http://www.techpowerup.com/222398/radeon-aib-partners-frustrated-at-amd
IT'S OVER, AMD IS FINISHED & BANKRUPT
>>54490115
Great Post!
>>54490115
Would be better if this piece of shit was knocked off, all they do is put shit out and broken promises
>>54490115
need INTEL to buy ATI
>>54490115
>AMD hasn't launched a new performance-segment GPU since 2012
That's bullshit though. Hawaii and Fiji have both released since then. I wonder how much Nvidia paid for that article? The all-expenses-paid trip to Texas was probably enough I guess.
AMD = fucked by inept leadership who seemingly have the 'just make it decent enough' attitude while their competition is always striving for more.
techpowerup.com = nvidiots who wouldn't wipe the cum off their faces after blowing Jen-Hsun.
This article makes no sense? Who are the partners? Where are the statements by the partners? Are they not taking part in the r9 400 series?
Then it says that there hasnt been a performance based gpu since 2012. The guy writing obviously meant architrcture, but even then wasnt fiji a new architecture tor HBM right? So hes still wrong.
I like supporting AMD and dont want to come off as a fanboy but something's genuinely off with this article.
>>54491749
and Tonga
>>54490115
Source: Their ass
Nice FUD-spreading though, nvidishills.
>>54490115
Fake or someone was mislead. The partners wouldn't tell some small european site that the new cards only 390 to 390x class performance when AMD has kept such a tight lip on the subject. Those partners would have their balls busted for leaking performance.
>>54490115
Such a poorly written article, it's obvious this is just anti-amd propaganda with no sources to back it up
>>54496460
http://www.nordichardware.se/nyheter/grafikkort/frustration-och-osakerhet-hos-amd-partners-infor-polaris.html
http://www.tweaktown.com/news/52056/rumor-amd-partners-showing-polaris-cards-computex/index.html?utm_source=dlvr.it&utm_medium=facebook
it provides the source at the bottom of the article. learn to read.
>>54496956
sources: nordichardware.se
that nobody can fucking read, sure buddy
>>54497013
>it's not in English so it can't be legit
AMDrone comprehension everybody.
AMD POORS ON SUICIDE WATCH
>>54490115
worst article I've read all week. ugh so badly written it hurts.
AMD is a step behind as always. nothing new here.
>>54497089
> so it can't be legit
Nvidiot.
>>54497107
>AMD is a step behind as always. nothing new here.
this
>>54497127
>being this mad
>>54490115
rebrAMD pissing people off for not ever making a new GPU in the segment and just rebranding over and over again?
who wouldathunkit
>>54490115
It's a blog.
>>54497157
pretty much. what would you expect from such a shitshow of a company.
>tfw there were actual people who believed amd would be competitive again
welp, sucks to be them. they can enjoy their $350 polaris 10 390x rebrand whilst i have my $379 ~faster than titan x gtx 1070.
>>54490115
Nice blog post
W1zzard is my hero.
>>54496956
Except nordichardware has no source mate, and these fuckers just copied their shit article and poorly translated it. It's shit covered in more shit, classic nvidiot believing everything he sees on the internet
Nvidia must be real scared of polaris, I haven't seen this much concentrated shilling since w10 launched.
this article is a joke, how much does nvidia pay these guys
>>54497492
>classic nvidiot believing everything he sees on the internet
you mean like all these amdrones who believe a paid shill on youtube? or the ashes of singularity benchmarks which haven't even been verified?
ok mate, ok. stay deluded.
>>54497496
DELETE THIS
>>54496956
DELETE THIS
>>54497496
BUTTMAD AMDPOORFAGS DETECTED
Poolaris had to get a respin, it will be months late
>>54490115
>AMD hasn't launched a new performance segment GPU since 2012
This whole article is a hit piece for Nvidia fqnboys to drool over.
AMD Friends
>looking forward to new gpus! hopefully we'll see great improvements all around! nvidia and amd are both releasing neat products, so excited!
Nvidia shills
>FUCK AMD BANKRUPT NVIDIA THE BEST! AMD FOR POOR PEOPLE LMAO NVIDIA MASTER RACE
>>54497496
>Nvidia must be real scared of polaris
you mean like the 2 or so threads full of amdrone damage control?
>the 1070 isn't really that cheap, it's secretly $1000
>p-polaris wasn't meant to go against pascal a-anyway
>w-wait till vega, y-you'll see
>n-no async
>i know all of this because my d-dad works for nvidia
>>54497587
AMD shills
>REEEEE NVIDIA KEKING MY GPU
>LE TESSELATION
>LE AMD MASTERRACE
>I HOPE NVIDIA DIES
Nvidia bros
>nice one, we get great performance for cheaper than we expected. more competition in the market for us consumers!
fixed it for you
>>54490115
DELETE THIS
>>54494661
pretty much this. i even own an nvidia gpu but this article is full on shill
What compels someone to be a fanboy?
>>54497768
http://www.techpowerup.com/212724/its-now-been-over-160-days-since-a-catalyst-whql-release
Man I fucking love TPU. They aren't pussy ass faggots held back by the fear of AMD revoking review samples.
