http://www.forbes.com/sites/patrickmoorhead/2015/11/12/advanced-micro-devicess-head-of-radeon-technologies-group-raja-koduri-talks-about-the-future/
HAHAHAHAAHA
get ready for the fourth generation of gcn 1.0 goyim
Please limit poo comments to 1 out of every 4 replies.
just poo in the gpu senpai
>>51320288
why is this company not bankrupt yet
>>51320288
>planned
He means possible we all know they have no money.
>mfw Tahiti rebranded yet again
>mfw Hawaii with GDDR5X
>mfw Fiji with HBM2
>mfw Greenland anal destroyer
>>51320288
DESIGNATED REBRANDED GPUS
E
Ș
|
G
Ǹ
Ȃ
Ț
E
D
R
Ȇ
ß
R
A
N
D
ǯ
D
Ǵ
P
U
Ꚃ
>>51320288
to clear this up, both AMD and Nvidia generally only introduce 1-2 new GPUs per year. The rest are rebrands.
This is the media latching onto something, that has nothing to do with AMD's current financial situation.
>>51320567
>implying
>>51320288
>next gen poo technology
hahaha
>>51320567
that's not true at all, nvidia has kept the same pattern of releasing a new architecture every 18 months for a very long time now
>>51320456
They're going pretty overkill with that HBM2, like shit nigga what are you even going to do with the 512GB/s HBM already gives you?
>>51320312
ugh, why do we allow /pol/ memes to spill over and ruin tech discussions?
Anyhow, I wouldn't look too much into it. AMD is forever destined to be behind nvidia on graphics tech, they just need to survive to make sure innovation still happens.
>>51320791
>/pol/ meme
This meme was from an Indian on /int/.
>>51320802
yeah, idgaf. If I wanted to know, I'd visit /int/.
>>51320288
>get ready for the fourth generation of gcn 1.0 goyim
Are you fucking stupid, or what?
GCN is not going away. Years ago 5 generations of the GCN arch were laid out.
They'll start selling better bins of the Fury X and Fury Nano. The dual GPU Fury card is also coming out.
On top of this they have two new dies.
If they had all new FinFET chips from top to bottom of their product stack no one would want to foot the bill.
>>51320791
>ugh, why do we allow /pol/ memes to spill over and ruin tech discussions?
How about you fuck off back to plebbit? The reddit exodus killed 4chins.
Anyway, hopefully they will be proper GPUs, with 2nd gen HBM, and proper tesselation so that the tesselation penalty will be the same for both vendors as opposed to 20% nvidia and 40% amd.
Other than that, POO IN THE LOO.
>>51320835
The thing is AMD will eventually cannibalize itself with Zen APUs and make low-end GPUs irrelevant.
They NEED to be competitive with Nvidia in the top-tier GPU market, unlike their processor tech which only needs to match Intel's midrange at a decent price.
No one looking for a mid-tier gaming experience will buy discrete graphics from 2017 onwards.
>>51320288
why does he look like his hair, moustache and glasses have been edited in
>>51320787
>They're going pretty overkill with that HBM2, like shit nigga what are you even going to do with the 512GB/s HBM already gives you?
HBM2 will expand the size of the memory. That's all it does (besides potentially being more stable).
>>51320288
Christ, why is Forbes so fucking homosexual?
>>51320863
pleddit is not much better, and I hate the layout.
>POO IN THE LOO
you too I guess.
>>51320861
>The stupid anti Indian racist
Indian is not a race.
>I NOW want to tell you something, sir
Fuck off back to reddit, sir.
Oh, and that's basically where I stopped reading.
>>51320891
That's pasta m9
>>51320864
>The thing is AMD will eventually cannibalize itself with Zen APUs
It is technically possible for this to accomplish this, but its likely that they won't do it for that very reason.
Raven Ridge will outclass the old 7750 and 7770 hands down, and may even outperform Nvidia's GTX 750ti. I wouldn't bet on it bring an absurdly large IGP right out of the gate. They still need to pull some punches so they can advertise big performance gains with its successor as well.
Intel isn't going to be putting eDRAM equipped Iris Pro chips in their full line of desktop processors, so there will always be a market for a $100~ discrete GPU. Each generation that same $100 is going to buy you more performance exactly as it always has.
>>51320906
It's probably by the same """"""person"""""" who posted it last night.
>>51320766
Yes then releasing only 2-3 chips based off that architecture.
