/w4n/ - Daily Wait for Navi Thread

>>61962722
fuck Navi, I'm waiting for next gen

>>61962722
JUST

>>61962722

I'm waiting for Navi as well. I've already bought a 1200 watt PSU, so I'm set.

>>61963174
I'm waiting for Next Gen, my 2000W PSU is ready

>>61963118
As the owner of a Fury X, I think I'm going to do the same... that is, if I ever do bother upgrading.
I can't even trust Navi at this point, and even if it does turn out great, the generation after that should only mean improvements upon that base.

File: 1502952324582.png (70KB, 852x944px) Image search: [Google]

70KB, 852x944px

I don't know anything about graphic cards

Will it have a massive die as well?

>>61963222
navi is going to be on the 7nm node that in cpus is targeting 5ghz as a normal operating speed, the process is either going to be really cool or the size makes it really efficient, its likely that this will bring at the very least vega lq down to polaris msrp on air.

its hard to justify anything right now

low end cards cost a fuck ton
high end cards aren't fast enough yet at the same time massively overkill... and with miners being cunts, any hope that a reasonably priced gpu for performance goes out the fucking window.

>>61963251
No, the point of Navi is making use of infinity fabric to make two small dies work as a single one, effectively increasing yields, decreasing power and cost.

What does the AMD GPU roadmap look like if you put in it's actual performance and efficiency for Polaris and Vega?
I'm guessing Navi looks impossibly far ahead

>>61963268
[citation needed]
We literally don't know anything about navi.

>>61963338
"Scalability" means exactly one thing in AMD-speak.

>>61963338
Guys it's going to be the best, AMD always makes the best and never lies to their customers. JUST WAIT (TM)

File: 1503019340135.jpg (131KB, 1080x1920px) Image search: [Google]

131KB, 1080x1920px

Buy vega now

File: big guy.jpg (50KB, 600x481px) Image search: [Google]

50KB, 600x481px

I give you the permission to wait.

>>61963834
Your beeg boi...

>>61963851
for you

bump for navi

>>61963522
Good luck, it's sold out everywhere already.

>>61962722
Why would I wait for navi when volta is 3 months away? Amd has proven they don't care about the high end market so 1180 and 1180 ti will own anything navi can produce.

>>61966434
>Volta
>3 months
lel

>>61966434
>They don't care about the high end market

They do care, but they just cant compete.
They should just re-stock the 570-580 to reset the street price, that would literally print money

>>61966621
Because right now it doesn't print money?

>>61966683
Surely their RX production was slowed/delayed because of Vega.

>>61966725
Probably not so much if they have very few cards. Regardless, both polaris and Vega are constantly out of stock due to miners. No surprise nvidia wants to jump on the train as well.

>>61966759
Not only Miners, but AI researchers and such are eating Vega up and buying in bulk.

The raw compute numbers are hard to beat.

>>61962722
when is the rx 490 or 590 coming out?

>>61966788
Why doesn't AMD just mine themselves?

>>61966788
True. Vega is still pretty good in segments where power consumption doesn't matter.

>>61966832
Well, they have a LOT of those petaflop P47s i bet that when they have no work they'll just mine away

>>61966434
>Volta
>3 months away
Leather jacket-man has already said he's in no rush to release Volta plus GDDR6 doesn't hit production to 2018 anyway.

>>61965873
In India maybe

>>61966813
Never. Like is aid earlyer. Amd gave up on the high end market after the 390.

Vega will get no high end card. Just the new "580" that is vega 64. The slap on a water cooler and oc it. But it is no 2/390 or furry x.

>>61967144
2018 is 3 months away. Volta is q1 2018 so 3-6 months.

>>61962722
IF YOU DON'T BUY VEGA THEY WONT HAVE THE MONI TO MAKE NAVI

prove me wrong

>>61967944
>>61966813
Isn't vega 64 the high end card?