That Russian site gamegpu tells it like it is too. AMD have since all revoked their samples. Now with no newest AMD Rebrandeon products Nvidia cards leads all their benchmarks, which amd fans aren't happy about. AMD just can't do anything right, even media censorship.
>>54497768
who knows. ask the amd owners (fanboys) since they seem to be the majority on every tech forum.
>>54497782
>he thinks WHQL matters
Remember: Nvidia's card killing drivers have been WHQL.
>>54497782
>Man I fucking love TPU. They aren't pussy ass faggots held back by the fear of AMD revoking review samples.
They'll be crying like all others did.
>>54497782
never forget
http://www.kitguru.net/site-news/announcements/zardon/amd-withdraw-kitguru-fury-x-sample-over-negative-content/
http://www.eteknix.com/things-go-from-bad-to-worse-for-reviews-of-amd-radeon-fury-x/
don't let the the cancer that is amd get away with this
>>54497800
>>54497800
And after a year or no drivers amd finally released their crimson driver that stopped the fan on all cards and killed it.
>nobody buying your rebranded shit
>stop drivers for a year to make everyone thirsty
>finally release it and make a big stink about it
>the driver kills cards, lol
>>54497828
Are we supposed to overlook Nvidia actually killnig cards when compared to a long gap between WHQL driver releases? Lets also not forget that AMD released a load of beta drivers within that period.
Also: moar driver hijinks.
http://modcrash.com/nvidia-display-driver-damaging-gpus/
http://www.zdnet.com/article/warning-nvidia-196-75-drivers-can-kill-your-graphics-card/
http://gadgets.ndtv.com/laptops/news/windows-10-is-killing-nvidia-graphics-cards-720443
https://forums.geforce.com/default/topic/862417/geforce-drivers/windows-10-geforce-drivers-are-killing-samsung-and-lg-notebook-lcd-display-panels/
(fyi for that last one /g/ actually tried to blame Intel for it)
>>54497800
Post proof of it killing cards.
Protip: You can't.
>>54497856
https://www.reddit.com/r/pcmasterrace/comments/3up8r2/amd_bugged_new_drivers_killed_my_gpu_and_others/
http://gearnuke.com/amds-new-crimson-drivers-bugged-killing-graphics-cards-report/
http://www.bit-tech.net/news/hardware/2015/11/30/amd-crimson-fan-bug/1
http://news.softpedia.com/news/warning-new-amd-crimson-driver-is-heating-and-killing-gpus-users-report-496867.shtml
http://n4g.com/news/1828663/amds-new-crimson-drivers-are-bugged-and-killing-graphics-cards-report
>buying amd housefire
not even once
>>54497892
AMD's drivers can't even do this even if they wanted to. GCN starts throttling itself hard when it reaches it's limit.
It's probably a few old cards dying from the heat and shitload of fake claims.
>>54497856
>ndtv
spotted the shitskin indian
>>54497656
>nice one, we get great performance for cheaper than we expected. more competition in the market for us consumers!
>http://wccftech.com/nvidia-geforce-gtx-1080-dx12-benchmarks/
16nm vs 28nm. +5fps. Amazing
>>54497606
>>p-polaris wasn't meant to go against pascal a-anyway
Polaris is r9 470/480 so this is actually true. I don't see anything to be mad about. There's a $100 "founders" premium on these cards till July. So I'll complain about it then when I know what Polaris actually is.
Anyone that thinks any modern drivers can kill GPUs is literal down syndrome retard.
>>54497930
>i-it's all fake, i swear!
like clockwork
>>54497930
This, the card will throttle and shut down long before heat kills it. Stopping a fan wont kill a gpu, feeding a lot of voltage will.
>>54497957
Its been baked into gpus for a long time - Nvidia's overheating killing cards is a good 6 years old. Nvidia killing cards via other means (namely by bricking the bios) is much more recent.
>>54497950
Nvidia didn't have hardware protection so they could. Not sure if they did it now.
>>54497930
the 20% was an actual thing >inb4 it still is.
But the problem was in the users who enabled AMD overdrive and never checked their temps.
>smells like pajeet...
>>54497938
nigga, there were rumors floating around wccftech saying polaris 10 is faster than a 980 ti. it was after i read those rumors when i finally stopped believing wccftech.
anyone who believes wccftech should literally kill themselves. i don't think they've ever been right, ever.
>>54497936
>wccftech
KEK, opinion discarded.
Why is everyone mad that Polaris isn't high end? Isn't the 480X in the sweet spot? Most people will be looking to spend that amount of money.
>>54498025
nvidia marketing
so good they made the 970 mainstream at high-end prices
>>54498025
no, the gtx x70 and r9 x90 have always been the best price/performance cards around the $300 mark. everyone is mad because nvidia are releasing a sweet spot card and amd aren't till next year according to their own roadmap
>>54498057
>x90 have always been the best price/performance cards around the $300 mark
Fuck no
5770
7870
280x
380
These were the price/performance cards for AMD in the last few years.
