet. I think it looks unique. Not ugly but not beautiful. At least somewhat pleasant to look at.
>>55016101
>Still cheaper than £400.
>Implying the 480 has comparable performance to a 1070.
>>55016255
About 20% less performance for half price.
>>55015947
To be honest im not sure if i even need 8Gb ram at 1080p.
I'm looking to get the best upcoming GPU for my wife's son. His birthday is in 3 months. Money is no object. What should I get?
>>55016274
>About 20% less performance for half price.
[Citation needed]
>>55014791
140$ more expensive than in murica.
This is the first time I used this meme.
AMD is finished and bankrupt this shit won't sell people will say fuck it and get a 1070.
>>55016307
>for my wife's son
Never thought I'd find an actual cuck on 4chan
>>55016282
i still play with 900mb gtx 560 just fine,because of consoles, graphic did not get much better...
you really need to look at spefic 2-5 titles that need more than 1 gig for 1080
>>55016307
>>55016324
Is this error in matrix or same fagg doing copy pasta shitposting?
>>55016324
>>55016358
If you're not going to help then don't reply.
>>55016013
yurocuck tax
>>55016307
>using a deprecated meme
>>55015570
Kek that image.
>>55014791
>300€
It would be much cheaper to order it from murica and pay import taxes what the fuck how?
>>55016307
Two 1080's
Four rx 480 for your Bull
>>55015728
>Who gives a fuck about tdp? Electricity is cheap.
How odd that Nvidiots have finally accepted this after a whole generation of MUH POWER DRAW shilling.
I wonder what changed their minds? :^)
>>55016411
nah most likely 200-250 euro and £200-£250
and still with that price point i'm fine with a straight conversion as pre import taxes the 480 should be around £150-170 for the 4gb version.
>>55015572
>calls it a rebrand
>1080/1070 Not shown on 2015 roadmap
>Pascal still cant ASync just like Maxwell
Who is the re brand now.
>>55016101
what the living fuck that should be £350 at most for a top end model.
>>55016573
Nvidia fanboys called the 390 a rebrand, yet Pascal is the rebrand this time around and they have the balls to claim it will be something revolutionary.
>>55014791
jokes on you the 1070 is 500-600 in eurpoor
Should I hold off a year before upgrading my GPU?
>>55016322
1070 will have more of a markup though...
>>55016717
maybe wait for Q4 maybe Q1.
>>55016630
>Pascal is the rebrand
God this fucking board is worse each day that passes. This has to be bait, but how can you call a fuckin' die shrink a rebrand you fucking idiot?
>>55016753
Little to no architectural changes is a rebrand.
All the performance of Pascal comes from clocks and more hardware.
>>55016753
Because it's the same architecture, just shrunk. It offers nothing new. Nvidiots will get into semantics about what constitutes a rebrand, but the fact is that Pascal is Maxwell at 14nm with higher clocks, and nothing more. That is a rebrand.
>>55015728
>390 came out, seemingly better than the 970, but it was late,
better than a reference 970, but an OCed 970 smokes it completely. the 390 is rebranded garbage.
Computer hardware is exempt from some customs charges or some shit.
http://www.dutycalculator.com/dc/109091-import-duty-rate-for-computer-hardware-is-0/
>>55014882
DP to DVI
HDMI to DVI
>>55016013
living in the first world is suffering i tell you
>shill believes they are undetectable
>>55014791
>>55016101
Enjoy funding those refugees.
>£
>€
Get off my 4chan you commie fucks.
>>55017024
Fuck off Trump
>>55016784
>but the fact is that Pascal is Maxwell at 14nm with higher clocks
Isn't Pascal 16nm?
>>55016013
8GB version
>>55016784
Re brand huh. Try get the same perf out of 28nm maxwell ams then you can call it a re brand. The 390x is a re brand however because it's literally the exact same as my sapphire 290x that has 1600mhz quad channel memory (reference 290x is 1250mhz ). And basically you are fucking stupid
>>55017707
ignore the HBM part
I'm hearing that 480s in crossfire lose to the 1070 in all games but aots. My bro works at kitguru.
>>55014791
>those taxes
I just puked in my mouth
>>55014882
>No DVI which is what the AYYMPOORFAGS use, shitty Korean cheapo monitors with DVI only
https://www.monoprice.com/Product?c_id=104&cp_id=10428&cs_id=1042801&p_id=12784&seq=1&format=2
$13 and works with korean 1440p dual link dvi monitors just fine
I know there isnt much info and shit out there.
But my 2gb GTX 770 is starting to act up.
Is this going to be a viable choice to replace it with?
I dont really game much nowadays...
>>55016784
I don't think that's entirely accurate. Not because I've read the specs and compared them (or rather an educated guess by analyzing the die diagrams) but because I've seen that several reviewers are saying that the 'IPC' on Pascal might be coming short from Maxwell. Some of those, are also defending the theory that the substantial gains on the new cards are mostly accounted by this renewed push for higher clocks and the die shrink itself (which mostly allows for higher clocks in the first place).
>>55018088
Which also could explain the ban on OC'ed 980 ti's and titans benchmarks for the first batch of the reviews.
>>55018043
you've got 3 weeks to wait. You really shouldn't be basing your decision on forum shitposting and wild rumors... or 1 benchmark made by the manufacturer himself.
Since your card isn't dead yet, I'd recommend to wait®
>>55014231
what the fuck it looks like shit, I was actually excited for the 480 too.
>>55018028
I'm hue.
21% tax would be a dream come true.
I checked the exchange rates for the pre order ont he 1080 yesterday and it came to 1.175 usd on a direct conversion. Which was not as bad as I expected.
Can I buy this on the 29th or will I have to get some meme shit instead?
>>55016784
>>55018088
Just to make it clear.
If it were the case of a simple die shrink, or rebrand as you've put it, the IPC shouldn't go down, theoretically. Of course, take into account that if they were experiencing some kind of manufacturing trouble that could very likely be the case.
>>55018370
if you were to wait you would at the very least get better deals on the 970's or may be even grab the 980's including the ti's for a better price. It depends on what will come from AMD, but nvidia is cannibalizing some of of it's own line, so whatever the case with the 480's, the overall prices should manifest a downward trend for Nvidia's last gen gpus.
>>55014791
$200 for reference design 4GB
$229 for cheap 8GB
$249 for high end custom 8GB
makes sense
>>55018453
I was just wondering if this card will come out on the 29th, or if that's just going to be some weird reference cards.
I'm looking for something to hold me over until Vega or the 1080Ti, my 980 recently kicked the bucket and the 1080 hasn't really impressed me.
>>55014231
So this is what my new GPU looks like. Hot.
