http://www.guru3d.com/news-story/po ssible-radeon-rx-vega-3d

File: amd.jpg (952KB, 2661x1638px) Image search: [Google]

952KB, 2661x1638px

Same device ID as the flagship Vega sample demoed a few months back too.

Good job AMD, first you show your incompetence with Poozen then royally fuck up Pooga.

File: 15282091_325948244455834_1493132613_n.jpg (158KB, 679x480px) Image search: [Google]

158KB, 679x480px

>>60146186
t. angry corelet with a shitty i7

>>60146186
>rumours
kek fucking maximum damage control kikes

>>60146231
>BUTTMAD AYYMDPOORFAG DETECTED

Look at the device ID in the Doom image, and look at the device ID from scrolling down

JUST WAIT
U
S
T

File: vega.png (44KB, 428x967px) Image search: [Google]

44KB, 428x967px

ayy

>>60146225
>that image
saved

File: amdinanutshell3.jpg (2MB, 700x5000px) Image search: [Google]

2MB, 700x5000px

>>60146230
I'm already saving up screenshots for new OC coming up with the release of Ryze. You subhumans never learn.

File: 3E91CC9800000578-4347068-CHECK_IF_YOUR_PARKING_BRAKE_S_STILL_ON_House_leaders_yanked_the_-a-4_1490389757094.jpg (68KB, 634x423px) Image search: [Google]

68KB, 634x423px

>>60146186
DELETE THIS NOW DELET THISSS

File: amdinanutshell.png (362KB, 607x608px) Image search: [Google]

362KB, 607x608px

As always

>>60146186
I can agree, they fucked up with Polaris and with hyping up Ryzen as the best gaymen machine. But this is just bad. Ryzen 5 is now the only good thing AMD has done

>>60146186
Little Vega, remember.

>>60146410
>Little Vega

Big vega is dual gpu isn't it?

I can already see it. "1.5X greater performance than GTX 1080"

>>60146186
>gpu clock 1200 mhz
into the trash it goes

Rebranded Fury X at14nm with 8GB HBMeme2 ?

>>60146265
>1200Mhz
>700MHz

>>60146186
it always lies on the price point. it could be literally the next shoa for nvidia

File: vega tpu.png (147KB, 1145x1076px) Image search: [Google]

147KB, 1145x1076px

>>60146493
>>60146512

the last leak was @ only 1000 Mhz

>>60146551
That's not a leak, that's a placeholder page, with completely wrong details on top of it.
Wrong GCN version, wrong bandwidth, wrong clocks, heck it even said April 28 release date a few days ago.

>performance estimate based on arch, shaders and clock

File: 1493341109921.jpg (34KB, 662x684px) Image search: [Google]

34KB, 662x684px

http://www.3dmark.com/spy/1544741

>>60146586
>10.5 inches/267 mm

That very accrute, I don't think it's bulshit.

>inb4 hurr durr techpowerupshill

>>60146636
That's the size of a regular Nvidia/AMD aftermarket card, we've already seen images of Vega and it's the size of Fury X.

Also arguing as if the fucking (wrong) dimensions which have been standard for years already somehow makes the technical specs written there not valid is beyond silly.

When did this board turn into /v/-tier console war garbage?

So it's 5% faster than Fury X? That makes zero sense, we're not talking a 28nm GPU there.

File: 1390964422957.jpg (13KB, 400x300px) Image search: [Google]

13KB, 400x300px

Vega has ~40-50% more transistors and 45% higher compute performance than Fury X.
If you believe a benchmark that shows Vega is only 9% faster than Fury X, then you have shit for brains.

>>60146186
You retards do realize that this can't be real, right? They could've overclocked a Fury X and gotten the same results.

lame

Ignoring the core clock, the memory clock is beyond suspicious, it could be some kind of Nano-tier card, but there's no reason to underclock the memory that much.

File: 47460-h80_small.jpg (78KB, 800x539px) Image search: [Google]

78KB, 800x539px

>>60146754
Maybe pic related will die, Fury X was a inefficient housefire.

>>60146186
I feel good now with my GTX 1080 purchase.

That being said AMD only redemption lies in pricing. If its going to be around $300 it could easily beat GTX 1070 which is around 350-400

>>60146769
>45% higher compute performance than Fury X.

Source?????

>>60146769
Fury X and Vega have 4096 cores.

>>60146844
>HBM2 on a $300 GPU
doubt

File: IMG_0741.jpg (42KB, 516x509px) Image search: [Google]

42KB, 516x509px

>>60146186
>mfw i bought amd stock

https://videocardz.com/amd/radeon-instinct/radeon-instinct-mi25

>>60146872
And Vega has around 1600MHz clocks and a new architecture.

>>60146769
>that delicious denial
The constant disappointment of AMDrones is something I look forward to every year.

>>60146853
MI15 is Vega with 12.5 TFLOPs

>>60146186
>slightly better than a fury x
i don't think this is real

>>60146926
*MI25

>>60146551
>Hbm 2
>1000Mhz
Well maybe it could possibly be 700, but not really, and totally not 1Ghz

Journalistic integrity at its best, all it takes is to look at the enterprise Vega card tflops to see that this is bullshit.