2015: GM200 and GM206
2014: GM204 and GM107
2013: GK110 and GK208
2012: GK104, GK106, and GK107
2011: GF114, GF116, and GF119 ; note the GF116 was internally GF106B
2010: GF100, GF104, GF106, GF108, and GF110 ; note the GF110 was internally GF100B
>>51320861
Indians have copypasta technology, but can he can poo in loo?
>Microsoft lead by currynigger
>goes to shit
>AMD/Radeon lead by currynigger
>goes to shit
I'm sensing a pattern here, and it seems to involve DESIGNATED SHITTING STREETS.
>>51320288
POO IN GPU
O
O
I
N
G
P
U
Who else poo in the gpu here?
>>51320880
Be gone non sjw scum, your kind isn't welcome on the chans.
>>51320874
I used to have a black version of his glasses and they just look like they're photoshopped in every photo of me wearing them
>>51320864
lol, amd isn't releasing any zen apus until q2 2017 at the earliest. we are getting 28nm excavator apus in 2016 with the launch of the am4 chipset
amd will mismanage itself into bankruptcy long before making themselves obsolete
>>51321281
Raven Ridge launches in Q1 2017.
AM4 is a socket, not a chipset.
Shitpost harder.
>>51321281
AMD taped out Zen back in like June and they've already received the first samples. They'll be ramping up production over the next couple months and it'll be available Q2 2016.
>>51321379
we're talking about AMD here, any potential release will be delayed at least 2 quarters and then paper launched
>>51321487
is this bait? amd themselves already stated q4 at the earliest for zen. what we're getting in q2 are excavator based APUs.
>>51321499
>amd themselves already stated q4 at the earliest for zen
Lol, no.
AMD was in danger of pushing back to Q4 because GlobalFoundries is shit, but they moved Zen production to TSMC and are ahead of schedule now.
>>51321499
And they're already losing more and more money and market share to intel.
They'll push zens launch up, 100%.
>>51320288
>Hurr GCN
Nope, AMD are ditching low end cards entirely in favor of APUs, meaning tonga and their new maxwell competitor will be low-mid tier while the 390s get refreshed to use even less power with GDDR5X and get dropped to mid tier, HBM cards will be top-tier with HBM1 cards being priced to compete with the 980.
>>51321564
>because GlobalFoundries is shit, but they moved Zen production to TSMC and are ahead of schedule now.
Impossible because of a contract they have with globalfoundries
Q4 2016 IT IS
>>51321564
AMD already confirmed their CPU's are on 14LPP
>>51321570
>And they're already losing more and more money and market share to intel.
what market share? AMD hasn't sold any notable amount of chips since 2010
i want zen to be a success, will probably buy it as long as it's decent, even if i need to wait until 2017. since AMD is dying it's likely zen will be a good deal no matter what happens
>>51321644
>AMD hasn't sold any notable amount of chips since 2010
Not even counting low margin consoles?
>>51321644
APUs, nigger.
Tons of low-end laptops are still AMD, and they're actually still pretty popular in pre-builts because retards see
>more corez
>more hertz
And go for it.
They're not dead yet, they need something to come out fairly soon though because skylake is fucking rekting their shit.
Hence why when you combine that with samples already being tested, it's very likely we'll see zen around Q2, probably Q3-4 before all the chips get released.
>>51320288
Current consoles can't even 1080p, why is AMD so obsessed with 4k?
>>51321735
They're going to render at 1080p30 in the future and upscale to 4k. I hope something like freesync is added to the next generation though.
>>51321735
Because they're betting the farm on VR. If VR takes off and AMD can obliterate sweep Nvidia under the rug on that front (which means they need to be good at very high resolutions at very high refresh rates), then they'll become the defacto choice for gaming.
>>51320288
They only need high end GPUs because Zen APUs will dominate the lower end market
>>51321764
What price range are "high end gpus"? I could have sworn that the most profitable gpu segment was the midrange 200-350 usd shit.
>>51321760
Just like they bet on cmt.
I don't see vr becoming a standard for a while.
>>51321786
Right those ones they'll only really need to make GPUs of the 380 and above if zen APUs can get at least 2 GBs of HBM on die
>>51321787
How long is a while? The Vive and Rift are going to be releasing consumer versions pretty soon. AMD will have Zen and Arctic Islands out in time for the next round of VR headsets, which will most likely be at 4K.
>>51321801
What if the mobo has hbm next to the socket?
>>51321838
Do you really think VR headsets will change faster than disposable tech like phones?