>>55018524
I think that Raja said "available at the stores, on the 29th" during the computex stream.
Since we've already strated experiencing these random vendor leaks, it's probably truthful. Or not very far from the mark.
>>55018731
Groovy. Thanks senpai.
>>55015570
>that curve
>>55016929
What is this image trying to convey? You would think most of /pol/ and most of /v/ would be celebrating this shit. I know I am.
>>55018817
Hot newt.
>>55018817
the implication is that it's written by a supposed gawker shill because the children around /v/ and /pol/ are very exciteable and like to believe shadowy corporations and government groups care enough to try advanced psyop campaigns instead of believing they're getting trolled
>>55014231
that is shit tier design.
>>55018161
wait, what?
are reviewers deliberately omitting overclocked 28nm shit from their reviews because nvidia told them to?
>new poorfag gpu
>gets beaten by old gen nvidia gpu
amdshills will defend this
>>55018948
Agreed, it looks like a cheese grater, quite a letdown as someone who was planning on buying a sapphire 480
>>55019422
YOU POSTED THE WRONG IMAGE
ABORT SHILLING OPERATION
DELETE THIS
>>55019422
>it's actually better than the 980
I think I'm gonna cum.
so, should I switch from a 295x2 to 2 480s?
ive been thinking about it but idk
pros:
looks to be slightly faster
way less power
way less heat
8gb vram instead of 4gb
cons:
4 lanes blocked instead of 2
heat exhausted into case
have to sell old card which is a bit of a hassle
anything i missed? if the performance really is fury level on this card i might jump and go for it, the 295x2 is really a killer package though and can go down to single slot if i get off my ass and watercool properly
>>55019525
Thinking about swapping my 980 for it desu senpai
![9310_ellesmere_cam3_02_0010_4K[1].jpg](https://i.imgur.com/zCGXQFq.jpg "9310 ellesmere cam3 02 0010 4K[1]")
>>55019442
are reference coolers okay? are they loud or something? i think the reference 480 looks sexy
Polaris?
More like POORALIS LMAOOOO
>people actually care what their hardware looks like
You put it in the case and never see it again, what's this meme about?
>>55019422
>7.5% higher graphics score
>wins 3 out of 4 graphics tests
>$200 vs $550
this is what you call a "generational leap"
>>55019422
I will cream myself if it's actually inbetween 980 and Fury performance. The RX480 would deprecate the entire GPU lineup and the 1070 would get the rest of the spoils. What the fuck man
>>55015366
That's because its one particular moron who for the last year or more, has always screamed 'AMDPOORFAGS' non stop because IT is quite literally a pajeet shill and is doing what it must to get paid.
>>55019556
It's a sub-150W card so it shouldn't be loud with any cooler.
custom coolers are mostly for overclocking
>>55019585
>middle range beating middle high range from last gen at less than half of its price
>not exciting
The 960 and 380 are shit "generational leaps."
>>55019590
Don't forget the 480 might have massive overclocking potential
>>55019581
no one ever sees my undies but i only buy undies that i think look good
b/c I KNOW it's in here. it's about me.
>>55019556
its only 150w it has to be ok.
Retard here, my buddy used to buy Nvidia card only due to phsyX, is this still an issue today?
Or does AMD have something for it.
>>55019622
>not just buying what's comfy despite looks
FOOL.
>>55019422
>less than 100 points away from an R9 Fury
>>55019673
physx is only in like 10 games, and it can just be rendered on the cpu these days
not an issue
>>55019581
>>people actually care what their hardware looks like
Where were you back when the AMDrones were calling the 580 cooler a "VCR Tape" because it was aesthetically similar?
>>55019673
there are only 2 games that even use GPU physx and it's only used for gimmicks that you don't even notice
>>55019422
>Fury wins on graphics score
>Fury wins on physics score
>980 somehow wins combined
What fuckery is this? I call bullshit.
It should be done with taking result A + result B, adding them together then dividing by 2.
Unless Maxwell can handle doing both together better then Fiji, but that makes no sense as its something that GCN's ACE should be good at.
Especially once an updated, DX12 version is released.
>>55019673
You could vote with your wallet and not support unnecessary proprietary bullshit. Nvidia are assholes, I used to have an old Nvidia card just for PhysX, but now you can't even do that when paired with another AMD card, fuck them. AMD has GPUopen, basically the antithesis of Gameworks.
>>55019622
>no one ever sees me in my undies
poor lad. We got you bro.
>>55019715
>It should be done with taking result A + result B, adding them together then dividing by 2.
No. It's a separate test. PhysX streamlines the process giving Nvidia an edge.
>>55019762
But you'd think that if it made that much difference, it'd have given the card an edge in a physics only test anyway.
>>55019762
the physics test is entirely CPU based and has nothing to do with physX. Don't spread bullshit if you don't know what you're talking about
>>55019673
Nvidia gave up on physics since it can be done with cpu. They're more about sabotaging the games itself with gameworks nowadays.
>>55019811
Correction: the physics part of the combined test is running some calculations on the GPU, but it's using Directcompute. Not physx
>>55019685
>>55019691
>>55019733
>>55019878
Thanks /g/uys. I might pick up the RX 480, then.
maybe this is what kyle from hardocp was on about. the reference cards were probably hot garbage with mediocre performance due to throttling (not uncommon for reference cards) but now its been revealed board partner cards will release on the same day as the launch day, amd have dodged a bullet by letting reviewers test out the cooler and higher clocked aftermarket cards. they probably didn't want another 290x situation.
>>55020094
Stop it already,Kyle. A card with a single 6 pin connector is not hot.
>>55020128
speculation. we don't know how it will be.
>>55020094
thermal throtling is a Nvidia meme this round.
>>55020192
only for the 1080. the 1070 reference didn't have if i remember the reviews correctly. also the aftermarket 1080's run very cool.
>>55020180
>physical evidence
>speculation
>>55016753
Nvidia have been using the same architecture since the 8000 days. All they've done is shrunk the cores down and added more of them. Every new release like Kepler or Pascal is just a slight update to those 8000 cores.
>>55020255
please, link me all these multiple temperature tests which have been conducted on the 480 which i seem to have missed.
>>55020231
>the aftermarket 1080's run very cool.
Blatant lies from a blatant shill.
News flash, 80C wasn't "very cool" when AMD ran it, it's not "very cool" when Nvidia runs it.
>>55020292
care to post an AIB 1080 running at 80c?
>>55020292
http://www.guru3d.com/articles_pages/msi_geforce_gtx_1080_gaming_x_8g_review,9.html
>The card's temperature under heavy game stress stabilized at roughly 70 Degrees C. We note down the hottest GPU reading, not the average.
http://www.overclock3d.net/reviews/gpu_displays/msi_gtx1080_gaming_x_review/19
you were saying?