This is either.
1. Some low power version ala Nano
2. ES, which makes more sense as neither core or mem clocks make sense

File: 1310483412100.jpg (36KB, 413x395px) Image search: [Google]

36KB, 413x395px

>>60146186
>barely above the 1070

This can't be real kek

File: AMD-Radeon-Instinct_Final-for-Distribution-page-017-2.jpg (830KB, 3999x2250px) Image search: [Google]

830KB, 3999x2250px

>>60146926
>MI25
>feel on the amd compute meme numbers

If I fowler your logic, the rx 480 is faster than gtx 1080...

>>60146329
>my 7700k offers 10% more gaymen performance compared to 8 cores clocked 1GHz less!!!!!!!

>>60146186
>ITT: faggots arguing about a POSSIBLE score with no concrete tests
/g/ is always eager for fanboy wars. fuck the underage faggots in here.

>>60146891
>Card Status: Unreleased
>300 watt
>This card has not yet been released, specifications are subject to change without notice

Paper launch

>>60147024

http://www.3dmark.com/spy/1544741

>http://www.3dmark.com/spy/1544741

http://www.3dmark.com/spy/1544741

>http://www.3dmark.com/spy/1544741

http://www.3dmark.com/spy/1544741

>http://www.3dmark.com/spy/1544741

>>60146983
it's a 45% uplift over Fury X
OP shows 9%
OP is obviously bullshit

File: Disgusted Beta.jpg (116KB, 1280x720px) Image search: [Google]

116KB, 1280x720px

>>60146186
I thought we were finally dwindling down on the non-stop Ryzen shitposting now that it's by far the best choice with the 1600/1600x. I can't wait for Vega shitposting.
This makes no sense whatsoever. It has the same cores as the Fury X on a smaller node with only minimal gains. Either this is a super downclocked version or there really is a smaller Vega. The latter would be nice, I'd love them to offer a smaller Vega at the $300-$400 price range.

File: 1d0416739c31389a56dafaf0a2e8cf79.jpg (25KB, 488x472px) Image search: [Google]

25KB, 488x472px

>>60146186

I'm pretty sure there will be 2 Vega cards out there. One to competive with GTX 1070 one to compete the 1080 (or maybe even the TI)

That being said, this looks like to compete against the GTX 1070, but seeing how the performance is relative the same, AMD only go for the cheaper pricing technique

>>60146186
>possible vega entry
ayyy, nvidiot shills hard at work.

>>60147069
The MI25 is not a dual GPU?

>>60147078
>One to competive with GTX 1070 one to compete the 1080
That would be rather pathetic, those cards have been on the market for almost a year now

>Just bought a 1080
Damn, what a relief.
Always bet on AMD to be a fucking failure.

>>60147086
http://www.anandtech.com/show/10905/amd-announces-radeon-instinct-deep-learning-2017/2

F U R Y X 2.0

>>60147086
Unless it has 32GB

>>60147100
>Just bought one of the worst price to performance cards out there.
Damn, what a retard.

>>60147068
this means absolutely nothing. If you're so eager for a fanboy war you can be patient for few more days and keep shitposting in the faggy muh Ryzen cores VS muh Intel Giggahurtz threads.

>C87F:C1
Lol, this thing is from fucking December

>>60147129
He said he bought a 1080, not a fury x

>>60147092
Yes that why I said pricing is where at it.

The 1070 is around $340-$400 on newegg. ($340 is a small version for ITX, the proper MSI/ASUS/Gigabyet/EVGA are all around $380)

If the Vega with the similar performance as the GTX 1070 goes sub $300 it may be worthwile, specially if you have a Freesync Monitor)

I just bought a GTX 1080 but no regrets, EU prices will be fucked over for the next 3-4 month after release

>>60147131
Why that mean nothing? How you can fake 3dmark ID?????

>>60147074
>Ryzen shitposting now that it's by far the best choice with the 1600/1600x
Yeah...no

>>60147148
>GTX 1070 goes sub $300 it may be worthwile

I don't see that happening, not with the leaked die sizes and the HBM2 memory

>>60147157
Have fun with your 4 cores at the end of this year when they're already getting starved right now.

How many times do I have to tell you to not reply to fucking bait threads? and sage every post if you're gonna reply.

>>60147155
>GPU could be underclocked
>The GPU BIOS could be tampered with to make it unidentifiable
>Could be an engineering sample
There is many possibilities and it is funny how impatient everyone here is to jump on the bandwagon of shilling with/against AMD.

File: 1491166649444.jpg (66KB, 568x612px) Image search: [Google]

66KB, 568x612px

What's worse is when it comes out it will perform 15-20% worse than it does now. Show me one case where modern AMD processors and GPUs haven't disappointed users in performance tests after release.

>>60147175
Nice memes kid

Have fun jerking off to Adoredtvs video's

>>60146484
there are likely going to be at least 4 variants of vega, 4gb/8gb cut/uncut mix and match.

at vegas absolute worst, it should be matching a 1080, especially if they focused almost entirely on alleviating the bottlenecks.

So Vega is a overclocked Fury X? That's nice.

File: 2017-04-30-15-32-24.png (14KB, 1002x215px) Image search: [Google]

14KB, 1002x215px

HAHAHAHAHA, look this AMDrone damage control post

>>60147169
AMD co-created HBM2 and we are not sure how expensive its.

I remember when there was a type of Second gen Memory what was cheaper to produce (and was more efficient) than Gen 1 (Maybe DDR1-DDR2?)

>>60147227
>it's not real reeeeeeeeeeeeeee!

>>60147227
This is some pretty terrible falseflagging.