>>51321838
I think 5 years at least. We can barely get 60fps 4k with sli/cfx.
HBM is definitely a step in the right direction though.
>>51321843
Dunno, might be too slow but that would be another option, although I don't think they have that much HBM to go around, it's much easier if they attach it to the die themselves
>>51321854
What do you mean? 4K already exists for phones, the Sony Xperia Z5 Premium is 4K. It's just not a popular display choice because there is almost no discernible difference between a 5.5" 4K display and a 5.5" 1440p display. But it makes a HUGE difference for VR.
>>51321864
Isn't hbm sensitive to heat or something? Just something I remembered randomly, sounds a bit silly though.
>>51321910
It just sounds like you think improved VR tech will arrive shortly after zen in a year
>>51321858
You're missing the point. You can get high framerate 4K, it just can't be as detailed as you'd normally expect. And that's fine because the VR aspect more than makes up for it. You'll just have a lot of Nintendo-esque games for the first few years.
>>51321912
If it's on die it'll be in constant with the CPU cooler so it should be fine
>>51321925
Yeah, that's what I'm saying. It's not rocket science. People make VR headsets with cardboard and their smartphone. Oculus is going to snap up a ton of 4K displays as soon as they can do so at a reasonable price, and that will be next year.
>>51320288
why does it look like his glasses are photoshopped on his face?
Why are Indian-run companies always seem so empty, passionless, and careless with their products?
>>51320288
>goyim
that is for nvidiots.
>>51321953
And the cpu heat AND gpu heat.
>>51320861
Just calm your poo m8
Try putting it in a loo
>>51321956
Do you honestly think vr will take off?
>>51320923
OP BTFO
>>51321735
so they can that ps5 contract.
>>51322035
But those will be much more power efficient than what's available now, especially since it'll be 14nm, look at the A8 7600 and A10 7800, they have pretty solid performance per watt
>>51322056
Yes, I do. Especially once it's properly marketed to normalfags so they can watch their football from the perspective of the people actually playing football.
That's frustrating, I was looking at low end gpus from amd and they're all old gcn rebrands, a 370 is 1.0. At least use 1.2 or something. Maybe nvidia's pascal line will have an update on the lower end.
>>51322131
amd would've been a lot better off if they dropped the rebrands by 2 tiers instead of 1 imo. should've put a fiji chip with 3000 processors in the $350 price bracket and put the 4gb 290x down to the $200 price bracket.
>>51321760
>If VR takes off and AMD can obliterate sweep Nvidia under the rug on that front
Doubt it considering Nvidia is adding hardware to work exclusively on the features needed for VR.
>>51322131
Expect the 980 to be rebadged as the 1070, and the 970 to be rebadged as the 1060.
>>51322167
Knowing nvidia, the price won't get rebadged...
>>51322066
You think they might be able to release motherboards with 8gb hbm for APUs?
That can dedicate 2gb for the gpu and 6gb for the system?
>>51320288
Makes sense, indfag here and a few of the latest names for their products have been named after rivers in india like kaveri and godavari.
Be on the lookout for ganga soon /g/
>>51322219
Would be a pretty expensive board, and a big waste if it failed.
>>51322250
I'd like that in a laptop desu.
>>51322219
>>51320288
>We need a new product
DON'T POO IN THE LOO
MOLD THAT SHIT INTO A GPU, NOW!!!
>>51321735
Probably just to not get steamrolled by ARM, too?
>>51322042
rofl
>>51320288
>AMD will start making graphics poo units
RIP.
time to buy scamvidia cards
>>51320864
>They NEED to be competitive with Nvidia in the top-tier GPU market
No they don't. Only an extremely small amount of people buy top tier gpus. Neither companies make any much money off those people. Only the dumbasses that buy them actually care.
Why are nVidia shills so MAD?
>>51320861
Nice copy pasta kek
>>51320982
G O O G L E
O
G
L
E
all this fud itt
>>51320288
POO IN LOO
AMD IS FINISHED & BANKRUPT
>>51320982
The question is - why the fuck does the US of A even allow it?
>>51320982
Salty Nutella is excusable.
>>51320787
astronomical amount of draw calls in dx12/vulcan will use bandwidth properly as i understand it
>>51324381
indian universities are literally shitting out millions of coders each year -> pooinloos are cheaper than domestic workers = more profit
https://www.youtube.com/watch?v=AfemCOLrKB0
POO IN LOO
O
O
IN
LOO
O
O
https://www.youtube.com/watch?v=_peUxE_BKcU
>>51321843
HBM cannot signal over long distances. Eye pattern would degrade into nothing.