>>54498057
The polaris 10 is hitting the $350 mark and the leaks plant it at a little over the 980ti performance, just like the 1070. Despite the companies marketing this segment as mainstream, doesn't make it not high end performance.
>>54498108
>The polaris 10 is hitting the $350 mark and the leaks plant it at a little over the 980ti performance
Says who?
>>54498104
5770 was a shit hole, 5850 before the price hike was the best at that time.
>>54498108
no. see >>54497994
actual benchmarks of polaris 10 cards put it way less powerful than a 980 ti or even a fury x hence why it's now being rumored that the vega10 series will come out to compete with the 1070 and 1080. polaris 10 was NEVER meant to compete with the 1070 or 1080, its the natural successor to the r9 x80 series cards.
>Nordic Hardware reached out to us with their report, saying that AMD's partners "won't have any new cards to display at Computex and the only Polaris cards promoted to them from AMD are R9 390/390X performance class but for a mid-range price. Great value but no sign of any GTX 1080 contender". I thought I'd reach out to an AMD partner, and received a quick "don't think so" from a board partner, and then I asked for clarification to which they said "as of now, no information".
Nvidia showed off Pascal
AMD Pajeet still busy praying to his cow
June draws near any time now
Pajeets saids no no we won't allow
Oh oh I know I got a plan, one sure to please the fans
Ploop plop there we go, what do you think, Lisa-chan?
Computex draws near but amd got nothing on their hands
Their future looking grim with an uncertain lifespan
>>54498057
Are you implying a 1070 will cost around $300? Because you're wrong.
If polaris 10 costs 250$ and performs better then a 390 with half the heat I'll be happy enough.
>>54498057
770 and 290 vendor cards were pushing $450 at launch anon.
>>54498171
i should rephrase
between $300-$400 price point. hence why vega 10 is and 1070 are both predicted to be $350-$400.
>>54498165
That's retarded they showed Polaris months ago already, and I'm to believe they are showing new versions of obsolete 390s at Computex when their new cards are reportedly releasing imminently.
>>54498211
Still wrong.
GTX1070 will cost $449, i.e. founders edition i.e. only edition for the next couple of months. You didn't fall for nvidia marketing did you anon?
>>54498350
>only edition for the next couple of months
>>54498211
Also the 1070 has gddr5, vega will have hbm2 with a relatively big memory bus. Even if it's at the same pricepoint vega will perform much better. I think vega is the enthusiast/high end card/flagship though.
>>54498384
>Gigabyte founder's edition
Wow, yeah.
>>54498384
I thought nvidia promised a non-founders edition for $100 cheaper.
>pic
We'll believe it when we see it. 3rd party manufacturers usually take 3 months. Not 3 weeks.
>>54498350
>amdrones still can't get their heads round the 1070 being such a good price/performance card
>they still delude themselves into thinking is permanently $70 more expensive
you know nvidia did good when they get so much hate from amd fanboys
>>54498400
That don't look like a founder's edition outline
>>54498400
All founders are ref designs. That's probably a Gigabyte 1080.
>>54498406
Can we PLEASE cut the drone/idiot shit? It's polluting the boards with unintelligible babble, e.g. "they still delude themselves into thinking is permanently $70 more expensive"
Poor b8 too thanks.
>>54498406
Yep, Nvidia and Intel, at any point, if they wanted, could strangle fuck AMD to bankruptcy if they wished for it.
>>54498350
>founders edition
Isn't that just a card with stock clocks and stock cooler from nvidia?
>>54498384
>stock cooler with custom plastic casing and sticker
Wow, neat.
>>54498400
The Xtreme cards are their top tier custom cards, not reference models, which they've already announced
>>54497151
dumb frogposter
>no new architecture since 2012
wow, thats fucking wrong though.
>>54497892
Nvidia never made a card that over heats NEVER!
>>54498530
refreshes don't count as "brand new", dumbass
The last genuine architectural change in AMD GPU was back when GCN first launched. Everything else was nothing but minor updates and rebrands.
If anything, with the October release it seems that AMD are very confident in their matchup against Nvidia right now, high end 490X cards will be out Q1 or Q2 2017.
>>54498711
Nope, this signifies that AMD is a bit desperate because they know Polaris simply lacks the grunt to challenge even the lowest-spec GP104 (assumed to be the 1060 Ti with 7.5 GB of GDDR5, the 1070 will have the full 8GB as seen on one of the earlier non-GDDR5X Pascal). They need to push something out that will be as competitive as the GP104 flagship, and Vega 10 is their only chance at that. It is unlikely that it will absolutely beat the GTX 1080 in every aspect (given AMD's horrible history of over-hyping and under-delivering with astounding consistency), but it will be competitive enough. I don't think that will change the market share in their favor, since people who want the best and the latest will already have bought the 1080 and very few people will want a second-rate, second-place "me-too" card from AMD, no matter how close it gets to the 1080.
>>54498771
>Vega 11
>6144 stream processors
>>54498841
I doubt that. Where is the source for this figure? Some random forum post speculating about what AMD should be doing?