Also
>reading comments on anything but RWT

File: 85866.png (35KB, 650x350px) Image search: [Google]

35KB, 650x350px

>>60146329
>gaymur forgetting the 140w TDP 6900k housefire
Top kek
>inb4 b-but MUH GAYMEZ

>>60147228
Then there's still he massive die size

Maybe we need moar wait™ or moar cores?

Is anyone dumb enough to believe this when you can see the clocks displayed in the same bench?

>>60147291
Look at this thread, Intelfags will latch to anything at this point.

>>60147227
kek

>>60147227
This guy is crazy
>>60147258
Everybody read the comments on news

>>60147207
Nah, fuck all the tubers, they all suck.

>>60147323
These are youtube-tier comments at best, how can you stomach that level of discourse?

>>60147227
>This is not real.

RX 480 has a Memory clock of 2000, he telling us that a Pro Duo has 1/3 of that.

I'm 90% sure this is Vega, seeing the Memory is at 700. Its most likely just a test version with a lower GPU and memory Clock

>>60147155
we know that passively the card hits what, 1525/1575, i forget what the pro card hits exactly, so the consumer version, that has a proper heatsink and 2-3 fans pushing air clcoks what, 20~% lower then passive?

I call bullshit on anything not clocking on par with the pro line till amd come out and tell us vega clock speeds.

>>60146186

>1200mhz
>eng sample.
>already proven to be on par with a 1080 with alpha drivers that aren't using its full potential.
> NVIDIA slaves in G try to bash it.

Good luck with that you useless niggers.

>>60147342
>know it's at least 12.5TFLOPs
>know shader count

It's not magic to calculate the clockspeed.
This isn't even up for discussion, consumer Vega may clock higher at stock or may not depending how much headroom they wanna leave or how much lower power they wanna go

>>60147348
*IF* Vega is on par with a 1080 = big disappointment

>ends up being a mobile chip due to the heavy underclocks on both clocks

>>60147379
I hate predicting prices but even so if its on pair with GTX 1080 at performance but cost $100 less, it may not be a dissapointment

But that being said, Nvidia just cut the prices of the GTX 1080 and call it a day

This is good. It means they have to sell the card for $350 max. I hope they'll do a Nano card this time around too.

>>60147366
And I don't want to do math, its some where in the 1500 range I listed.
At amds worst, its on par with a 1080, at where I think they will land, half way between a 1080 and 1080ti, and at best, playing ball with the 1080ti or titan Xp, really it all depends on where they land with alleviating bottlenecks.

>>60147379
Not really, but then again I want amd to fuck off from the consumer gpu market for a generation or two, just because of how much I hate nvidia owners and their 'fuck amd, now my nvidia gpu costs more' line of logic.

>>60147469
>I want amd to fuck off from the consumer gpu market for a generation or two

They already did that.

Wasn't there a leak with the same PCIID months ago in compubench at the same clockspeed and it was between a 1080 and 1080ti?

Hard to believe they didn't get clocks up at all in that time.

>>60147379

Read the rest of the post. If the card is on 1080 level with alpha drivers and showed off so early in its development. Then it should be at 1080ti performance level upon its release. If it was a finished product six months ago with all optimizations done the it would have been released by now.

Aka it's a 1080ti similar performer that will cost a bit less and that's a good thing for everyone.

>>60147502
But Doom is not an AMD game?

>>60146186
>>60146265
That's exactly what a Fiji card with 8GB memory would perform like at these clocks. It's not VEGA. It's not even a die shrink.

>>60147502
>If the card is on 1080 level with alpha drivers and showed off so early in its development.

It's 1070 level with alpha drivers and low clock. They'll try upping the clocks to the breaking point like they did with Polaris, Fiji, and Ryzen, so it'll reach 1080 level barely, but then it will overclock like shit.

>>60147548
Maybe VEGA is a more power efficient Fury X

>>60146329
For all that effort you put into your shitpost, it's utterly worthless as Ryzen is an incredible success for AMD. Spend your time better.

File: Untitleds.png (15KB, 1677x139px) Image search: [Google]

15KB, 1677x139px

>>60147592
Nice.

>>60147580
The GPU is 15% bigger than the GP102, don't be ridiculous.

File: sss(1).jpg (81KB, 852x587px) Image search: [Google]

81KB, 852x587px

>>60147612

>>60147612
>Believing lies from Agner the AMD shill

File: 1492892250335.jpg (76KB, 858x555px) Image search: [Google]

76KB, 858x555px

>>60147658
>Agner is now an AMD shill
Oh how times have changed.

>>60147590
It doesn't justify what we know of its size, then. Something isn't adding up.

>>60147580
Yeah, no, it probably won't overclock at all. Not even Vega 20, which is planned for next year. Vega 20 is going to be another Polaris 20.

>>60147689
Shut up, anything that shows how good Ryzen is in server applications is obviously an AMD shill.

>>60147725
Vega 20 is AMD's first FP64 GPU since Hawaii

File: 1492890544839.jpg (62KB, 856x548px) Image search: [Google]

62KB, 856x548px

In know this is a GPU thread, so i'll just leave it at this.

File: AMD-Vega-GPU_2.jpg (46KB, 630x354px) Image search: [Google]

46KB, 630x354px

>>60147725
The chip is not bigger

>>60147774
You can size it up easily using the HBM stacks, it's around 520mm2

>>60147580
>but then it will overclock like shit.
for the love of me I can't understand why this is a problem. What is wrong with purchasing a card which is already overclocked to its limits? I'd take that anyday over tinkering with a chip which could've had better performance if the manufacturer wasn't so lazy.