AMD has a severely fucked naming scheme right now

>>61967957
Anon it takes 1 to 2 quarters from production to market. You won't see consumer volta till at least midyear

>>61967997
If 64 is the high end why does it compete with nvidias mid range card the 1080? If amd was into high end they would compete with the 1080ti.

>>61968313
>1080
>Midrange

My RX 480 8GB runs most games fine on high at 1440P using my Freesync 40-144Hz monitor. I can even run a lot of games on ultra.

I think I will skip this generation. If I had not been forced to upgrade my monitor two weeks back when my old one died I would have waited and gone with Gsync + 1080. I wish it was easy to send the monitor back for a refund and start over.

>>61963338
well if the slide in OP is anything to go by, Navi is at the very least a die-shrink.

Which uarch they choose to base it on though is anyone's guess.

former AMDfag here. I'm never buying another AMD product ever again. I just hope nvidia doesn't screw me over as hard as AMD.

>>61968313
There are only two GPUs that are more powerful than the GTX 1080.

>>61968797
>former

>>61968797
Terrible bait

>>61968040
and why can production on the gpu die not start before gddr6 is readily available?

>>61962722
... isn't this the same road map for Zen, but with names changed?

>>61969234
1080ti and?

>>61963268
Yeah massive latency on CPUs
Now massive latency on GPUs too
What could go wrong except massive bottlenecks that are already around

Navi will definitely blow Volta away.

How can NVIDIA ever compete?

>>61970411
Next gen after Navi is called Zimbabwe due to mines. It's rumored that on next gen amd will drop all video connectors to focus on their target segment: cryptocucks

>Next Gen
will we need access to a nuclear reactor to power it?

How many Intel/Nvidia shills are in these threads? I always see a Just wait(TM) and all that, but I don't know how many there are.

>>61970472
>making a joke means you hate AMD
I have an amd processor and an amd video card in the computer im shitposting from right now. You need to learn to take a joke.

>>61970488
I know there at people who joke, I'm just looking for genuine shills. They probably won't respond anyway.

>>61970432
I've ever used ATI/AMD cards for my gaming rig but I switched to NVIDIA in 2015.
There is nothing which can compete with the ti cards.

But I have to point out that the average price for NVIDIA ti cards raised by about 100$ if you compare the maxwell and the Pascal generations.

After all competitive AMD cards are important for keeping NVIDIA from becoming an Israel company like Intel.

>>61970472

>IT IS OK TO BULLY INTEL/NVIDIA.

>IT IS NOT OK TO BULLY AMD. REEEEEEEEEEEEEEEEEEEEE

Grow a pair.

>>61970557
>After all competitive AMD cards are important for keeping NVIDIA from becoming an Israel company like Intel.

Too bad AMD just shat the bed with Vega.
On Steam survey since Feb 2016 till today they lost 6% of market share (from 26% to 20%).
Vega fail will just keep this decline, and next year Volta will only increase it.
When Navi comes in 2019 they will have like 10% marketshare and then developer support will dwindle.

>>61962722
Next to Navi will be in PS5
Amd don't fuck this

>>61970390
>Anti-AMD shills are this retarded
Even NVidia knows MCM GPUs are the next step, retard.

http://research.nvidia.com/publication/2017-06_MCM-GPU%3A-Multi-Chip-Module-GPUs

>>61967203
>shills are this terminally butthurt Vega is selling out

>>61970111
Titan X Pascal.

>>61973273
Miners sure do appreciate those gaming specific hardware features.

Faggot.

>>61973701
Titan X (pascal) and the Titan Xp

>>61969234
>>61970111
>>61973701
>>61974517
1080 Ti, Titan X (Pascal), Titan Xp, Quadro P6000

File: Quadro_GP100_Top.jpg (1MB, 4000x2746px) Image search: [Google]

1MB, 4000x2746px

>>61974578
1080 Ti, Titan X (Pascal), Titan Xp, GP100? (pic related), Quadro P6000, GV100

File: 1503105784638.png (484KB, 1874x620px) Image search: [Google]

484KB, 1874x620px

>>61974775
You forgot the rest?