I doubt Vega 10 will even have 4096 stream processors at launch. 14nm lithography has been a bitch to refine and the yields simply won't be there.
>>54498817
>HBM2
>Bigger die
>Smaller node
Vega 10 will wipe the floor with 1080
>>54498862
If you extrapolate from Polaris 10, then Vega10 will be in the region of 350mm^2. That's perfectly feasible. First gen 28nm chips were just as big.
>>54498932
>Vega 10 will wipe the floor with 1080
>it's another "amd will rek nvidia" episode
no one is wiping the floor with anyone, especially not amd. remember when the fury x was meant to wipe the floor with the titan x with its hbm2? sorry mate, its not going to happen.
>>54498932
>HBM2
Doubtful, since HB2 NANDs and interposers have been constrained by supply
>Bigger die
Size doesn't always matter, since AMD is putting a truck load of garbage onto their dies to "modernize" the now-ancient GCN architecture. GCN 1.3 is as much of a refresh as Pascal is, admittedly.
>Smaller node
Not so. Samsung-Glofo 14nm FinFET is roughly the same physical size as the aborted 20nm design from TSMC, its just that the gates are made smaller so they can market it as "14 nm"
>>54498984
Seeing how Nvidia already start asslicking DX12 and Async i don't see how Titan X or any haswell series will be relevant anymore.
Fury will be the new 290x for another couple years while Haswell will gone in less than a year.
>>54499063
>haswell
u wot? this ain't a cpu thread
>>54499063
Typical AMDrone who doesn't even know what he's talking about.
>haswell
kek
>>54499063
>>54499109
MAXWELL HASWEEL MAXWELL
Fuck!
>>54498817
10 rupees have been deposited into your account
>>54499132
Wrong, and I also don't used Nvidia.
>>54499166
If only I could get pay for telling the truth I would never have to work a single day of my life!
>>54490115
>AMD hasn't launched a new performance-segment GPU since 2012
Guess they missed Fiji. An easy mistake with Nvidia's superior marketing and all.
>>54494661
This
It's also ignoring the fact that Vega will undoubtedly beat Volta by a significant margin
>>54500699
see >>54498984
people need to stop with this shit about companies destroying each other by big margins. there's not secret sauce in any of this.
Last exciting AMD gpu launch was the radeon 7000 series, Back in 2012
It's now 2016. Time to shit or get off the pot AMD
>>54500678
You mean that "overclockers dream" card when Nvidia dropped the GTX980Ti ridiculing AMD and clearly showing to everyone that AMD "will never have this"?
>>54500798
I mean beat it to market: leaked specs show an objective superiority over any current gpu and nvidia has Volta road mapped to 2018
People who think AMD is finished because they can't compete at the highest tier for only 4 months are stupid, is what I'm saying
>>54494661
Agreed. It's just a shill article.
I've used both, went from HD5770, to HD5770 crossfire, to gtx 680, and I'm 2017/2018 will upgrade to whatever I can get a good deal for.
Not sure what I had before the hd5770, i believe it was a 9600gt?
AMD has clearly had performance GPUs though, HD7950/7970/7990, r9 290/290x/295x2, fury/furyx/fury duo.
>constant rebranding
Pretty sure the only thing that hasn't been rebranded are the performance segments.
>>54497855
Pretty sure that was never confirmed, and it only affected a small number of people with third party software for fan control
>>54497864
Here's an older card dying from nvidia's drivers
https://www.youtube.com/watch?v=sRo-1VFMcbc
New rumor
Polaris delayed until October. AMD's been lying a ton and they're going to get sued
>>54497855
>tfw reddit is smarter than 4chan because they tested and debunked this myth instead of shilling for nvidia.
Sometimes the autism I see on here is painfully retarded, take a minute to do some research before posting your bullshit.
>>54501325
>amdrones find these clocks and massive heat output acceptable
>i-it's not amd's fault, i-i swear!
>i like playing my games at 10 fps with my card at 94 degrees
>i-its the Gaming Evolved™ way
wew
>>54501370
My 290x plays Witcher 3 at 1440p 60fps at 70c, but ok senpai. I guess I should be mad about that
>>54501545
Sorry I forgot to add I play at low/medium settings.
>>54501600
Only foliage distance on medium, everything else at high/ultra
>what is overclocking
Top kek
>>54490134
Great Trip Code!
>>54501370
Are you on drugs?
Are you okay?
>>54501545
>>54501600
>bragging about low settings
come again?
>mfw using an fx-8350 for the past 5 years and still getting over 60fps in muh gaymes while having superior encoding speeds
>>54497782
WHQL is literally a tax you pay windows for an extra shiny certificate. It means jack shit
people who get mad that things aren't marketed to them as shiny and elitist are fucking retarded. Like Apple and Nvidia shills. You literally like it because it's marketed as 10/10 premium even though it isn't
>hurr nvidia has so many whql drivers!
>nvidia has better same day driver support even though they just rename the driver and do nothing to change it!
>wow, only apple has retina displays!
etc.
>>54501633
>>54501600
Samefag, or posting from 2 devices.