File: amd-chip.jpg (86KB, 900x507px) Image search: [Google]

86KB, 900x507px

>>60147789

>>60147816
Yeah? That one is 601mm2 IIRC

Couldn't AMD just dieshrunk Fiji, added a bit higher clocks and get this? Makes little sense.

>>60147833
The shit I can't understand is why both chip have 4096 cores?

>>60147865
But has 50% more transistors, ain't that funny.

Cores are obviously more powerful, first Vega slide said that it'll clock higher and have more IPC

>>60147854
this x1000, Fury 2.0.

>>60146186
>possible entry with no driver.

Really makes you think.

>>60147877
If they kept the same amount of ALUs what point would there be to doubling their transistor budget if they aren't increasing IPC? Would be just a waste of silicon

You know, just doubling up Polaris 10 in size and significantly reducing the clockspeed would give better performance than this, it would also be a 20% smaller chip as well, not to mention it would have been released in September last year and be much cheaper.

Alternatively, a diehsrunk Fiji would do the same.

So unless AMD is literally braindead, and I really doubt it considering the job they did with Ryzen just recently, this 'leak' should be taken with a good ladle of salt

AMD's tflops claims are on base clock, so the Mi25 is 1530MHz without turbo.

This is looking rather nice.

https://www.youtube.com/watch?v=m5EFbIhslKU

Too low contradict official released info, can be nano, mobile or ES, or a mix of them.

Hmm

>>60148152
>TFLOPs based off base clocks

Proof?
If this is true I'll have to some recalculation of Vega's performance in my head.

>>60147792
>for the love of me I can't understand why this is a problem.

It raises the heat output and power usage greatly. 1070 and 1080 use fuck all electricity, and overclock great - because they are using low clocks by default. The 7970 had extremely low clocks by default too, to the point that they could do a GHz edition later.

Polaris ended up with power usage problems and high heat output due to being clocked way the fuck high. Ryzen had to get its clocks maxed too, to match Intels offering.

>>60148263
If heat and power is an issue then you can always underclock the card. It is a far more stable and easier process than overclocking.

>>60148254
11:53 here >>60148176


Also basing tflops off turbo clock is not preferable since turbo is variable, you're never guaranteed top turbo bin in all workloads, you're guaranteed the base clock always outside of power viruses that the drivers purposely downclock

File: grab024031.png (429KB, 640x480px) Image search: [Google]

429KB, 640x480px

>>60146186
You know that it is just this piece of information again, rediscovered but without realising it is still the same bit?

Since it runs at 1000 MHz base, 1200 MHz turbo, it is a prototype with non-final clocks, measured in February.

https://compubench.com/device.jsp?benchmark=compu20d&os=Windows&api=cl&D=AMD+687F%3AC1&testgroup=overall

>>60148298
The problem is that AMD has poor performance so they HAVE to max the clocks.

Nvidia does not. They have way better performance, so they can get away with using low clocks and lower TDP. If AMD somehow matches them in performance? They release a factory overclocked card and that's all they need to do.

>10% higher performance than Fiji 2 years later

Someone must be pretty retarded to actually believe this when Polaris has similar performance to a 390 with half the ROPs.
Just using Polaris uArch at that shader count would give them more performance, this would literally be worse than Bulldozer

When they test these GPU's on 3DMark etc they always keep them at lower clocks and exrapolate from there. Plus alpha drivers. It's normal for them to appear to be underperforming. Even at launch it will probably underperform just like Ryzen and the RX480 but both are now performing as good as if not better than Intel/Nvidia at a lower price point.

It's real nice that people actually bothered to explain what's going on, but you do know this thread will be reposted soon with the first 20 posts being wojak edits and greentext and you're gonna have to explain again?

>>60148331
It doesn't matter for me who has better performance than the other. What matters is that when I buy a card I want it to be at its maximum potential. I couldn't care less about heat and energy consumption. I don't run a data center so that is not an issue for me.

>>60148381
I'm done explaining, if people want to listen they'll listen, if they don't, I'm not wasting anymore time drilling through someone's skulls that Vega isn't an overclocked Fiji.

Anyhow, computex isn't far off

>>60146186
why are intel and nvidia so much better at agitprop. aymd literally can't keep up with the business acumen and propagandic skill of a bunch of nerd jews

>>60148444
I really don't understand why they would let a 50% underclockled test run out in the wild. fucking poojeets

If AMD is moving from 4 shader engines, I have serious doubts of the "leaks" showing it having 64ROPs, {4,5}*16 is no longer 64.

Vega is probably a bigger redesign than I thought, no wonder we have literally zero core config leaks even though the chip was taped out November

>>60146186
HAY GUISE AYYYMD IS LE FINISHED AREN'T YOU EXCITED TO OVERPAY FOR 1% PERFORMANCE "UPGRADES" EVERY YEAR NOW THAT THERE'S NO COMPETITION LEFT? I'M TOTALLY NOT A SHILL OR A BRAINDEAD RETARD BTW....

>>60148522
if it forces shitty pajeets to actually learn how to code and manage memory. ok

UHH PEOPLE
HEY PEOPLE ARE YOU FUCKING BLIND

LOOK AT THE ID

687F:C1

687F:C3 leaks have been out for a while, C1 is literally a ES from December or earlier.

>>60148385
The problem is that AMD has to push their cards to the max, which raises the TDP to Pentium 4 levels.

Nvidia does not, so they get a more economic card.