>>55020270
you don't need to run temperature tests to know that a 100-130W card is not hot.
>>55020306
Not the same anon but I was just reading this one recently when I saw the post on the first page.
http://www.guru3d.com/articles-pages/msi-geforce-gtx-1080-gaming-x-8g-review,9.html
480 performance
>>55020331
http://www.anandtech.com/bench/product/1714
http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1080/31.html
>>55020331
>roughly 70 degrees C
>on an open test bench
So 80C in a case shouldn't be uncommon, gotcha.
>>55020342
Yes, you do. Especially if there are indicators that 14nm process suffers from manufacturing problems that could be leakage.
>>55020331
>you were saying?
Pretty much this. >>55020382 Couldn't have said it better myself.
>>55020377
>talking about AIB 1080's
>retard posts reference 1080
top kek
>>55020382
funny you mention that. here's a detailed video showing the exact opposite :^)
https://www.youtube.com/watch?v=Jb6C6fd8tGc
it's even using the same msi 1080 from the reviews posted >>55020331
:^)
>>55020342
If the cooler is shit it can be hot. It's not like making a good cooler for sub 150w videocards is hard these days so expecting it to run hot is retarded.
>>55019556
Reference coolers are almost never ok, they're louder and less effective
>>55020409
>using the smiley with a carat nose
>>55020409
>funny you mention that. here's a detailed video showing the exact opposite :^)
I don't give a shit about some youtuber test. If you don't think putting multiple components that generate heat inside a case don't make the ambient temperature go up you must be a retard.
Then there's the youtube comments idiots.
>with proper airflow it should be cooler than open test bench
Epic meme, people say this but it's never the case.
>>55020472
>this damage control after he was well and truly BTFO'd by hard, tried and tested facts
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
>>55020472
I'd stop posting if I were you. You are making AMD owners look like retards.
>>55020366
For $200, how can you complain?
>>55014885
Like a dream
Should I sell my 290x lightning and use that money to buy a 480?
>>55020565
No, you should continue to shitpost until that card catches fire and burns your parent's house down around you.
>>55019535
The 1080 still cant beat the 295x2
So maybe you should try it, idk, if you get a nive waterblock for the 480's they could use just 2 slots
>$199
looks like AMD for life.
>>55018375
Except that maxwell was so heavily tuned for 28nm that a die shrink to a completely different process is going to cost them some of that IPC performance.
>>55020610
even a 1070 beats a 295x2 on average over several different titles
>>55020640
Im talking about Sheer Powah not fps's in your goyworks games or single gpu tests
I love my NVidia card but I want Radeon to succeed. I would love to switch back.
>>55014836
Anon please tell me when a higher end card will release. I want Polaris and 980ti-tier performance.
490? 480X?
>>55020681
Check videocardz.com
They had an rx 490 leak
If its true
Which i doubt
Nvidia is seriously fucked
It would be an unfair fight
>>55020681
>I want Polaris and 980ti-tier performance.
We don't know anything about Vega except that it will carry HBM2 and is probably fat Polaris.
>>55020681
2017. Rumors were that it was being "pulled to October" but there's no corroboration of any kind.
>>55020700
I certainly hope we're not waiting until October to get something better than the 480. I was expecting a bombshell at E3
>>55020694
With HBM2 being included, sounds like it'll be a thousand dollar plus card. Not interested in anything that goes over $650, maybe $700 if it's really fucking good.
>>55020727
>I certainly hope we're not waiting until October to get something better than the 480
Then buy a 1070 or something. I'm sitting with a 970 waiting for high end Vega since GTX 1080 is not good enough, i.e. it doesn't run 4k and 1440p at desired refreshes. Oh and yes, of course it's going to be expensive with HBM2.
>>55020727
Ah ye i was worried about that but i've been reading that HBM2 will have way better yields and since HBM doesnt scalate like 1GB/2GB/3.5GB but rather 4/16/32, even in bigger quantities it may be cheaper and also less power hungry, since the fury series memory controller technically meant adding 50w to the card's TDP
It may be 600-700 time will tell
Doubt amd would make a single gpu 1000 dollar card, that was just once on the 4000 series i think
>>55020745
I'm just in a bad position, got a 980ti, hated it, stepped up to 1080 ACX, if I keep it unopened I can get my $600 back and go back to AMD (Crimson is so much better)
Just bought a 21:9 1440p free sync monitor too. I want free sync.
>>55020767
>Doubt amd would make a single gpu 1000 dollar card
Firepro
>>55020781
Out of curiosity, why did you hate your 980ti?
>>55020781
if you already bought a 1080 just use it. AMD doesn't have anything to compete with it until much later in the year. They just aren't going to go in and fight at the ultra high level because it's bad business. They're going to dominate the mid-low range for the next 8 months, then update their high end chips with Vega and probably repeat the whole process over again in another 8 months after that.
>>55020802
Drivers. I was on the 390 before. I really like Crimson. It's not a bad card, I just feel like a retard paying double for 5-10 frames more and a higher setting here and there. Not worth it.
>>55020819
>I just feel like a retard paying double for 5-10 frames more and a higher setting here and there. Not worth it.
Of course not. Anything higher than the 390/970 was retarded and the 1080 is an even more retarded buy.
>>55020819
Do you still have your 390? If you do, return the 1080 and see what Vega has to offer. Otherwise, just use the 1080.
>>55020817
Well I mean if AMDbro can confirm that something a bit better will be released on Polaris in the next month or two that'd tide me over, but I really want AMD, especially with that freesync monitor.
>>55020836
Needed it for the Vive and development.
>>55020863
>Needed it for the Vive and development.
>Vive
I'd rather wait for second generation products or price drops. Current prices are ridiculous.
>development
Fair enough.
>something a bit better will be released on Polaris in the next month or two that'd tide me over
I'm almost 100% sure they won't release something bigger than Polaris in these few months. Maybe the 480x if you're lucky, but that would mean close to 980Ti/Fury X or above it.
>>55020787
Consumer grade cpu enthusiast or not
Not "Prosumer" whatever the fuck that means
>>55015689
fuck off samefag
>>55020342
with AMD you know it will be hot
>>55020472
Damn son you just got your ass ravaged
Also samefag
>>55020382
>>55020400
>>55020915
Perhaps I was not explicit enough. AMD would charge $1k for a "consumer grade cpu [sic] enthusiast" if they could justify it, ie. the performance was comparable to the Titan X over all the other cards. AMD prices there products based upon what the market permits as demonstrated by the Firepro.