>>61974984
I'm waiting for someone to respond to you with an even bigger picture with more shit inside

>>61974984
>vega from boot to 100%
https://www.youtube.com/watch?v=UxQdS0pbpKo

File: 1503105784638v2.png (568KB, 1874x620px) Image search: [Google]

568KB, 1874x620px

>>61975974

>>61973240
I might as well keep my 390x until mcm comes out as current monolithic gpus just don't have the raw power for 4k

>>61974578
And 980 Ti.

>>61973240
>glue a bunch of 1050ti
think of the yields.

>>61962722
It probably will be fine.
Not AWESOME, not even close to the impact of a ryzen, just well fine.
Also i hope the low end Vega don't suck as much as the high end, because a Ryzen+vega SoC would be excellent if the vega part don't fuck up.

>>61974705
They need review the GP100 desu

>>61962722
Unless Vega 56 turns out to be a decent performer at 4k with some OCing I'll be waiting for Navi and Zen 2.

>>61974984
Replace generator with Vega

>>61977648
w-wait till raja turn on all the features with drivers update.
just wait.

>>61977745
I mean, I don't doubt DSBR and primitive shaders and all that shit will help since it saves on memory bandwidth and that seems to be Vega's problem, but I'm not expecting a huge performance increase.

>>61977815
DSBR is working, Vega is not bandwith starved.
Primitive culling through prim shader is not working.
Frontend is choking bretty fucking hard, given the ALU count.

>>61977831
>DSBR is working
>Primitive culling through prim shader is not working.
>Vega is not bandwith starved.

front end choking the back end somehow not bandwidth starved

>>61977923
idort here

why is vega 64 so slow then?

t. 390x owner

>>61977937
Just like Fiji, current Vega has too much ALUs given the triangle output (that's without prim shader).
>>61977923
OCing both core and HBM2 does basically nothing.

>>61977973
So is it bottlnecked in software?

It cant pump out enough poly's to use the ALU's fully?

>>61977973
for 64, it gains a ton of free performance on 56 with OCed HBM

>>61977996
pretty much, that's why prim shaders are a feature but they still didn't make it work driver is basically barely working alpha
the hope they will make it work by AIBs time, but I wouldn't hold my breath probably due to next crimson iteration october

>>61977831
I don't understand. How does OCing the memory on Vega 56 put it so close to Vega 64 if memory bandwidth isn't the issue?

>>61978003
No shit, 56 is segmented by memory bandwith.
>>61977996
>So is it bottlnecked in software?
Sort of. It's the drivers.
>It cant pump out enough poly's to use the ALU's fully?
Yes. As evidenced by workstation benches - Vega does bretty fucking nice in geometry-bound shit with prim shader culling working.
>>61978025
56 has 800mhz HBM while still having whopping 3.5k ALUs.
56 to 64 is 1080ti to Titan Xp or 1700 to 1800x-tier upgrade.

>>61978051
but what if, what if they will manage to feed ALL THE ALUS later on, which might happen because vega has hardware for it unlike fiji
how big of a gap 500 ALUS would be?

>>61978065
>how big of a gap 500 ALUS would be?
A bit bigger than 1080ti to Titan Xp gap.
It's not much.
You just need to OC 56's memory stacks.

>>61978070
so we come back to a conclusion that 1600(x) and 1700 is the only cpus worth buying and 56 is the only gpu
I don't know, their marketing team seems out of touch, what's the point of wasting yields if nobody going to buy it?

>>61963522
Stop lying about msrp

>>61978097
>I don't know, their marketing team seems out of touch, what's the point of wasting yields if nobody going to buy it?
They are not wasting yields, everything remotely good goes for Pro line or MI25.

>>61978022
thanks, i might as well just go ahead with my ryzen build and wait

>>61967957
>>61968040

Not to mention that the Volta 2080 Ti will probably be launched in mid-2019, just in time to do battle with Navi. So we're still a little under 2 years away from seeing what the next level of gpu-perofmrance will be like, whether that comes from Nvidia or AMD.