No way 2 people can be this retarded and in the same thread at the same time
>>54498057
this is sort of true, but Polaris 10 is actually supposed to be at that sweet spot right out the gate and offer near fury speeds. Sure the 1070 will outperform it, but it's also $100 more expensive
>>54498658
>minor updates and rebrands
the jumps from tahiti, hawaii, tonga, and fiji all added new features in the hardware, improved power consumption, improved IPC, and more. Stop pretending that "muh rebrands" is any kind of argument against AMD, especially when the rebrands beat new nvidia cards
>>54498984
>with what we know about the hardware of the new card, the performance should be unparalleled
>no dummy, amd will be bad because uhhh... amd is bad lol
>>54501803
Did the core architecture change like it did between Maxwell and Kepler? No?
Then it's a fucking rebrand
>>54501661
the guy who wrote
>low/medium settings
was an nvidia drone in denial
this guy>>54501633
is either the same one or just completely retarded
>>54501824
GCN1.2 != GCN1.0
>>54497828
Techpowerup is a known NVidia shill site.
>>54501325
That powertune has been built in the cards since at least the 5000 series. I could run my 5850 with a single Accelero S1 heatsink and no fan on it - it just scaled itself down whenever temps rose too much.
>>54501880
Are you stupid? How is mentioning the release of vendor graphics card shilling? It would be shilling if they wrote articles calling amd bad or something like that.
>>54501906
Nvidia shills refuse to accept anything positive about AMD. They don't realise AMD usually does things first to push the industry forward. My HD5770s have dedicated tessellation units, and then nvidia comes along with crysis 2 and gameworks stupid over the top tessellation. I see no frame drops setting tessellation to 2x, and slight drops at 4x, but 64x is just extreme overkill
>>54501906
Not true, my HD7950 reference fan stopped working and my card hit 110ºC whihc by then I had to shut it down.
Older cards can't downclock as newer 200/300 series, older cards have predefined clock values even the boost versions of the 7950/7970 are the same.
>>54502016
Back in the ATi 8500 they had tessellation, it was never adopted by anyone though
>>54502016
tessellation is part of the dx11 specifications without it you only get dx10.1 certification.
>AMD hasn't launched a new performance-segment GPU since 2012.
wtf what was fiji then?
i'm a 980 ti owner myself but really what was fiji then if it was not a "new performance" segment???
>>54502125
all the enthusiast buy the 980ti or titan x, only amd fanboys got the fury x.
if youre an enthusiast you dont buy what was "almost" first place
>>54502155
so that makes fiji a rebrand?
wtf?
>>54501803
>with what we know about the hardware of the new card, the performance should be unparalleled
They literally said the same thing about the fury series. There is no secret hardware sauce that'll cause anyone to be unrivalled. Amd have said every single gen that their top cards will BTFO nvidia and be "unrivalled". They've given everyone false promises so far. Nothing they can do will change my mind about them now. Nvidia are just plain better at the high end and they've showed this. I'm getting a 1080 ti, amd aren't good enough in my eyes.
>inb4 fanboy hate comments because I'm buying nvidia
>>54501880
>aibs, not even Nvidia sends them free cards
>"mind reviewing our cards nigga?"
>k sure bra
>suddenly it's an Nvidia conspiracy
>>54490115
>AMD hasn't launched a new performance-segment GPU since 2012
>Fury X wasn't performance-segment
What a load of shit
>>54502185
No, it just makes it that shitty card no one but idiots buy.
performance segment is something like the Titan X or like the HD7970 was.
>>54502223
you still have yet to explain how you think 4096 GCN 5.0 stream processors at 14nm manufacturing could possibly not out perform the current 1080
I wonder who's going to buy the $1500 fury pro now that the faster $599 1080 is announced.
It's literally dead on arrival.
>>54502408
Fury Pro almost doubles the raw performance of the 1080, it probably will be used as development kit for DX12/Vulkan VR, and for niche GPGPU stuff
>>54502421
>Fury Pro almost doubles the raw performance of the 1080
>two 28nm GPUs that could barely keep pace with a stock 980 Ti is somehow double the performance of a new 16nm GPU that's 40% faster than a Titan X
This cuck actually believes this
Holy shit, this level of delusion is staggering
>>54502421
So potentially enterprise guys getting it free from a company write off but realistically nobody because enterprise is so up the cuda ass that isn't even funny
>>54497936
>114% Performance at 60% of cost, 80% of TDP, at ~55% Die size is a failure
C-check mate nvidia fags.
>>54502456
It does 16 TFLOPS, the 1080 does 9
AMD has always had much more raw power than most Nvidia cards
>>54502457
>because enterprise is so up the cuda ass that isn't even funny
That's why it's niche GPGPU, and not GPGPU
>>54502495
TFLOPS does not matter. The fact that the Fury Pro has no market alone is worth the look on AMDrones' faces when AMD stock prices plummet over the coming months.