>>60148331

Lol. This is amazing /v/ faggotry that hasn't been mocked until the faggot that posted it hangs himself in his closet.

A fury X running at 1500mhz, so not taking into consideration the upgrade to HBM 2 or architectural refinements, would perform better than that.
If they even manage 10% "IPC" increase, 1500mhz base clock, 8gb HBM2 it's going to be extremely competitive and thank god for that.

My fear is that it will be voltage locked. I'd still probably buy it anyway because I want freesync. This isn't even hype, it's just logic. We know how the fury x does at 1050mhz, we know they put effort into architecture improvements, and we know the mi25 runs at higher than 1050mhz so logic says that 1080 performance would the bare minimum.

>>60148631
Perf/watt of Fiji and GM200 was within spitting distance, that was AMD's last highend GPU at the same die size and Maxwell had Tiled Based Rendering.

>>60148474

Vega is an all new architecture meant to take what GCN was shit at, and then make it better.

>>60148677
GCN really has only two big things it's shit at, geometry throughput and pixel fillrate.

One can be addressed by a wider frontend while the other is basically ROPs

>>60148631
Drop the "economic" card meme will ya? No one here lives in a rural village in Africa. And I doubt anyone here plans to run 100 GPU's for some server farm or something.

>>60146329
is this supposed to be falseflag making intel fanboys look stupid? because it's working

>>60148663
You're buying a cut down MI25 300W
Vega 10 is 225W

>>60146186
>>60137069

lmao

>CL_DEVICE_NATIVE_VECTOR_WITH_FLOAT=1

It's underclocked and for all intents and purposes only using half its cores

>>60148806
>reprogrammable shaders disabled

What the fuck

>>60148806
Shh, stop talking actual knowledge, you're interrupting the circlejerk.

>>60148806
that's retarded. why would they do such a thing. muchless let it out into the wild. are they fucking autistic holyshit

>>60148840
It's actually pretty common, awhile ago there was a leaked benchmark of an Intel i5 6-core running at 3.5 GHz base w/ no boost. Obviously this is an engineering sample that got leaked. https://wccftech.com/intel-coffee-lake-desktop-6-core-4-core-cpu-leaked/

>>60148840
Because this might not be from AMD, you know people mug cards/CPUs off the assembly line all the time? Some get caught too.
It also could just be some engineer doing some quick testing.
Anyhow, looking at the C87F:C1 ID, it's the same GPU used in December.
Here's something newer, C3, but as that's still pretty old.
https://www.pcgamesn.com/amd/amd-rx-vega-sisoft-benchmarks

>>60146844
>>60146873
Yeah.

Vega is going to be $500 minimum.

>>60148806
NOOOOOOOOOOOOOOOOOOOOOOOOOO!!

>>60147421
You're fucking delusional.

The Fury X was supposed to be AMD's best but it couldn't even beat a 980 Ti.

>>60148834
>chip that's basically taped out has major hardware circuits disabled

It's pretty fucking normal, why do you think it takes over half a year from tapeout to shelves?

>>60148925
Fury X was two Tonga's glued together, dude it kept the same frontend as the much smaller Hawaii's, that can mean only a bottleneck.

>>60148986
>replying to OBVIOUS bait

>>60148910
>NOOOOOOOOOOOOOOOOOOOOOOOOOO!
www.nooooooooooooooo.com

>>60146329
I've been screenshotting vega for a month

Seriously tired of every release AMDumbfucks say "X performance for Y dollars!!!!!!" Then when it comes out and isn't anywhere near that they say
"I don't see what's the issue here it's a great value nobody said it would be X performance this is what we were expecting"

File: 2017-04-30_163153.png (169KB, 948x677px) Image search: [Google]

169KB, 948x677px

>>60148599
Reminder that this is actually Vega.

Proof is pic related.

>>60148994
>Implying Nvidia fanboys aren't that stupid

>>60149017
It's fucking /g/, no one here is THIS stupid to know nothing about Fiji's shortcomings.

>>60149014
>Inb4 anon believed Ryzen was a failure
>Inb4 posting gayming benchmarks as "proof"

>>60149016
Why did you quote me? I never said it wasn't Vega, I just said it was an early sample, with disabled reprogrammable shaders to boot.

>>60149039
Ryzen was decent though
We aren't even talking about CPUs right now

>>60149014
The 1600 and 570 are by far the best value on the market at the moment, I don't know why you're shitting yourself.

>>60149044
I'm quoting you so you have a decent source, anon. Got your back desu sempai.

>>60148925
Please explain to me what the relative performance of the Fury X compared to the 980 Ti has--at all--to do with a logical inference of where Vega's performance should slot based on technical details of Vega's design?

You literally just typed out your own cognitive bias and posted it. Logically the two things have absolutely nothing to do with one another.

>>60149034
>no one here is THIS stupid
>people actually claiming a 7700k is faster than a 6900k

They're far more stupid.

>>60149053
That's obvious bait.

>>60149048
>BUY AMD GOY BEST VALUE
seems like everyone who buys AMD is pretty jewy with their shekels

>>60149063
I wish.

>>60149053
A 7700K is faster than a 6900K in single-thread performance.

>>60149080
Which manifests in low resolution synthetic benchmarks while the 6900k obliterates it in any CPU load and equals it at normal resolutions.

When you have to create a unrealistic situation of the 7700k to win, you've lost.