>>61978984
Cutdown GV102 is late Q4 2018 at worst.

>>61963163
WAIT™

>>61962722
>tfw no money to buy a new rig right now and components are too expensive anyway
I'll happily wait and see what happens desu. I'm sure Vega will improve with time due to better drivers though I'm not sure my next card will even be a Vega or Pascal.

>>61962722
I've lost all faith in AMD.

>Fury launch was a disaster
>R9 3XX launch was a pathetic rebadge fest
>Polaris launch was miserable
>Vega launch is a cataclysmic disaster the likes of which nave never been seen before

If this trend continues Navi GPUs will turn into tiny black holes that suck up your PC before blowing up and obliterating your house. At this point, it might be a matter of public safety to keep AMD from making any more GPUs.

>>61979134
>Fury launch was a disaster
What?
>R9 3XX launch was a pathetic rebadge fest
Okay.
>Polaris launch was miserable
What?
>Vega launch is a cataclysmic disaster the likes of which nave never been seen before
What?

>>61979134
>Fury launch was a disaster
Disaster? You sure haven't seen Vega's launch thus far, have you?

>>61979204

Not him but Fury X launch was not perfect. The card was so overhyped, the drivers was not mature so the card was slower than the 980ti at the launch, the AIO and the power draw triggered a lot of people, the card was the the ''overclokers dream'' nor the Titan Maxwell killer and the Nano was ridiculously overpriced at the launch.

Also honestly, Vega is far worst.

>>61979312
was not the overclokers dream*

>>61979312
>Also honestly, Vega is far worst.
Why? The only bad part is that RTG failed to make primshader culling work for launch.
That and supply issues, but that happens every bibeocard launch, including NVDA ones.

>>61979222
>No stock
>No overclocking capabilities
>Massive power consumption
>Bad performance
>Broken pumps on their AIO liquid coolers

How is that not a disaster? The only thing at which AMD is good is failing to deliver on their promises. They've been doing that for half a decade now.

It's no wonder Nvidia's market share is getting bigger and bigger.

Does NVIDIA scale their geometry perf with the number of Polymorph engines?

>>61979337
The chip is bigger and have more transitors than Titan Xp GP102 but match only with the GTX 1080 with +50% more power draw.

>>61979641
Right now.
Also it's only 15mm^2 bigger than GP102.

>>61979645
Yeah but do you know the size of the GTX 1080 GP104?......Vega is a massive fail.

>>61979664
It competes with full GP102 when it works.
It doesn't work in gayms right now.
So you can postpone your screeching for later.

>>61979394
>Bad performance
it was faster than 980, and 980ti didn't exist

>>61979684
Get a FE, people bought the RX for gayming. They don't give a fuck about cumpute or pro aps.

>>61979694
>980ti didn't exist

Does NVIDIA scale their geometry perf with the number of Polymorph engines?
>>61979704
Why would i buy FE if it don't need 16 gigs of memory?
And it'll perform well in gaymen in a month or two (i don't really know when the next major Crimson update lands).

>>61979694
>980ti didn't exist

>>61979694
> 980ti didn't exist
It literally did.
Fury X was pretty bad.
It was a bottleneckfest.

>>61979694
>980ti didn't exist
The 980ti was here before Fury X

>>61979694
>980ti didn't exist
the abosolute state of amdrones

>>61979694
>980ti
>launch 06/01/2015 $649
>fury x
>launch 06/24/15 $649
?

>>61979684
>Vega will compete with 1080 Ti """""when it works"""""

It doesn't even beat the 1080

The fuck are you on?
Muh magical drivers
I mean Vega 56 AIB should be okay, but holy fuck you are retarded

>>61979793
>a card that competes with P6000 with not compete with cutdown P6000
?
>The fuck are you on?
Enabling uarch-defining features is important.

>>61979845
>JUST WAIT
how much amd pays you

>>61979850
Just like everyone just waited™ for Maxlel to finally beat fucking GCN?