>>54502537
>moving the goalpost
Good work senpai
>>54502546
>choosing an irrelevant goalpost
Not my problem if no one wants to play on your retarded pitch
>>54502456
>>54502495
>>54502537
>>54502546
>>54502563
>Fury Pro is marketed to gamers
Dumbasses
>>54502563
But you're
>>54490115
not going to lie, I'm getting pretty upset that AMD hasn't said shit of substance for Polaris 10 and they let Nvidia beat them to the punch when it seemed AMD would release well before Nvidia
I JUST WANT YOU TO DO WELL AMD WHY DO YOU ALLOW YOURSELF TO HAVE SUCH A SHITTY MARKETING DEPARTMENT
>aib partners probably knew months before the polaris announcement that full blown gcn/vega will be available at a later date since amd wants to refresh low to higher-midend for shared competition
>now aib partner suddenly gets frustrated
bullshit article
>>54502664
you act like a 700$ card will find customers in the mainstream section
>>54502380
but that has nothing to do with techpowerup claiming amd has not released a new performance based gpu since 2012
amd has released a new performance based gpu since 2012, and its fiji.
actually fiji wasn't their only new performance based gpu since 2012. hawaii pro too was a new performance based gpu released in 2013.
amd might not have an entire new catalog in all categories, but they have been releasing new performance based gpu's.
>>54502958
Don't forget 290x
Fury pro is fucking sexy
>>54502958
techpowerup wasn't claiming anything they were quoting some people, techpower up is a news aggregator they rarely do reviews on their own.
And performance base means the top of the line GPU, the strongest single GPU available that money can buy, which AMD hasnt had since the HD7970.
>>54502380
Fury x was actually pretty good.
Wtf
https://youtu.be/NzOD0WaUJVs
Thoughts? I think they're correct when they talk about AMD (20 min mark I think). AMD will need to pull a rabbit out of a hat to be able to compete with just low end shit and them only focusing on the low end would be dumb.
>>54503011
They don't call it the "Fury Pro"
It's the "Radeon Pro Duo"
Fury is just for the single GPU cards.
>>54503538
If you want the top of the line performance you only have Nvidia available.
>>54502480
>114% performance at 55% die size is a failure
yes
it is literally just an overclocked 980
>>54505592
They traded blows you retard
>>54503756
> them only focusing on the low end would be dumb.
Its to secure OEM and laptop deals which is 1) where lots of money is and 2) areas where Nvidia massacre AMD.
Focusing on the low and mid ranges is the smart move to regain market share, which is what AMD have stated they are going to do.
>>54494661
techpowerup is hideously biased towards nvidia. it's like the daily mail or the guardian of GPU news.
>>54491732
this is by far the worst suggestion i have ever seen
and yet somehow people genuinely think having an out-of-country company having a monopoly on american market (or any other for that matter) is somehow a reasonable plan
>>54490115
At least there's some smart people in the comments
>>54490115
>techpowerup
>btarunr
who did he interview?
his mom?
where is his source?
his ass?
>>54505725
No it doesn't, the Fury X isn't even on second place, that place is for the GTX980Ti and the first place is for the Titan X.
Once cooled properly the Titan X BFTO everything even a highly clocked GTX980Ti.
If you want the best performing card you can buy with money that card is from Nvidia.
>>54506044
>If you want the best performing card you can buy with money that card is from Nvidia.
The most expensive card, sure.
The best performing? Maybe in select benchmarks that ignore compute performance and don't use DX12 or Vulkan. Generally, no.
>>54494661
MSI should be worried, their Nvidia cards sell less than EVGA or Gigabyte probably less than Asus too.
They got lucky with the 390 and their card overclocked better than the Shappire version for a while and it was the recommended 390 card to get for months on the overclocker forums which tons of people got even over the Shappire one which has better cooling.
>>54506070
Nice delusions faggot, Titan X is the strongest card money can buy.
Your performance/ratio price are only poorfag excuses, you probably buy used too and have never been first place on anything.
>>54505896
amd is dogshit
>>54490115
>own 290x
>still plays all games AT MAX SETTINGS
>meanwhile nvidia keeps releasing new flagship $899 GPU
>"we totally promise this one is better, what are you, poor or something?"
>meanwhile still no dx10 support
>290x is faster than 980ti in native dx12 games
wasn't the 980ti released around the same time, last year?
>>54506101
>implying you can even afford one
>projecting this hard
>>54506172
oh irony.
Thats very "rich" coming from an AMD shill.
>>54506223
>AMD shill
>don't even own an AMD card
Buying a Titan X is retarded, you're better off buying a 980ti and using that $300 elsewhere.
>>54506101
>Titan X is the strongest card money can buy.
even the 295x2 is a faster card
>>54506256
>even the 295x2 is a faster card
Dual card*
>>54506283
no, it's a single card, it's dual gpu, you retard.
>>54506287
semantics.
>>54506287
meanwhile there's no dual-GPU titan and never will be
so "the best money can buy" (and fit into a single system) is either
n * SLI titans (n gpus)
n * CFX 295x2s (2n gpus)
the latter will always be higher performance.
>>54506319
290X2 needs crossfire profiles, with out them which isn't rare it performs worse than a 290X
>>54506293
nothing to do with semantics. it's a single card.