>>60149079
Nah, /g/ is full of endless shitposting. No one here is that stupid. sandley rheinallee

File: Your+bait+is+bad+and+you+should+feel+bad.png (4KB, 622x626px) Image search: [Google]

4KB, 622x626px

>>60149072

After the lauch
>We are aware of some games performance issue, ad we are working hard to fix them.
>Vega was designed for low end workstation.

>>60149034
Everyone here is that stupid.

>>60149121
Vega is designed for ML market actually.

File: IMG_1826.jpg (521KB, 960x686px) Image search: [Google]

521KB, 960x686px

>>60149113
>b-bait

File: 1493522127671.png (57KB, 625x626px) Image search: [Google]

57KB, 625x626px

>>60149072

>>60149080
it's so slim it doesn't even matter. It's the clocks that make the difference, otherwise the 7700k would be a shit cpu

File: thick wink.jpg (178KB, 542x926px) Image search: [Google]

178KB, 542x926px

i have a r9 390, what kind of increase would i get from buying the vega equivalent? 30%-50%? seems not worth it atm. when i upgraded from my hd7870, i got like a 100% increase

if the next generation is a big jump, it might be worth the upgrade for me though. i wonder when it will be worth to upgrade my 4790k

File: 1492103803448.png (104KB, 582x1048px) Image search: [Google]

104KB, 582x1048px

>>60149147
>Ryzen
>Stuttering

>>60147200
>what is an athlon64
Also see>>60147259

>>60149147
>1826
Made me kek watch newfags think this is an actual phonepost

>>60149175
>thinks that's stuttering
https://youtu.be/YSDTPPM9qsc

>>60149175
HNNNNNNNNNNNNNNNNGGGGGGGGGGGGGGGGGGGG

Oh wait it's that faggot again.

File: 1492876169454.png (1MB, 1110x5110px) Image search: [Google]

1MB, 1110x5110px

>>60149176
Kek. A literal house fire!

>>60149205
I thought he got banned.

>>60149175
>Le Intel stutter maymay
>Ryzen is silky smooth just like my hook nose

>>60149217
For half a day. Or he's evading.

File: 1493400743726.jpg (2MB, 3840x1032px) Image search: [Google]

2MB, 3840x1032px

>>60149198
>Being this mad
Sorry, wrong post.

>>60149172
Given how 1080 is incredibly far ahead of a 980, I would expect at least a 50% increase in performance going from Hawaii to Vega. I'm more on the Navi train since we might get small GPUs on a MCM, making a high end card cheaper due to more yields.

>>60149111
Nah, people are fucking stupid here.
Some retards here actually believe the voltage/clock curve is linear.

Laugh as you want. But AMD has already sold millions of Vega chips to Nintendo for their next SNES Classic.

>>60149147
>IMG_

>>60149238
>still thinks screen tearing is stuttering

>>60149251
I admit, this got a small laugh out of me.

>>60149198
m-m-m-u-u-h d-d-r-o-p-p-p-e-d f-r-r-a-m-m-e-s

>>60149246
>Some retards here actually believe the voltage/clock curve is linear.
No way people are that retarded.

>>60149258
>Frametime
Can you say that again? I think you stuttered, anon.

>>60149278
>STILL thinks the picture was portraying stuttering
oy vey goy

File: 1492879532221-g.png (359KB, 600x900px) Image search: [Google]

359KB, 600x900px

>implying I'm not hype for this card, whether the leaks are true or not

If that Vega leak proves to be true, we're still going to be getting a card more capable than a 1070. The 1070 is already a card that is perfect for 1080p/1440p 60fps gaming. If priced right, this would be perfect for AMD to become more relevant in the GPU market. Just like Ryzen. Ryzen wasn't/isn't the best for gaming, but the 3-10 fps difference wasn't enough to warrant an extra $90-$140 for a comparable Intel build. This will also apply to this Vega card compared to 1080/1080ti.

I separated myself from Intel and their Jewish tricks when Ryzen launched. I'll do the same if Vega proves more capable than my 980ti. Not only do I get a card that will smash anything on 1440p, but has proper async support, and will do freesync.

t. Nvidia user looking to get off this ride.

>>60149294
>>60149160

>>60149302
Hello fellow Nvidia user!

I too use an Nvidia card and I am tired of this ride, I am pre ordering AMDs new flagship VEGA as son as I can!

-sent from my FX8350

>>60148806
From where do you get this stuff?

>>60149327
>Fx-8350

If you're going to try and pretend shill post, at least make it current you faggot. No one in their right mind would still be using FX given the choice.

>>60149302
>amdummies are already backtracking to "well if it gives me Z performance instead of X that is still great for me I hate Nvidia"
This is happening sooner than expected

>>60149396
>Anon doesn't know
>>60148806

>>60149380
Don't be rude

I can't send from my 1700X because my RAM keeps crashing my computer. I however look forward to pre-order VEGA and don't worry it gets better!

>>60149380
If you were unable to discern that he was joking, then you are most likely quite literally autistic and should go get a diagnosis.

>>60149407
>bullshit post with no evidence

>>60149426
This was a retard bait/troll thread from the start.

>>60149414
I feel you, dude.

I can't use my 6900K either without my computer bursting into flames; and whenever I email Intel for any service, all they recommending me to do is "delid", or whatever that means.

>>60149374
https://compubench.com/device.jsp?benchmark=compu20d&os=Windows&api=cl&D=AMD+687F%3AC1&testgroup=info

>>60149434
>Being this fresh out of /v/
Kek.

>>60149459
You know the 6900k can't be dellided right?