>>61962722
if they fail with navi (navi being the first card amd can actually throw money at it because they actually HAVE money now to throw) they better just sell rtg and be done with it

>>61963338
we know navi will be mcm since 2014....

>>61979948
They'll have no iGPUs if they sell RTG.
>>61979955
We don't.
Just "scalability" in AMD-speak means roughly one thing.

>>61979958
yes we do there is a shit ton of rumors from back then talking about a card of amd being an mcm

>>61979981
You mean patents for chiplets and interposers?
That's not only for GPUs though.

>>61979994
no i mean an interview of raja
https://www.youtube.com/watch?v=24rFeHo1S-o

>>61980014
Cutress should interview him about making of Vega.

>>61980014
>interview of raja
He lost all credibility.

>>61980084
>he
Who?

>>61980082
>>61980084
oh yeah vega is a flop on games a card that is literally fiji on a die shrink makes raja lost his credibility..
lol either you are a shill or you knowledge is dangerously lacking

>>61980106
It's not a foot in gayming. It's just late and the product is basically unifinished. Hence why the launch is staggered as hell.

>>61980106
When you lie and spin shit, you lose credibility. It's that simple.

>>61980124
Where did he ever lie?

>>61980121
it doesnt have gaming drivers thats all
in pro segment the card shits on nvidia
>>61980124
he lied? where?

>>61980136
Well it has right now.
These are alpha-quality though. Which is pretty bad for a hardlaunched product.

>>61980127
>>61980136
most recent is how RX will be much better than FE in gaming

>>61980168
It will be much better than FE at gayming.
Whenever it'll start working though.
Anyway FE will also be good at gaymen once they merge the fucking driver branches.
It's a fucking shitfest right now.

>>61980143
indeed but lets get real here i think that amd decide to throw all the money on the pro segment because back then they simply didnt had the money to have 2 lines developing for it

i bought 2 64 sapphire's ones and ok i took the risk knowning that this is amd in 4-6 months the card will be awesome but right now not even the power managment is working (i see power flux even on idle that spikes +30+50watts for nothing)
nothing really works its like the card when it senses a 3d program just activates everything and get done with it
as far as i have gone (along with people that told me to test shit on beyond3d the card is basicly this a fiji on a die shrink with bigger pipeline and almost everything of the new tech is disabled or not working to their full capacity especially the culling and triagle buffers )

>>61980188
The only thing that is not working properly RN is culling in shaders.
That and DSBR needs tweaking.

>>61980207
nothing is working hbcc /power profiles/culling-triangles/simd pipelines profiles nothing really
which considering its running on fiji drivers im not suprised..

>>61980222
It's not exactly Fiji drivers.
But yes, consumer Vega is basically not ready for launch.

>>61980228
it is it buffs 8 triangles only while amd on the whitepaper states vega can buff 17 somehow the card gets fully utilised even if you are playing the simplest 3d game from 2002 but its underperfoming quite a lot in the same time

>>61980243
>only 8 tris
At least PDA is working.

I have my nuclear reactor ready.

>>61967957
>august is 3 months away from january
Burger education, I presume?

>http://www.gamersnexus.net/industry/3025-amd-statement-on-vega-pricing-retailers-fault

We're innocent, goyim. Now go forth my drones and defend ayyymd because we're the good guys.

>>61980476
I guess Polaris cards selling over MSRP is also all evulz ayyemdee.

>>61980492
Did you even read the article. It's completely different.

>unironically have to wait for an upgrade to my 750 Ti
well, I guess I'm fine with medium-high settings at 720p

>>61980508
It's literally the same.
I don't even know why the fuck is it even selling, given the lackluster launch day performance.

Oh wew NGG path is truly disabled for both wangblows and open-source Radeon drivers.
Basically it's not ready yet.

This launch is both untintentionally hilarious and retarded.
Mostly pants-on-head retarded.