>>55020472
dude there's nothing left, move on
>>55020952
Ah yes i know
I remember some athlons going for 1200 even though they offered not much more performance than a 400 dollar athlon but since it technically was the most powerful cpu back then it was expensive
I mean you can always justify a price one way or another but ATI's (RTG whatever fuck you) strategy is mostly price perf unless nvidia fucks up like back in fermi vs evergreen
>>55020681
They did say they will be releasing polaris cards from $100 to $300. We'll just have to wait for the 29th and see what the $250 and $300 cards are.
http://www.3dmark.com/compare/3dm11/11257751/3dm11/11263084/3dm11/11263252/3dm11/11167887/3dm11/11167781#
Crossfire benchmarks released. Looks so shit compared to the one they showed off
>>55021053
Probably because it is a brand neutral test.
>>55016282
8GB will be some overkill for a 1080p level card, because the most console devs have to work with is ~5.5, and from this I doubt even the worst of 1080p games will draw much more than 4-5 (the game engine takes up some of the console's 5.5 allotment too)
The increasing trend though has been caching as much game assets as possible into VRAM, which is why for instance GTA:V can have some pretty awful VRAM usage for what you get graphically.
Right now it's probably way too much, but if you keep your cards for 3-4 years it might be the perfect amount. 2012 era 2GB cards are getting long in the tooth even at 1080p, when it comes to VRAM.
It's also a perfect amount for 4k Crossfire, for whatever that's worth.
>>55021058
Far left result looks a bit underwhelming 2bh. I can't tell what's causing the performance gap between the same card in the far left benchmark and the one to the right of it.
>>55018910
This,
Tinfoilers don't realize things like Gawker are just generic internet businesses where one guy (like Denton) decides to pander to a specific audience (impressionable kids and liberals) in order to make a lot of cash with shitty gossips and bait articles.
Same goes for Alex Jones who panders to right-wingers and sells snake oil.
All of them deserve the gulag
>>55020982
>ATI's (RTG whatever fuck you) strategy is mostly price perf
You assume they have a choice. They haven't had the chance to lead the market since the 5000 series.
>>55021053
These benchmarks on the left basically confirm the 480x to be around 970/390 levels of performance.
>>55021053
18060/25803 = 1.43
43% scaling really isn't good, but this is only on some unfinished drivers... I wouldn't expect more than 60-70% at best in dx11 though.
>>55015570
>4GB means 3,5 + 5GB
That would make it 8,5 GB
>>55020836
>5-10 frames more
http://www.anandtech.com/bench/product/1715?vs=1721
>>55021173
Yeah i know, best of luck for then.
Hope they consider going back to ati since i loved that name
>>55021175
>These benchmarks on the left basically confirm the 480x to be around 970/390 levels of performance.
where are you getting that from? It's a completely different chip, the :C4 instead of the :C7 that's been confirmed to be the 480, and is probably even on older drivers, the test date was back in April...
>>55021228
>I'm blind
It clearly shows the two left benchmarks are the c7 aka 480. Match those scores up with the graph of 3dmark performance results.
>>55021266
The AMD shillforce is out. Marketing for AMD has started a little early.
>>55021407
Nigga what? I'm saying it's 970 performance and I've provided evidence. All these fanboys keep saying it's fury performance when it clearly isn't.
>>55021458
I'm not accusing you of anything. I'm saying he isn't blind. He's just blindly defending his employer he is shilling for.
>>55021458
It has a graphics score of 18060. How exactly is that 970 performance, you stupid shill?
I really don't want to get a new video card when my current setup isn't even 3 years old yet but my first venture into a dual gpu setup has been very disappointing. I just recently went to a 1440 monitor and my single 290x does alright. I'm wondering if I should just wait another year before getting a new card. When the games supports crossfire the 290x is amazing but maybe like 2 out the 10 games I play support it so the 2nd card is just collecting dust. I really have to question why crossfire and SLI even exist, is it just for benchmark bragging rights?
>>55021458
You are blind as fuck moron
>>55021504
it beats the fury x at $199
Nvidia is fucking FINISHED
>>55021266
>>55021458
14461 makes it between 290X and 980 on your graph, and it's the most recent test.
How is that below 970 in performance? 970 only scores 13548 on your graph, unless you're somehow blind to that score and instead are just focusing in on the 13141 score from 2 days before.
>>55021504
http://www.3dmark.com/3dm11/10884635
Barely even better than a 390
>>55021682
oh and this is completely ignoring the graphics score of 18060 which the 970 can't even touch.
>>55016547
Shows you dont pay the bills. It might be cheap for you as mummy & daddy pay for it.
>>55016671
£399 of the nvidia official site.
>>55021712
electricity is cheap as fuck unless you're a eurocuck
>>55016929
Man, /pol/ sure is deluded.
The press hasn't been free since the proliferation of television. Fucking nut jobs
>>55021720
Confirmed, he lives with mummy & daddy
>>55021745
>>55021712
>mummy & daddy
confirmed europoor
hows the immigration going? :^)
>>55021711
http://www.3dmark.com/3dm11/9694504
Fatty American kids up at 5am posting on here!.
Keep begging dad for that better gpu anon you can do it!!
So, no R9 490 before september/october ?
I currenly have a 560Ti and wanted to upgrade
480 looks nice but i'm willing to spend 100$ more for a better card
>ITT nobody understands 3dmark
>>>/v/
>tfw all amd fanboy hopes were crushed by one guy who actually bothered to check and compare the 3dmark results
enjoy that $200 390 performance i guess.
>>55021769
>only manages 18464 after a 26% overclock
Let's run the numbers on that.
GPU core is 1050, that particular chip is at 1328.
Main score on the 970 GPUmark page is 15730, that chip manages 18464.
1328/1050 = 126%
18464/15730 = 117%
26% overclock gets you 17% more performance, wow what amazing clock scaling.
>>55021866
Better than pascal thats for sure.
>>55021769
>http://www.3dmark.com/3dm11/9694504
That card's overclocked, numbnuts
>>55021866
Yeah that is pretty impressive. Maxwell can clock pretty high whilst also scaling well. They really did well with those node optimizations down at nvidia when I came to maxwell.
>>55021858
Actually, that looks like it is $230 390 performance. The results are for the 8gb version. That does kind of leave a question of how much of a degrade in performance will result from only having 4gb of VRAM.
so like when can we expect to be able to order this shit in europe? like amazon and shit
>>55021921
who cares. the retard claimed the 970 "can't even touch" that score, and he was proved wrong. Also that 970 is only factory overclocked. Check the specs for the gigabyte g1 gaming 970. It has no further overclock from the user.
Imagine what a juicy OC you can get from a 480 with good cooling.
>>55014231
>>According to our information,
i wish i could beat the fuck out of people that write this bullshit. no one cares about your rumors regarding garbage graphics cards. get fucked.