>>60149449
Well, obviously, though I guess it counts as progress that the anti-AMD brigade is having to bait/troll about the as of yet unreleased Vega since Ryzen is now performing too well to convincingly bait/troll thread about.

>>60149492
Man, really? Then what was Intel recommending me to do all this time?

>>60149459
You should have bought an AMD Ryzen CPU! It is not to late to change!

Yours truly,
-Designated

>>60149509
Probably use the loo

Most likely got lost in translation

File: 1492946586387.png (158KB, 2033x909px) Image search: [Google]

158KB, 2033x909px

>>60149511
Dude, it's already too late; Skylake-X is almost out, and I hear it's going to BTFO Ryzen out of the water.

Pic related is already showing more progress than Intel had done before! Can't wait!

>>60149529
>benchmark shows possible vega entry running 8k 500fps
>WOW NVIDIA IS SWEATING AMD IS GIVING THEM HELL
>benchmark shows vega entry running badly
>SAGE FAKE NEWS

>>60146186
Some day AMD will be dead in the ground and there will be celebrations in the street

It's like watching a retard try to make tech, it's just embarassing

>>60149547
Show me at least one benchmark where Vega runs better than anything GP102-based.

>>60149547
>Implying that couldn't be done for both types of shitposts

>>60149548
We kinda need them though
Without AMD Intel and Nvidia are going to be fucked

>>60149548
>>60149547
Yes yes, you can all have your thread where you sit here jerking your tiny little hateboners over AMD. It's really pathetic to watch, though.

>>60149548
>Unironically wanting a monopoly

>>60149586
>you're just haterz
Literal millennial logic

>>60149530
Poor attempt at role playing. Took me out of the immersion.

>>60149039
ryzen was pretty bad tho. amd failed entirely in marketing and got demolished by intel within minutes of release atleast in the minds of consumers

>>60149587
>implying having pretend competition is better than no competition

>>60149540
This is fake news! Don't fall for the leaks my AMD brother! Intels skylake-x platform cannot compare to our memory speeds and PCI lanes! Don't worry we will break that 98% market share Intel has on servers just wait!

Signed
-venu malesh

File: 1493235815447.png (2MB, 769x8779px) Image search: [Google]

2MB, 769x8779px

>>60149608
Wait a minute, what's you agenda, anon?
>>60147612
>>60147634
>>60147689
>>60147757
>>60149215
>Inb4 gaymang benchmark retaliation

>>60149608
>atleast in the minds of retards that made their decision entirely on day one GAYMUR benchmarks

>>60146225
>"flagship"
you know theres a big and small vega right?

File: 8e8.png (8KB, 509x619px) Image search: [Google]

8KB, 509x619px

>>60149618

>>60149645
that's all that matters it seems. amd lost that battle

This thread sucks. Just look at what you are posting, this is worse than console wars on /v/

I hate neo /g/

>>60149639
>Ryzen releases
>Suddenly everyone on /g/ uses CAD and renders 8k premiere videos
>Naples is announced
>suddenly everyone on /g/ owns a server room and has a high electrical bill and needs pci lanes

>>60149663
Why not just shitpost on /g/, if gayming is your thing?

>>60149608
Intel just had a terrible quarter and their stock took a big hit in response. Unless the entire PC market quietly experienced a sudden collapse, it stands to reason that their low revenue relates to losing market share to Ryzen. We'll know tomorrow when AMD releases their quarterly report.

>>60146186
>possible entry

fuck off.

>>60149663
You live in a tiny little bubble of fanboys, retards and trolls.

>>60149691
That's not my point my dude. AMD got beat in they eyes of most consumers by intel with an ez marketing campaign that probably only cost them 1/1000th of ryzens launch. in the

>>60149696
Their stock went down a dollar
I'm shaking in my boots, whatever will they do with that 70% market share

>>60149690
Kek. What the fuck was /g/ supposed to discuss before Ryzen
>Bulldozer against Skylake
Now off you go, silly /v/ermin.

>>60149690
>Intel's only advantage is single core performance
>SINGLE CORE IS ALL THAT MATTERS
>MUH 10% HIGHER AVERAGE FPS
>Intel switches to 6 cores, negating said advantage
>HAHA JUST KIDDING, INTEL IS STILL BETTER THOUGH FOR SOME REASON

File: amdvega.jpg (156KB, 633x758px) Image search: [Google]

156KB, 633x758px

>>60146186

>>60149744
>Ryzen
>Barely same single core performance as a 2500k
>it took this long to reach the 2500k
>moar cores meme

>>60149764
>amd only finally caught up after Intel got lazy and sat on their ass for 5 years and are still having trouble
Good image they got

File: 1492949268415.png (143KB, 665x699px) Image search: [Google]

143KB, 665x699px

>>60149764
>In gayming
AHHHHHHHHHHHHHHHHHHHHHHH

>>60149731
>an actual intel stock owner
How does it feel to have made 20% return over the year, when AMD stockholders made >270%?

>>60149775
it's all that matters companies that matter get prebuilts. pcbuilders are majority gaming oriented

>>60149782
AMD stock shills are truly subhuman

Neckbeards who spend all day shilling AMD on forums in hopes they will finally be able to move out of mommy's basement

>>60149782
I'm pretty sure going from sub <$2 to $13-14 is far more than 270%

>>60149782
Only stock I own is AT&T

File: ayyymd.jpg (439KB, 1920x1080px) Image search: [Google]

439KB, 1920x1080px

AYYYYYY

>>60149795
>All that matters

>>60149799
>i hate free money
Okay.