>>61980476
gn being gn lol its amd fault that the cards sells for 600 -800 due to miners and retailers wants to make profit

one more proof as to why gn is a fucking retard when it comes to amd

File: 1495147543587.png (6KB, 472x508px) Image search: [Google]

6KB, 472x508px

Why does AMD only release two graphics cards while Nvidia releases an entire series?

>>61981141
Vega10 works as a nice testbench for the rest of the Vega products (being Vega11/12/20).
Right now they are busy implementing NGG into drivers.

Best bang for buck gtx 1080 SKUs are going out of stock. O I am laffin.
Buying an 11gbps aorus 1080. Fuck freesync

>>61963215
nah buying a 3000 watt modded psu for 5 generations to come

>>61981097
>Some review site call amd bullshit
>It's not amd fault

At least on Nvidia cards drop over time when tech is more accessible and not grown price over time

Also fuck cryptocucks

>>61981475
amd sets the msrp
retailers just skyrockets the price because of cryptominers

ITS AMD FAULT

>>61980228
How much more time it needs to get ready? 8 years?
Fury on paper was hell of a card, in practice it's a mess. And not only done it again, it double down on Vega

>>61981489
Vega's price skyrocketed not due to cryptocucks.
The supply is not there.
And the demand is pretty high.
>>61981496
>Fury on paper was hell of a card
It was a gigantic bottleneckfest even on paper.

>>61981516
SO ITS AMD FAULT THAT THEY CANT KEEP UP WITH THE SUPPLY

GOT IT

REDDIT SPACE

File: images (31).jpg (16KB, 488x301px) Image search: [Google]

16KB, 488x301px

>>61980082
You mean the potato of poo tech that shill for amd?

>>61981536
>anandtech now shills for AMD
BWAHAHAHAHAHA
WHAT'S NEXT, THG IS ALSO SHILLING FOR AMD YOU FUCKRETARDED BR APE?

>>61979040
Stock, yields and prices won't thought it's the fury x all over again

Gonna sit this gen out and wait for mcm Navi and mcm volta around 2019.

Sadly my 390x is a good card but I am sick of amds bullshit

>>61978051
>Yes. As evidenced by workstation benches - Vega does bretty fucking nice in geometry-bound shit with prim shader culling working.
So you believe that both primitive shaders and primitive culling are working in pro drivers, and that's why Vega10 pumells the shit out of GP104 in pro rendering applications already?

>>61981670
It only shits on full GP102 in Creo.
They're dead even everywhere else.
GP102 clocks higher on average though (sasuga GPU Boost 3.0).

Hell, they would'nt demo WX9100 so fucking actively unless the shit was working properly.

AMD really did fuck up here and fine wine just wait won't apply because in 2 years or less mcm gpus will be out on 7nm and will absolutely destroy monolithic gpus.

>>61981704
>just wait 2 years to buy a new gpu :^)
When will the wait faggotry end

>>61981679
Yes, I know, my point was asking whether you believe that primitive shaders are implemented in the pro drivers? That would actually explain a lot.

>>61981729
They are.
No way Polaris-like uarch (no prim shaders) could even HOPE to touch full GP102.
Like, no chance.

>>61981740
That makes sense. By the way, I want to thank you for baiting me into actually learning something about GPUs in all the w4v threads.

raja fucked us over

he needs to be fired

>>61981797
>us
Who?
Cryptokikes get out.

>>61981727
Hell I waited from 2013-2016 to get a 390x

Had a 290 at launch and the driver's back then where hilariously bad to the point where a 770 4gb oc was just as fast and 1/2 the power and heat output so I downgraded.

When Nvidia launched the 1080 it was too expensive here so I got a 390x last year and it's been going great ever since.

Ironically it's CPU bound by my old i5 4690 once I get a ryzen it will be even better.

AMD rtg really did fuck up hard, I miss the old ati that could actually stick it to Nvidia and drive down prices for the average gamer but ever since 2013 it's been a absolute shit show.

Not even gonna bother waiting for Navi tbqh and almost every GPU I've ever owned long term has been amd ati radeons.