>>55021945
It's more like 390x performance and the 8gb will only matter for high resolutions.
>>55021858
The part you are missing it's that it's sub 150W 390 performance.
The 480 is miles more efficient per watt which means there's plenty of room for performance in the architecture. If they make a Polaris that consumes as much energy as a 390 it will have amazing performance.
>>55021974
Oh, sorry, I guess I was wrong.
Here's my revised statement: when the 970 is clocked up to the moon it can barely beat out this unreleased chip with unpolished drivers at a much higher power draw.
Does that make you happy?
>>55022002
The same goes for nvidia.
>fastest single gpu on the planet
>uses 100w less than a 390 in a typical gaming load
>>55022002
Do we know the watts used? We only know the TDP which is 150 watts.
>>55022044
Yeah Nvidia has amazing power consumption, I don't think AMD will ever trump them on that.
The 1070 (and 1080) will be the king of power consumption for a while, the 480 can't touch it there.
However the 480 is only $200, until Nvidia can compete at that price point, AMD will have the best card there.
>>55022046
Yes the AdoredTV guy did some math and it came out to like 95W in Ashes of Singularity.
That's pretty crazy efficiency
>>55022073
Here's the link
https://www.youtube.com/watch?v=PJ5wYVu-tM8&t=10m
>>55022073
>Yes the AdoredTV guy did some math and it came out to like 95W in Ashes of Singularity.
You mean this one?
>https://www.youtube.com/watch?v=PJ5wYVu-tM8
The one where he noted some "sketchy" IQ differences?
>>55022032
>clocked up to the moon
>gigabyte g1 gaming 1178 mhz core and a variable boost of 1329 max
>core is only 128mhz over stock
Lol people can easily achieve 1500 mhz core (see: core) on a 970. That factory overclock is nothing and no where near maxing the card out.
You sound butthurt. Feel free to try again next time though :^)
>>55022073
>>55022081
That's all theoretical, wait for the actual tests.
AMD states the TDP is 150W, I don't think they're going to be fudging the numbers up instead of down.
>>55015313
Who cares, it's better than my 670.
I would take that for 200 bucks.
>>55022093
It's obviously going to be less than 150, probably around 100
>>55022081
Ah, yes. The math he did based upon AMD's claim of 51% efficiency. But we don't know if that efficiency was in regards to multi-gpu scaling, watts, or even fan speed. It only says "51% utilization."
I wouldn't call that an independent, or even reliable, report of watts used.
>>55022062
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1070/25.html
These new cards have the best performance per watt in every resolution. Pretty damn impressive.
>>55022046
It's a single 6pin connector so 150W is the absolute max it can draw without overstressing the PCIE connectors specs.
Realistically it's 100-120W
>>55022097
Why would you say that? The load power draw is always higher than the TDP in actual tests.
>>55022102
>new technology is better than old technology
>impressive
Someone on reddit said they would have around 177k cards ready for release.
Will it be an actual launch this time?
>>55022106
You're still speculating. I'm not calling it an unreasonable speculation but it is still speculation. The previous point still stands. The only thing we know is the 150watt TDP.
>>55022102
Yeah Nvidia has been working on power efficiency in their cards for a while now, they really have that down.
>>55022088
Power draw scaling compared to performance gained on Maxwell is the worst since Fermi. For a 20% clock increase you see a 30-35% power draw increase, and the 970 already draws 180W under gaming load.
>>55022115
Considering how well maxwell was optimized for 28nm with its fantastic power efficiency and performance, it's impressive that they kept a similar level of efficency when dropping down to 16nm. They could have just given us much better performance for the same power draw as maxwell yet they've given us much better performance with a much lower power draw.
>>55015264
I have Sapphire R9 380 and I agree
>>55022073
https://youtu.be/tn4oIa4erVs
>>55022133
Also to be fair, AMD's Fiji architecture in the Fury and Nano is quite impressive power wise, hopefully Polaris will be even better.
>>55022146
smaller node naturally drops power draw. It's not a big achievement to gain energy efficiency by going from 28nm to 16nm.
>>55022126
seems likely considering die size.
>>55022164
>smaller node naturally drops power draw.
Provided there's no leakage like with Fermi.
>>55019556
its called polaris for a reason, they run cool and therefore dont have to be cooled as agressively making it quiter
>>55019601
The 970 is 145W and the blower reference cooler is pretty loud.
Just wait for the 3rd party open air coolers, they will be nice and quiet.
>>55014231
This is what an actual launch looks like Nvidiots, not stores recieving 6 fuckins cards hahahaahha
>>55022211
145 TDP, 180W actual power draw.
>>55021688
You stupid.
>>55022142
It uses just over 150w on average under a typical gaming load. That's much better than its direct competitor which draws about 110w more on average under the same typical gaming load. Even with a meaty OC the 970 doesn't reach anywhere near that 260w mark.
>>55022229
This
Nvidia have serious supply problems, I don't think their shitty die shrink has good yields at all
Feels like a soft Fermi launch
>>55022251
Yes and the 480 will be similar.
Blower coolers are always louder than the open air ones, just wait for them if you want a quiet 480.
>>55022300
480 will have a MAX draw of 150w due to the hardware limitations. If the card is using the MAX draw, then any peak that happens will shut down the system or the card.
This ofcourse is purely fantasy narrative created by delusional fanboys. Reality of the matter is RX480 will probably be around 110w+-10 or so. The extra room should be there for overclockers and situationals.
Fury X 2 fucking when?
>>55022300
it can't because it only has a single 6 pin connector. 150W is the absolute maximum it can draw.
>>55022334
That's not how it works
6 pin pcie can draw as much as it wants, 800W is possible
>>55022330
>dual gpu
Why?
>>55022351
>tfw no 2 pin 10 billion watt draw
>>55022351
Explain pls
>>55022325
A blower (reference) cooler at idle is like 20W on a 970 and it's still way louder than open air.
>>55014791
>jie vis dar nurodo kainas litais
>>55022358
No I mean the sequel
>>55022390
I doubt we will see it. I actually liked the name.
>>55019625
150w is max. Maybe it will draw only 100w. I mean it has only one 6pin connector, it must be very efficient. If they built this exactly with 150w then you cant oc shit because there is not enough power.
>>55022372
The Jews want you to think you need a bigger power supply but actually you just need to mod several PCI-E connectors into one so you get 4x6 pin into one end and can supply 2kW to your graphics card for higher OC.
>>55022410
Never knew there were so many dumb fucks on /g/ . Well I now know
>>55022409
150W is not the max, TDP is just the maximum power for typical applications.
When you run a stress test all GPUs go over their TDP and the 480 will be no exception.