It can only be placed on package. Even HMC encountered the same problem, and they had to create a special variant with a unique interconnect fabric to address it.
>>>51321912
DRAM isn't inherently sensitive to heat, but HBM stacks themselves run fairly hot even under normal conditions. In current implementation a die stack will operate over 85c. That doesn't leave them a ton of headroom, but you also have to remember that it is a tiny chip only consuming about 3w~
The heat is due to its small size and density, not high power consumption.
>>51322035
This isn't a problem.
>>51322219
AMD's upcoming Raven Ridge is a DDR4 platform. Its possible that it will have on package HBM used for the IGP, and for compute tasks, but it will still use normal DIMMs for system memory.
The HSA spec has exemptions that allow it to break itself and fully implement context switching for classes of workloads, and it does this on the fly simply by looking at pointers.
I'm waiting for poo2 stacked poo with high poo content, ensuring pooing is even more efficient
>>51321066
pf o.k. kid
>>51322913
Marvelous hair. Am jelly.
>>51321994
it is. look at his right ear, the ear piece of the glasses are phasing through his ear. but why would someone do that? because glasses are a sign of intelligence in india
>>51320874
>that fuzzy as fuck outline of his hair
god damn
>>51321644
> AMD hasn't sold any notable amount of chips since 2010
Why do you dumb little kids think you can just lie without being called out?
>>51321564
> but they moved Zen production to TSMC and are ahead of schedule now.
Same question as above, you tech illiterate retard.
AMD's GloFo's dedicated customer. They are contractually obligated to meet a wafer supply agreement, so AMD has to produce the bulk of all their CPU logic at GloFo. Porting a design is a tremendous undertaking especially since 14nm and 16nm have wildly different design flows.
The Zen based Summit Ridge CPUs and Raven Ridge APUs are GloFo chips. They are not fabbed by TSMC.
>>51321636
They didn't confirm what specific node they would be using, only that GloFo is making some 14nm FinFET chips for them.
14nm LPP and all its variants passed qualification at the end of Q2 this year, and entered ramp up phase.
Its likely that the Zen based Summit Ridge CPUs would be using 14nm LPP LVT due to the high frequencies it can target.
>>51325728
and pic related
Fiscal year 2011 AMD sold well over 60 million CPUs all together.
They did nearly that in the first 3 quarters of 2012 alone.
>>51320288
Opinion: Into the loo
AMD is pretty much dead.
Look at their market value.
They are dirt cheap to buy up, but no company with a right leadership would buy this dead horse.
>>51326047
They'll never die. Intel can't have a monopoly
>>51326047
>he actually thinks the stock price is the value of a company
This is an 18+ site.
>>51326085
Yes it is you idiot. It's a public company.
If you buy 50%+1 stock, you pretty much own the company.
>>51326092
This is an 18+ site.
The stock value is not even remotely representative of the value of a company. AMD's IP alone would be worth billions of dollars, their operating business, facilities they own, and etc would all be factored in if a party were to acquire them.
If you bought 50% of their stock you would have a controlling share, and a seat on the board. You do not own 50% of the company, you retarded underage kid.
>>51320882
Relax Pajeet. You must relax your bowels when you use the loo. I know it seems much easier to just shit in the streets but you must control yourself Pajeet. This is the 21st century. Remember Pajeet, POO IN LOO!
>>51320982
Microshit went to shit under Ballmer, fuck you talking about
AMD lost, Intel and Nvidia won.
AMD is playing catch up game, but you can't play this forever.
AMD is in big debts, no profit or whatsoever for years.
>>51320288
>only Two New GPUs planned for 2016
Well, it's not like I planned to "upgrade" to some low end shit, so as long as the top of the line is new shit I'm happy.
>>51320791
>>51320882
Krishna pls go
>>51320766
>700 series was all rebrands except the 780/Titan
>b.. but it's bad when AMD does it
>>51321735
They're trying to force the
>we added more RAM so it's better than Nvidia meme
>>51327369
>complaining about rebrands in the 28 nm era
Blame the shitty fabs for taking nearly a decade to get to 14nm.
There's only so much architecture can do without a die shrink.
>>51320288
>Raja Koduri
Oh, so that's the reason amd is so shitty
>>51327426
How to rebrand all your GPUs and have your new architecture lose to older Nvidia cards.
>>51327420
>complaining about rebrands in the 28 nm era
Sounds like Nvidia marketeers to me
>>51320791
Poo in loo.