Sadly I just can't be assed waiting for decent driver support 1-4+ years after launch

>>61979717
>(i don't really know when the next major Crimson update lands)
october

>>61983247
That's late.

>>61983286
that's amd. they are on a budget.

>>61983303
Ryzen stronk, they have money to spare right now.
But their schedule is tight, they've yet to launch RR, Vega11, and the rest of the EPYC stack (besides 3 top SKUs currently available).

>>61983348
they still don't have money to spare. next year this time around, they would have the money thanks to epyc. but not right now.

And I'm genuinely curious about Vega 11. How are they going to cut the Vega chip down to replace rx 580, is mystery to me.

>>61983392
2.5k ALUs.
GP104-sized thingy with GDDR6.

>>61983410
>GDDR6
not ever. they will use hbm2, that's a given. 2.5k alus are way too low. it has to perform better rx 580. unless they go with performance for watt. vega 11 would much more efficient.

>>61983442
Anon, GCN scales now due to primshaders.
They may also use fabled low-cost HBM from Samsung (HotChips is near, we'll hear about it soon).

>>61983466
>GCN scales now due to primshaders
we didn't even see yet. we really don't know anything about it.
>They may also use fabled low-cost HBM from Samsung
all of the vega cards have samsung hbm. the price won't change. there are no other one making it. Hynix fucked up.
>we'll hear about it soon
I'm really tired of waiting.

Just like NVIDIA adds GPCs going up (2 for GP106, 4 for GP104. 6 for GP102), AMD will add/remov ALUs. More ALUs = faster geometry culling.

>>61983504
Not HBM2, anon.
Low-cost HBM.
And yes, we saw werkstation performance of Vega10.
It definitely scales more than decently.

>>61983522
>Low-cost HBM.
isn't hbm1 dead? being stuck to 4gb and all.

>>61983539
Not HBM1.
Low-cost HBM (it's literally called that, google Samsung's HotChips 2016 presentation).

>>61983577
>Low-cost HBM
ok, that looks promising. did we see any samples? can it be functional at december?

>>61983629
Only those that are fuckbuddies with Samsung (the likes of AMD and such) know.
But yes, it looks more than promising.
Especially the part about lowering the TSV count.

>>61983647
just the thought of it getting into consoles and ryzen apus make me support amd even more.

G-d bless Joe Macri & his team for creating new memory standards (that actually work and are usable).

File: 1498726462918.jpg (108KB, 960x960px) Image search: [Google]

108KB, 960x960px

>tfw my next gpu will have to be nvidia at this rate
i hope they have decent loonix support

File: Rod Broward.jpg (561KB, 1500x1714px) Image search: [Google]

561KB, 1500x1714px

>>61963522
Don't forget Fallout 4, and the season pass, and Skyrim for Switch.

File: quadro-spec3_0.png (75KB, 1208x548px) Image search: [Google]

75KB, 1208x548px

>>61979845
depend

>>61986236
http://www.tomshardware.com/reviews/amd-radeon-vega-frontier-edition-16gb,5128-6.html
>Since we’re comparing graphics cards made by more than one manufacturer, their drivers make a significant difference in the results. Thus, we're shaking up the benchmark suite a bit by deliberately not using the SPECviewperf freeware suite. It just doesn’t reflect the current state of professional workstation and CAD applications. Its workloads are just too short and old. Also, some of its benchmarks don’t even start without the presence of professional graphics cards or their professional drivers.
>Understandably (but also unfortunately for reviewers), AMD and Nvidia optimize their drivers to target popular benchmarks. This practice diminishes the value of results we generate since performance differences become unrepresentative of other workloads or card behavior in general (we're talking about you, Maya). It all makes for great marketing, though.

>>61976500
980 Ti is not more powerful than the 1080, it's slower than a 1070.

>>61977831
>Primitive culling through prim shader is not working.
Discard is working.

>Vega FE and RX Vega are out
>AMDrones still think it can compete with a 1080 Ti
Wait for Navi you clueless fucking fanboy cucks maybe you'll finally be able to beat a 1080