>>55022428
It can't you mongrel it only has 1 6pin connector
>>55022427
Trolled :3
>>55022428
it is max, it's physically max that 480 can pull.
>>55022387
>jis rašo lietuviškai mongolų ekonominės raštvedybos nuotraukų dalinimosi svetainėje :D
>>55022437
Don't forget the socket delivers power also
>>55022447
Socket + 6Pin = 150W max
>>55022428
150 is the max hardware specification limit.
The presentation showed power at 150w NOT TDP.
TDP is NOT the max for power consumption, but rather a rated power usage for the gpu done by the company.
>>55022427
Yes, and they're generally referred to as amdrones.
>>55022455
That's wrong it's 200W on socket alone.
>>55015728
>Who gives a fuck about tdp? Electricity is cheap.
Nvidiots do, that's why they like maxwell, after all.
>>55022437
I've read somewhere that OP's version may come with an 8-pin connector
>>55022473
kek
>>55022469
kek
>>55022478
Power = heat
More heat = louder fans, and heating up your living space to uncomfortable degrees
>>55022445
>jis daro tą patį
>>55022473
Socket is 75W
6 pin is 75W
>>55022493
False. More power doesn't mean higher voltage. Higher voltage means more heat. You fucking Jew
>>55022500
Actually the power is literally the heat output.
Where do you think the heat comes from?
>>55022497
How can that be when 8 pin is 173W?
>>55022473
http://www.tomshardware.com/reviews/power-supply-specifications-atx-reference,3061-12.html
The 150W are the max power draw of the RX 480.
However, I'm sure board partners will release cards with an 8-pin connector or two 6-pin connectors to give the card more overclocking headroom.
>>55022497
You can get up to 150w from 6pon, if you don't mind running it out of spec. The other two wires are just ground. It's more important how good the power delivery circuit is on the card itself.
>>55022511
I don't know. That's how it is.
>>55022510
No it isn't, Skylake-E is 140W TDP but runs cooler than older 95W TDP chips you fucking electricity kike
>>55022511
8pin is 150W
Poorfag here, only have 300-325 USD to spend on a card.
Which one should I get? Would the 490 be cheaper than the 1070, or would nvidia drop their card's price?
>>55022428
stop it already.
AMD's last generation 6pin "150W power" card was the R9 270, which consumed 120W in gaming, 140W peak load in furmark.
RX 480 will be about the same.
>>55015728
>Who gives a fuck about tdp? Electricity is cheap.
AMDrones used to when the GTX 470 outperformed the Radeon 5850 and the 480 outperformed the 5870 though it wasn't called "TDP" back in those days.
>>55022527
Architectural improvements my man
>>55022527
You don't understand heat output do you?
It's not about how cool the chip runs, but how much heat it puts out into the environment around it.
For example, you can get a better CPU cooler to run your CPU at a lower temperature, but the heat output remains the same and actually circulates even more heat out into your room.
>>55022542
You're fucking wrong stop posting
>>55022535
Why would you say that when AMD themselves are saying it uses 150W?
>>55022530
I'd just wait for the 490.
The 480 is nice and efficient but it's still just an entry midrange card and not really something you can use for a long time.
The options are either pay a lot for a 1070 now or wait for the 490.
>>55022527
>TDP
>Thermal design power
>Thermal design
>Thermal
>>55022547
It's like you don't even understand basic physics you fucking moron
>>55022552
AMD always overestiamtes their power consumption, while Nvidia lies about it.
>>55022553
I have a long long time to wait. I would probably wait another 2 years, the 970 is good for me at 1080p.
>>55022567
Yeah no real point in upgrading if you play 1080p. The 970 is enough for now.
>>55022547
Sorry I'm smarter than you.
>>55022530
490 is not released.
1070 is going for $450-$500. MSRP is $380 but I don't see any 1070 coming down to that price for a while.
>>55022542
People have two different metric for heat.
The temperature of the graphics card itself and the temperature generated due to power inefficiencies.
>>55022553
I wonder, will 480 be better with low end CPUs or will they be as much of a bottleneck as for 970
>>55022566
You do realize how much that sounds like "Why my guy says anything he's telling the truth but when your guy says anything he's always lying?"
>>55022542
So this is what they're referring to when they call amdrones sweaty neckbeards. That housefire tier heat output must be really warming up that basement. At least it keeps their McDonald's warm.
>>55022552
They don't say it uses 150W. They say it can draw up to 150W in a worst case scenario. That's how they measure TDP.
Typical gaming power consumption is always lower.
>>55022583
>*When my guy says anything
Fixed
>>55022583
You do realize it's been proven almost every time a direct power consumption comparison is made, right?
The 970 isn't actually that much more energy efficient than the absolutely power mad beast that is the 390
>>55022585
Not too far from the truth.
I had a 390 and had to return it because it was a fucking space heater.
>>55022583
It's true in this case, would you rather somebody underestiamted or overestimated a card's power consumption?
>>55022599
>You do realize it's been proven almost every time a direct power consumption comparison is made, right?
Oh really now? Every time I look it seems the only time Nvidia is lying is when an AMDrone cherry-picks an extreme situation outside of regular gaming loads. Most every review I've seen has put regular gaming loads on target with Thermal DESIGN.
>>55022603
>It's true in this case
Prove it. Show me when an Nvidia card under normal game load is underestimating the TDP.
>>55022599
That's not even remotely true.
See: https://www.techpowerup.com/reviews/Sapphire/R9_390_Nitro/21.html
GTX 970 max power consumption: 201W
R9 390 max power consumption: 365W
>>55022571
>>55022567
My advice? If sell 970 and get RX480.
If you can get atleast $200 from that, then you'd be better off with RX480. EVEN IF the RX480 is only 970 level performance (which the benchmarks shows is around 980 level), the card will perform better in the coming DX12 games. So you'll automatically get a slightly better card without costing you anything. Plus with better warranty/longer support. There's also the support for h265 hardware decoder.
970's value is only going to go down rather than up. Meanwhile RX480 value will go up overtime.
tl;dr you lose no money from getting rx480 but gain value and performance (after selling 970).
>>55022614
Lol the 390 uses 70-100w when watching a fucking blu ray disk movie on your pc.
>>55022614
AMD and nVidia have different philosophies when it comes to TDP.
AMD sees it as maximum power
nVidia sees it as typical power
both ways are correct, what is not correct is you motherfuckers arguing about what is the "right way to display TDP"
>>55022614
I don't give a fuck what they draw when running LoL in 720p, things as important as power draw should ALWAYS be stated as a wost case scenario
>>55022631
Did you miss the point? Someone stated that Nvidia lies or underestimates its TDP. Are you trying to prove this assertion by discussing a different product by a different company?