Fight me faggot.
>>51327465
Sure thing bro, tell that to AMD forcing their shitty HBM onto new fury x's and selling it for $650 with little to no performance advantages and rebranding it's entire line top to bottom with a shitty factory OC.
>inb4 Direct X 12 async compute performance
>Implying by the time any actual games worth a shit using DX 12 come out there won't be 14nm cards that blow everything right now out of the water for cheaper.
>>51327495
Why are you so upset?
Is nVidia not paying you enough?
Or do you only get paid if you get proof of conversion?
>>51327511
>u mad bro?
get back to reddit faget.
>>51327426
he got managing position 5 or 6 months ago, if anything you should blame previous execs.
>>51327495
Those shitty factory OC rebrands are still beating shiny new Maxwell cards even in DX11 benchmarks.
Either Hawaii was great or Maxwell really fucked up
>>51327158
>AMD is playing catch up game, but you can't play this forever.
Ironic when you consider they pretty much pushed for all the most recent technological innovations.
>>51327566
no, they beat reference maxwell cards. even factory oc'ed 970s can beat the amd competition, a proper OC will see a 970 outperforming stock 980s and 390x's.
>>51327569
what innovation? selling a rebrand for a higher price than it sold for the year before?
>>51327569
marketing> actual hardware
I seriously hope they will push on content part next year, alien isolation and tomb raider ran too well for their own good.
>>51327582
Except all recent benches, review and users say otherwise. If anything it's equal at 1080p, with tiny lead at 1440p.
>>51327566
The answer is both.
Well, maxwell isn't 'fucked up' - nVidia made a conscious choice to sacrifice compute capability so they could use more floorspace for geometry and tessellation SIP blocks.
28nm wasn't supposed to last this long, so drastic measures had to be done.
AMD chose to shrink designs and just try for more shaders - that uses lots of power and doesn't always work out well in existing codepaths, but will work better for DX12
nVidia chose to optimise for right now, maxwell won't age well, but it's the best-in-class for DX11 at 28nm.
>>51327605
DirectX12, Vulkan, Adaptive sync.
inb4 "but AMD had nothing to do with DX12 !"
because they completely led the way with Mantle of what developers wanted, and DX12 was pretty much inexistant before Mantle was revealed.
And it's even more obvious when you see AMD old cards on new tech, they perform much better than same age nVidia cards. Building for the future rather than optimizing for current benchmark performance.
>>51327631
those reviews are comparing amd's cards to reference cards tho
>>51327635
This,
The architecture is at the point of literally sacrificing one thing for another because of lack of space on 28nm.
It's no use having more compute performance when your card is going to get shit on at every level with 14nm coming up.
>>51320791
>ugh
Back to >>lebbit, pajeet you turd
>>51327654
are you retarded or are you watching "newegg" videos all the time? No decent reviewer compares OC card to non OC card.
Reminder that GeForce 900 series is only 3 chips, 1 released in 2014, 2 released in 2015.
2 distinct chips per year is perfectly normal.
>>51327686
every single reviewer does you idiot, maxwell chips destroy amd once overclocked
Please hold me /g/
I bought AMD stocks for 2 Shekels, 1500 shares.
I am fucked, it's dropped down to 1,8€ already.
>>51327647
There's a reason why those APIs aren't being used by any of the latest games.
It's because it's literally going to take throwing away years and years of work and starting from scratch.
Which literally means you won't see any DX 12 games worth a damn until 2017-18.
Even now some games still run better in DX10 mode because lol console ports.
>>51327707
hope and pray that AMD survives 2016 and doesn't sell off radeon
>>51327695
>posts firestrike
Okay, thanks for proving me right.
>>51327707
Nigger, the first rule of stocks is to buy when it's low; not when it spiked up for a bit.
>>51327707
Short it to hedge your bets. And buy Nvidia with the rest.
>>51327695
>synthetic benchmarks
>cherrypicked press samples
>mfw this thread is literally the India experience
>wading through literal shit in order to get to the content
Kek, no wonder you guys love this meme so much. 4chan IS like a literal curry site.
>>51327733
I thought it was already low, and it would atleast stabilize around the 2-3€ line.
I mean, AMD is totally underrated in comparison to the likes of Nvidia, where 1 share costs 29€.
>>51327721
>APIs aren't being used
They are, on a console, xbone pretty much uses dx12 analog. You say it like dx12 is something revolutionary - it is the same fucking thing under different sauce. It actually, in theory, should lower effort and unify workflow for consoles and pc.