>>54506319
>the latter will always be higher performance.
the latter would be higher performance if it was crossfiring separate cards anyway. SLI is garbage that hasn't been changed in a decade and crossfire scales much better.
>>54506341
>nothing to do with semantics. it's a single card.
It is semantics, mr autismo.
Do you know what context is or do you need to look it up on a dictionary.
>>54506350
you need to learn what semantics are you gigantic retard. while you're at it learn the different between a 'card' and a 'gpu'.
>amdrones have gotten to the point of denying the 980 ti is the superior card in the enthusiast market even though it btfo's the fury x in 95% of games instead of 3 amd sponsored games and a glorified tech demo
How can you fall so far
>>54506044
Pretty sure a fury x is faster than a Titan x in some titles.
That counts as trading blows.
>>54506585
Also don't forget this is the Nvidia biased TPU, so you can add on a few extra % for all AMD cards.
>>54506585
>980 ti 15% faster at 4k
>probably double that at 1080p
And they said the fury x was going to destroy the 980 ti. They even said it'll destroy the 980 ti at higher resolutions. Welp, glad I didn't fall for the amd meme.
>>54506670
if you believe in benchmark sites..
i lately happen to have strong misbeliefs in almost every site.. every benchmark ever could be bought by either company
>>54506341
>sli hasnt been changed in a decade
>they literally just announced a major upgrade to SLI with the 1080 announcement
>>54491749
I scanned the comments, the author is claiming that those cards are "non-mainstream enthusiast" cards, and not "mainstream performance" like the 970 or 980
>>54506670
They even said it'll be overclockable by a good amount
I'm just waiting for Polaris to drop before i make a decision, sorry that i'm not running out to buy a 1080 immediately mighty nvidia shill gods.
Maybe my friends are cheapskates, but i don't know anyone with anything higher end than a 970 or 390x, Nobody in my circle has any of the super high end cards and i'm pretty sure outside of people who want the PC equivalent of Ferraris most people have mid to upper mid cards.
>>54507220
I'm using a 260x :D
>>54506670
>Reference Fury X 5% faster than GTX980Ti
>Comparing overclocked cards to stock cards
>>54507296
Were you expecting better from an nvidiot?
>>54507220
I had a 970 and was happy until 3.5 and then has to sell it due to the shame. Got a used 980 and figured fuck it I'll get two. For a while I was happy but one day they weren't the best card anymore. I said fuck it and sold them for two 980 Tis. 970 is a gateway drug.
>>54502125
Fiji was a HBM tech demo not a performance GPU.
>>54501325
Leaving your card at 94 degrees for 35 minutes straight is not smarter.
>>54502421
>with 4gb of vram
>>54508047
>Not knowing Vulkan and DX12 can adress each 4GB HBM individually in crossfire or on the duo pro
Same for the 295 as well
>>54508012
educate yourself
>>54508012
It's for science. And for proving nvidiots spouting bullshit wrong. Therefore it's well worth it
>>54507112
>the author is claiming that those cards are "non-mainstream enthusiast" cards, and not "mainstream performance" like the 970 or 980
wat
>>54508089
>muh whole 2 dx12 games
>>54509248
Oh and its only on a per game basis
>The news arrived through Nordic Hardware saying that AMD's partners "won't have any new cards to display at Computex and the only Polaris cards promoted to them from AMD are R9 390/390X performance class but for a mid-range price. Reports say Polaris 10 can't hit 850 MHz reliably, whereas Nvidia already demoed the 2000 MHz marker with their GPUs.
>>54507112
So I guess the 285 and 380x don't count?
Or did the author intentionally ignore that?
That's what AMD gets for wasting time on freetard drivers. This wouldn't have happened if they were more like nvidia.
This thread is the purest form of cancer
>>54509855
No wonder they basically dumped it and moved straight to Vega
>>54490115
What about the r9 380 it's way better than the 970 and I have a 970. I won't buy Nvidia for this go-round because they lied to me about how the memory performs in my card. When a game takes 100MB of memory more than 3.5Gb I lose 20% to 50% of my FPS
>>54511736
You got any factual proof of this backed up by multiple sources?
>>54511736
https://youtu.be/COEwgUI6D3w
https://youtu.be/p6MDUjFiUR4
https://youtu.be/IW-H0qNw38M
https://youtu.be/ljHgKO1E3s4
https://youtu.be/vG21VQDpz2E
You were saying?
>shills shit posting sourced on "vague rumors" when Polaris has already been shown to perform at 60fps 1440p on the card
>>54512008
60fps 1440p for Hitman*
>>54509855
>>54510395
Polaris always was Low Mid range. It was never made to compete with high end cards.
>>54511815
nvidiots will defend this.
>AMD fags don't use tower coolers on their card
>Nvidiots aren't using tower coolers either
enjoy your broken card in 2 years
>>54490115
As someone who doesn't know much about graphics cards, can someone explain why people seem to get so aggressive when defending Nvidia? Are the cards actually that good to deserve such rampant and unapologetic fanboyism?
>>54512214
Amdrones will defend this
>inb4 oc
They clearly say in the video that their 390 is over clocked by 70mhz from the factory and they couldn't oc it more.