>>55022618
I didn't say that
>>55022619
>Furmark
>Factory OC 390 vs stock 970
>
>>55022571
Reckon the 1170 or 590/580 would be out by 2-3 years? I would only get 1440p once it becomes dirt cheap and is pretty much the norm, but that is still a far cry away.
>>55022630
No way I can sell the 970 for 200, I would be lucky to get 150.
>>55022635
>AMD sees it as maximum power
Well then, AMD is calculating TDP in the wrong way.
>The thermal design power (TDP), sometimes called thermal design point, is the maximum amount of heat generated by the CPU that the cooling system in a computer is required to dissipate in typical operation. Rather than specifying CPU's real power dissipation, TDP serves as the nominal value for designing CPU cooling systems.[1]
Kind of funny how this is cited to how AMD calculates TDP for its CPUs. Either they calculate differently for their GPUs or they changed over the years what the standard is.
>>55022566
No they don't.
AMD says the R9 390 has a TDP of 275, yet under max load it uses 365W
TDP is always an underestimate, doesn't matter if it's Nvidia or AMD.
>>55022645
>I didn't say that
You stated in >>55022603 that "it is true [AMD] always tells the truth and [Nvidia] always lies" in regards to gpu TDP. Or did you respond to the previous post with "It's true in this case" without understanding what "this case" was about?
>>55022657
Check your craigslist and you'll find out. Remember, not everyone keeps up with the latest and greatest GPU news. So you'll find one buyer.
>>55022644
Calm your tits. I'm simply stating how inefficient the 390 is in general.
>>55022663
I think it's just ATi thing, I remember this debate going for a very fucking long time.
>tfw when you know 480 will not even outperform the 970
>>55022682
But even the 390 did that
Faggot
>>55022682
It should be faster than the 970, but even if it's not at $200 it will be the best card for the money.
>>55022636
>things as important as power draw should ALWAYS be stated as a wost case scenario
Perhaps you should tell that to AMD when it comes to calculating their CPU TDP.
>>55022679
>I think it's just ATi thing
It may be, but I'm going to go with what has a written record to support it for the meaning of TDP. If AMD wants to use a more stringent standard that's their choice but that does not make Nvidia's choice to use the recorded standard "a lie" as was asserted.
>>55022678
Thank you for admitting that you DID miss the point or else you are a self-entitled twit who responds to other people's posts without regards what is actually IN their post.
>>55022666
Are you a moron? That's a beast aftermarket card, not a reference. The reference 390 doesn't have 2x 8 pins.
>>55022630
Selling your 970 and buying a 480 is a side grade and isn't even worth it. Even if the 480 is 980 levels of performance you can stick a nice oc on that 970 to match that performance. Also there are less chances you'll get fucked over in gameworks games compared to owning amd.
>>55022666
>60watt difference in gaming
kek
>>55022678
>Peak
Reminder that Nvidia throttles their cards in Furmark to produce lower powere consumption readings. AMD runs the benchmark as fast as possible and therefore gives a true power reading.
Neither is a realistic representation of gaming power consumption.
>>55022636
Oh, and you might want to address what >>55022666 wrote about the 390. It seems AMD doesn't calculate based upon "worst case scenario."
>>55022700
>you can't OC RX480
Nice try. If there's a remote chance of RX480 getting 10% OC, it will be within Fury's reach.
>>55022583
>>55022619
>>55022635
>>55022666
>980 TDP: 165W
>Actual Power draw: 177-185W gaming
http://www.tomshardware.com/reviews/nvidia-geforce-gtx-980-970-maxwell,3941-12.html
>Fury X TDP: 275W
>Actual power draw: 220W gaming
http://www.tomshardware.com/reviews/amd-radeon-r9-fury-x,4196-7.html
>>55022696
Why are you so mad amdrone? Go back to bed.
>>55022700
If you can sell your 970 for $250 or higher, then it will be worth it for sure.
>>55022710
It seems like the reference 390 has a 6 and 8 pin making a max of 300w, meaning 275w at stock is probably accurate.
The nitro has 2x 8 pin meaning 375w max.
>>55022729
>Why are you so mad amdrone?
I'm pointing out how AMDrones claiming AMD always tells the truth and Nvidia always lies about TDP is wrong and you think I am an AMDrone?
You fail at basic reading comprehension or you're so enamored to see yourself posting on the internet you don't bother to read what you are responding to before posting. Pic very related in regards to which I think is more likely.
>>55022697
It's the same with reference, the R9 390 always uses way more than the GTX 970
The 390 is one of the most inefficient cards out there
>>55022676
Unfortunately where I live, people who build their own rigs are pretty up to date with the latest news so.. I'm not going to be able to sell it for 200.
>>55022689
screw bang for the buck itll still be a pile of dog shit
>>55022746
>It seems like the reference 390 has a 6 and 8 pin making a max of 300w, meaning 275w at stock is probably accurate.
Are we abandoning the point about "worst case scenario" now for "stock" or "typical" load?
>>55022724
We're talking about max power draw, not "gaming"
>>55022759
>I don't care if it's good, it's going to be bad
Good argument.
>>55022765
>We're talking about max power draw, not "gaming"
What do AMD or Nvidia claim is their "max power draw?" I thought the only specification they give in regards to power draw is TDP.
>>55022724
You do realize that games don't stress a GPU to 100%?
Try furmark or unigine valley to see what they really use.
>>55022767
No sorry I dont say it was going to be good.
Bang to the buck so $250 card $250 expectations.
>>55022751
Someone woke up a bit cranky today
>>55022754
390 has bigger die and also it's packed with more, everything.
>>55022768
They don't state max power draw. That's the point.
They say the TDP, which is max power draw using "typical applications".
Apparently "typical applications" does not include benchmarks, which always go over the TDP.
>>55022778
Packed with more inefficiency.
>>55022774
You might be wrong less often if you had the sense to know when to stop posting. You tried a personal attack last post and it failed then too.
>>55022779
>They don't state max power draw. That's the point.
Did I miss something? I thought you were rejecting the part of his post that TDP because it is not max power draw. Did you post something retardedly redundant because I'm pretty sure we started with some idiotic "AMD always tells the truth and Nvidia always lies when they state their TDP."
>>55022788
Imagine that, more calculations = more power consumption
>>55022805
>the part of his post *about TDP because it is not max power draw.
Fixed
>>55022765
>>55022772
why the fuck would literally anyone care about furmark or valley?
I'm not playing those pieces of shit, I'm going to be playing games with my GPU.
Jesus the shills are out in force today.
>>55022792
Shhhhh
Shhhhhhh
I think it's time to stop posting today. Mommy's calling you back to your crib.