>>60149821
yes

>>60149809
>Steam
When did this board get so infested?

>>60149764
You can cling to your tiny little island of increasing irrelevance if you want. Even Intel is moving away from it. Zen 2 will even up the single core performance anyway.

>>60149799
>this level of bitterness
Sorry you missed the train, but you have only yourself to blame. Should've known that /g/ is always right.

>>60149799
>He's mad about getting free money

>>60149839
Ages ago.

>>60149840
>wai..it for zen2 it will be better

kek

>>60149869
>Ryzen already has a higher IPC than RefreshLake

Didn't your mom teach you to report shitposters?

File: 1492153304675.png (47KB, 708x462px) Image search: [Google]

47KB, 708x462px

This is what happens when you let an H1B run the Radeon group.

>>60149829
>>60149850
>>60149852

AMD stock shills coming out of the woodwork

I have a real job, I don't spend all day shilling for AMD in hopes that my pathetic stock monies will go up.

>>60149840
>8 cores at 4-5 Ghz
i'll believe it once i see it
>it's not like amd would be unable to do that, we 'll just get zen2 housefire edition

File: 1492949616071.jpg (205KB, 1600x917px) Image search: [Google]

205KB, 1600x917px

OHHHHHHHHHHHH SHIIIIIIIIIIIIIIIIIIIIITTTTTTTTTTT

https://www.nytimes.com/2015/01/07/technology/intel-budgets-300-million-for-diversity.html?_r=0

>>60149939
You got a real job avatarfagging and posting wojak edits? I'm jealous, I still have to walk a few minutes to mine.

>>60149939
>i hate free money
Okay.

>>60149939
>Not having a job + free money
Sad!

>>60149939
>Intelel mad because he missed literal free money
:)

>>60149876
hahaha

File: Screenshot_20170501-004103.jpg (461KB, 1252x1965px) Image search: [Google]

461KB, 1252x1965px

>>60146986
>Enjoy your Intel Pentium single core perfomance

:^)

>>60149958
>>60149959
>>60149988

>free money
>I spend 24 hours a day shilling for AMD while living in my mom's basement

Really makes you think

File: Untitledd.png (36KB, 633x613px) Image search: [Google]

36KB, 633x613px

>>60150009
>>60149639

>>60150027
>i
>hate
>free
>money
Okay.

>>60150027
Hey, you're here with us, schlomo.

>>60150027
>Literally denying free money all in the name of brand loyalty
J U S T

>>60146329
/g/ doesn't care nearly enough for the gaming performance.

>>60150011
>1800X single core performance on par with 6900K at half price
Yeah, really got burned there.

>>60146186
amd poo'd the fuck out

>>60150039
>>60150056
AMD shills so delusional they think the stock market is free money

>>60150011
I guess you're right. Why don't we all just buy i3s?

>>60150078
>i hate free money
Okay. Why are you even posting here?

>>60150078
>He didn't see the signs and missed out on it
>He's now in damage control
How are your stocks doing from Intel?

>>60150088
>maybe if I repeat it enough times, my delusions will become true

>>60150104
>i hate free money

>>60150011
Remember when I said up in the thread there are people retarded enough who would rather a 7700k instead of a 6900k?

I present to you, Exibit A

>>60150104
>He doesn't know that statement can go both ways
Heh, AMDonkeys, a-am I right?

Stop bullying the retarded.

>>60150027
>Bought AMD stock
>stock multiplied several times over
>not free money
You would be more convincing if you weren't so obviously desperate to make excuses for your bad decisions. You realize nobody needs to do anything but hold stock to make these gains, right?

>>60150136
No

>people mad at AMD stock owners
Very few of the people in /g/ invested $10,000+ in AMD, so it doesn't matter if they invested.

Most of the people (see: NEETs) only invested $100 max and built a new PC with the money they made.

>>60150144
At least he didn't invest in a dying company his conscience is clear

>Vega MI25 runs at 1525mhz while air cooled
>Vega consumer card (releasing with water cooling) running a popular benchmark 1000mhz
What did AMD mean by this?

i know that image in the op is fake but i'd be happy for amd card with 1070 performance for less money so i can get that 1440p 144hz freesync monitor.

>>60150168
AMD isn't in any danger of dying anymore. The increasing share price allowed them to settle and/or transform practically all their high interest debt a couple months back.

>>60150204
1070 is too weak for 1440p 144Hz unless you drop settings significantly

File: 1493283204085.jpg (61KB, 500x667px) Image search: [Google]

61KB, 500x667px

>Meanwhile at Nvidia and Intel
"Oy vey! I smell a penny!"

>>60150065
Why do AMD fanboys always jump to the 6900k
You know nobody bought that right?

>>60150185
The MI25 is not real

>>60150209
They are still in debt

>>60150302
According to Intel's Q1 sales drop in Client with a full quarter of Kabylake revenue, nobody bought a 7700k either.

>>60150322
Ouch

>>60150318
Only high interest debt is bad though. Literally every company takes on low interest long term debt to facilitate growth.

>>60150322
Actually they did

>bump limit

NEW THREAD

>>60150417

>>60150417

>>60150417

>>60146186
So instead of a housefire flagship it's goint to be 1070 performance, possibly lower price and most likely better power efficiency? Guess I should care even a little bit about it after all.

No way this is the flagship vega, why the fuck would amd release an gpu that's only slightly faster then the R9 Fury X?!