Two chinese dudes made a dx12 port for age of wushu in two weeks, imagine what actual team can do.
And a reminder Deux ex comes out february with dx12 support, whole year before next gen of gpus.
>>51327707
Losses aren't losses until you realise them. Prices will go back up. Nobody actually believes AMD is only worth 1.6 billion USD.
>>51327738
very bad timing to buy NVDA
http://www.barrons.com/articles/amd-could-be-profitable-in-24-months-1447419030
>>51327782
It had a spike of about 25% in a few weeks.
>>51327756
>>51327729
if you're actually trying to deny that maxwell oc's better than amd's rebrands i have a bridge to sell you
>>51327799
Cherrypicked press samples do not represent the average cards you can buy in stores. If they did, then all cards would be sold with higher stock clocks.
>>51327799
>comparing against 900mhz 290X
Nice shill graph.
>>51327787
>They are on console
Literally no game on Xbone currently even uses DX 12 you stupid fuck.
The only game that's going to use it is a shitty MMO called Fable Legends.
>They made a port for Age of Wushu
>shitty crap mmo
>anyone giving a shit
Top kek, try convincing Bethesda, Ubisoft, EA, Activision, Square Enix and Rockstar to port their deeply entrenched console engines into a completely new API and see how long that takes...
>>51327849
the majority of meme cards oc far higher than the OC's shown in these benchmarks you shill
they got 1350mhz core, the majority of 970s can do 1500-1550mhz with no voltage change.
>>51327799
You actually might be retarded.
>>51327855
the 290 is what competes against the 970, they're both in the $250 price range these days.
>>51327879
Except every 290 has 1ghz+ clocks, they are comparing reference 290, which hasn't been available in over 2 years.
>>51327869
>>51327879
>>51327895
the reference 290 is the recent $200 290, not sure where you got the impression they can't be found. they're actually the only 290 still available to purchase for a reasonable price.
>>51327879
idiot
>>51327868
>no game on Xbone currently even uses DX 12
are you trying to tell me that APi they use has no relation to dx12? Do you only speak in marketing terms and if it doesn't have a shiny label it doesn't exist? Do you seriously think MS made dx12 from scratch?
90% of it was made as internal API for xbone, they ported it and trying to sell it as a new shiny thing.
>anyone giving a shit
It's about what is possible. ANd big companies have dev cycles, my bet tomb raider will be dx12 in spring.
All AAA will go dx12 next autumn - simple reason it's cheaper to make.
>>51327929
thanks for proving my point that the reference 290 is indeed the only card competing with b-stock 970s, shill
>>51327929
I wouldn't say he's an idiot, but that does show a cheaper 290 vs 970 at only 60mhz different clock rates.
>>51327916
I just had a look on Newegg, I'll be damned, you can actually still buy them in USA - They've been gone for ~2 years now in APAC and Yuroland.
>>51327970
290x is still around, but they seem to perform slightly worse than 390 for the same price, they shouldn't though
>>51327956
>b-stock 970s
Wish this meme would just die. They're literally never in stock and the only reason it exists on EVGA store is so it can marketed by shills like you. Go ahead and check. They won't be in stock.
>>51327956
Stop trying. Even the people you are trying to trick here know you can't buy a 970 for $240 or even $300.
>>51328001
it is right now
http://www.evga.com/Products/ProductList.aspx?type=8
>>51328020
No it's not.
>AMD
in to the loo it goes
>>51328020
your own link proves you wrong, shill
>>51328037
https://en.wikipedia.org/wiki/K%C3%BCbler-Ross_model#Stages
you're in denial son
>>51328049
You're fucking retarded.
>>51328061
he gets $0.01 every time he bumps this thread
>>51328061
yeah, i'm retarded for pointing out that the reference 290 is 30% slower than a factory oc'ed 970
AMD should make GPUs that actually perform up to the standard of the competition if they want people to buy them
Dear retards,
970's are essentially b-stock 980's. Nvidia doesn't specifically lock out 980's so that they can sell them as 970's
970's are literally the 980's slightly gimped sibling.
There is no b-stock 970. the 970 is already b-stock to the 980. As for the factory overclocking. Maxwell OC's so well that any factory OC is pretty irrelevant considering how much headroom you have when the card runs at 70c on load lel.
>>51326092
>If you buy 50%+1 stock, you pretty much own the company.
>>51320288
>indian
dropped