>>55022805
I'm saying that "gaming" use doesn't show the max power draw.
If you want to see how much power the card actually draws under a worst case scenario, you use stress tests.
>>55022808
Except when it's not.
Notice how the 1070 is twice as fast as the R9 390 with half the power consumption.
>>55022836
Thank you for proving my point that you don't have the sense to know when to stop posting. Is it because you have in immature need to have the last word? Let's test it out.
I expect you will respond to this post with another shitpost. To which I will respond "It almost hurts to be right so often." You know what is coming in my next post so if you did not have an irrational need to have the last word you would respond to it before it is posted in my next response. If you respond to that as well you prove to have an immature need to have the last word.
>>55022765
>max power draw
No one cares about the metric when considering cost. Those "max" spikes happen in benchmarks only and arent related to real world performance or cost metrics.
Most people complaining about power draw complain about costs and heat. Neither of those are affected by the max power draw. The gaming performance is where the power draw/performance/power costs/heat generation will take affect as that will be the real world results in 99% of the scenarios.
>>55022688
>390
>200w more power consumption
>OC required to reach 970 level
>970 OC
>390 getting cucked
JUST
>>55022842
>I'm saying that "gaming" use doesn't show the max power draw.
What is the "typical application" of a consumer GPU? Gaming? Thus TDP would represent the "max power draw during typical applications" for the two cards mentioned?
>>55022832
It's important because you need to know how much power the card could draw under the worst scenario.
For example, if you are buying a power supply and need to know how much wattage you need.
Or need to know how much cooling you're going to need in your case to handle the card.
Or need to know how much power your system is going to use for your power bills.
>>55022842
>Thus TDP for the two cards mentioned would represent the "max power draw during gaming?"
Fixed
>>55022860
Thank you for proving my point that you don't have the sense to know when to stop posting. Is it because you have in immature need to have the last word? Let's test it out.
I expect you will respond to this post with another shitpost. To which I will respond "It almost hurts to be right so often." You know what is coming in my next post so if you did not have an irrational need to have the last word you would respond to it before it is posted in my next response. If you respond to that as well you prove to have an immature need to have the last word.
>>55022895
It almost hurts to be right so often.
>>55022875
Yes that's correct.
The point is that the TDP is meaningless because it's a best case scenario.
Let's say you have a game open on one monitor and a video open on the other monitor, it will go over the TDP.
>>55022900
It almost hurts to be right so often.
>>55022856
>1070
>twice as fast as the R9 390
learn to read benchmarks. It's only 50% faster
>>55022901
>The point is that the TDP is meaningless because it's a best case scenario.
No, it is a typical scenario. I'm sure we could pick out some old games that don't use more than 30-40% of the gpu's capabilities to determine its best case scenario.
It seems you function under some false dichotomy. If it isn't the worst case scenario it is the best case scenario?
>>55022926
What does your post have to do with what was written in >>55022913 as I don't see it has anything to do with worst, best, or typical use scenarios.
>>55022884
but if i literally never run those things i would never run into those problems. I could buy a power supply, budget for electricity, and buy sufficient cooling based on the gaming power draw...
how many people actually run their cards at 100% 24/7 using furmark and valley? who would even care about this shit except angry techbabbies whose identity is entirely based around the company they buy shit from?
>>55022913
It literally is a best case scenario. TDP is basically marketing language.
See this: https://www.techpowerup.com/reviews/Sapphire/R9_390_Nitro/21.html
GTX 970 TDP is 145W, yet during peak gaming it uses 191W
R9 390 TDP is 265W, yet during peak gaming it uses 277W
>>55022931
Post wasn't supposed to link to your comment. Just a general image when discussing power draw vs TDP.
>this buttmad poster who came out of nowhere and is getting angry over all of this shit
Top jej
>>55022940
Do you have two monitors?
Do you ever run more than two applications at once?
Do you ever run games that stress your GPU more than average?
Congratulations, you're already over the TDP.
>>55022944
>It literally is a best case scenario
Repeating yourself does not make your position any less wrong.
>TDP is basically marketing language.
>I want to rewrite what TDP means because I don't like how it doesn't represent what I want it to represent
I'm sorry but you don't get to unilaterally redefine the common usage of words in the English language.
>>55022959
Sorry that I'm right.
>>55022962
You failed to address a refutation and now claim you are right?
Are you nine years old? Because you argue like someone who is nine years old.
>Repeating my position rather than addressing any attack on it makes it not wrong.
>Claiming I am right means I right.
Let me guess. You also believe you are right unless you are proven wrong?
>>55022957
Please, link me to information that proves this.
>>55022978
Sorry I don't see any facts in this post.
>>55022962
Leave him. He's probably some 30 year old man who is taking this shit way to seriously. He thinks his fancy vocabulary gives him credibility. If he asks you a stupid question just paste it back to him. Fucking retard probably got molested as a child.
>>55023003
That sounds exactly like "I'm right unless you prove me wrong."
>>55023016
Still no argument to be seen.
>>55022992
Quite a bit here: https://www.techpowerup.com/reviews/Sapphire/R9_390_Nitro/21.html
>>55023024
If you act like you are nine years old expect to be treated like you are nine years old.
Be quiet, child. The adults are talking. Why don't you go read about Russell's Teapot? Then you might learn that people don't have to "provide facts" to "prove you wrong." It is your obligation to prove yourself right and just repeating yourself over and over with your fingers in your ears saying things like "Still no argument to be seen" or "I don't see an argument" doesn't fulfill that obligation.
Now I expect, just like the other child (if you are not one and the same) you also have an immature need to have the last word but I cannot be bothered to do the same test twice in the same thread.
Go ahead and pat yourself on the back for getting the last word and not being proven wrong.
>>55023063
That's a lot of words to say nothing.
>>55023063
If you act like you are nine years old expect to be treated like you are nine years old.
Be quiet, child. The adults are talking. Why don't you go read about Russell's Teapot? Then you might learn that people don't have to "provide facts" to "prove you wrong." It is your obligation to prove yourself right and just repeating yourself over and over with your fingers in your ears saying things like "Still no argument to be seen" or "I don't see an argument" doesn't fulfill that obligation.
Now I expect, just like the other child (if you are not one and the same) you also have an immature need to have the last word but I cannot be bothered to do the same test twice in the same thread.
Go ahead and pat yourself on the back for getting the last word and not being proven wrong.
>>55022682
>Nvidia fanboys constantly moving the goal posts
>pretty soon they'll say it won't even perform like a 380X
>>55022547
Temperature delta and heat output are different things. You're the one who is wrong.
![xxxuez83[1].png](https://i.imgur.com/mSjjLvN.png "xxxuez83[1]")