>>54512313
nvi ots will d fend th s.
>>54512214
Isn't this game meant to be pretty broken for both sides?
>>54512400
Very broken...390 still diggin that bitch out.
>>54512400
No its perfectly fine.
The real game with issues is gears ultimate
http://www.extremetech.com/gaming/223858-the-new-gears-of-war-ultimate-edition-is-a-dx12-disaster
And also quantum break with even the higher end cards not able to achieve a locked 60fps on even medium settings. Amdrones will defend this though.
>>54512308
Mostly because people who worry about Nvidia having issues are overly concerned AMD users who worry on behalf of NNvidia owners.
GTX 970 bought almost on release day, have had zero issues with mine that couldnt have been fixed, any problem with games get patched really fast and so does driver problems like that one where you would get a black screen when upgrading if you use two monitors, unplugging the second monitor, updating the drive and then plugging it back was a easy workaround yet you still see lots of AMD people making a big deal out of it.
Meanwhile at AMD, the 380 was released almost 5 months ago and still no fix, whats even worse is that AMD don't comes out and say it is a hardware defect and get a RMA(Nvidia does this on their forum or support if they can't replicate your issue they even help you get in contact with the vendor) so you're stuck with a card that could be fixed by software or a card that could just have a hardware problem but you can't be really sure because AMD won't confirm either so all those people are waiting for a fix that might never come because it can't be fixed by software.
https://community.amd.com/thread/199291#2718444
>>54512362
>gtx 970
>cheaper than a 390
>overclocks to gain 20% extra performance over a 390
Wew. Amdrones will defend paying more for less and paying more for a rebranded 290 which is $100 cheaper.
>>54512447
That sounds like a developer problem not an AMD problem. Do you even brain? Nvidiots will say "Amdrones will defend this though."
>>54512490
>Mostly because people who worry about Nvidia having issues are overly concerned AMD users who worry on behalf of NNvidia owners.
Holy shit this so much.
>>54512362
You don't own a GTX 970, do you?
With currents drivers the GTX970 is on par with the 390X on Far Cry Primal now, and the driver that increased performance is like 4 drivers releases old already.
>>54512491
vi ots w l DeFend ThiS.
>>54512491
Not sure where you're from but a 390 is like 40 AUD cheaper... which admittedly is only like 25 USD right now
>>54512513
That's the fucking point you retard. Instead of everyone calling out the devs for shit ports they instead praise the devs because their amd cards get fucking 5 fps over the nvidia competitor so they're satisfied with the shit Performance as long as they're doing better than nvidia.
They are literally defending shitty developers. Watch how quantum break benchmarks are posted in this thread soon by amdrones.
Remind me, does nvidia do approximation architecture graphs like amd?
I remember there was something for maxwell.
>>54512518
You're retarded of you think owners didn't care. If you want to swap anecdotes, the only people that worry about bad AMD drivers are overly concerned nvidia owners.
>>54512540
>bad cutscene performance
>way superior in game performance on nvidia cards
I'll take that 2bh
>>54512580
You mean whitepapers?
Nvidia releases whitepapers for their architectures with each release, the whitepaper cfor the full Pascal chip is out already, they might release one for the 1080 too maybe more detailed.
>>54512584
No. No one cares about amd drivers. Everyone laughs at them, not cares for them compared to amd owners who are always like >d-don't buy nvidia they'll cuck you in 3 years even though I have no evidence and I like talking out of my arse.
>>54512584
>The only people that worry about bad AMD drivers are overly concerned nvidia owners.
No surprise really, most people who buy their first GPU buy AMD because budget concerns, then reality hits you and you realize what mistake you made by buying AMD, yet you don't see all those people crying about the AMD boogeyman, instead they just stop buying AMD GPUs that don't work.
>>54512597
DEFEND THIS!
>>54512638
I have evidence though - I sold my 770
Have zero issues with amd drivers, didn't have them in the year 2007 either when owned another amd card
I just love installing drivers these days, with nvidia it meant absolutely nothing for performance - zero benefit
amd - oh wow I gained 7 fps for free!
>>54512681
Don't worry, Pascal supports async compute, right? Right?!!
>>54512638
>no evidence
Once Nvidia stops optimizing a previous generation for games their performance tanks. See: any Kepler benchmark on a modern game. I don't think they're specifically nerfing them but they aren't improving them either.
Look at the 7 fucking 80 ti getting cucked by midrange AMD and Nvidia cards while the Hawaii chips it used to compete against destroy it now. The same shit is going to happen to Maxwell because Nvidia's performance is heavily tied to specific DX11 workarounds. But it's all about those game ready drivers and getting good benchmarks on the day the game comes out, right? Who cares about longevity?
>>54490115
>techpowerup
That site is an even bigger joke than wccftech.
>>54512681
No one should defend this game. It's fucking shit. I bought the intro pack for £8.80 off greenmangaming yesterday and I played the first 20 mins and was so fucking bored. How is everyone saying this is the saviour of hitman. Fucking terrible game.
