Vega Frontier in game mode performs like 1080

What's the difference between Vega Pro WX

https://videocardz.com/70526/amd-preparing-radeon-pro-wx-9100-with-vega-gpu

And Vega FE?

wasn't this supposed to be faster than a titan Xp? a bit underwhelming imo.

>>61129767
http://support.amd.com/en-us/download/workstation/certified

Pretty much that and mostly that.

>>61129781
The results are very inconsistent across 12 different runs. The guy was using it with a crap 550 watt PSU and an mITX case/motherboard. It was obviously throttling/downclocking itself.

>>61129703
>MUH GAYMS

same people who complained about ryzen 7 getting trashed by like i5s

>>61129703
DELET

>>61129703
it basically over for poor vega I was gunna go for one but the delay delay delay then deliver trash just makes me wanna go nvidia next built. gunna wait for volta atleast Nvidia makes me believe they can deliver sometimes

>Tested on a potato by some mouth breather.

Nope!

>>61129850
so what you're saying is AMD fucked up or AMD fucked up?

His channel (May be doing some more videos) https://www.youtube.com/channel/UCtUuZ6l_xTfGGEx3efxYUkA

His Disqus where he's been posting shit https://disqus.com/by/klaudiuszkaczmarzyk/

Take all results with a grain of salt. His setup is pretty janky and he's been getting inconsistent results.

>>61129850
I'm a mouth breather and fuck you!

BUT HOW GOOD ARE THEY AT MINING ETHEREUM

>>61129932
pretty bad if 1060's are beating them for less than a third the price

>>61129803
In my experience, having too small of a power supply ends up in blue screens NOT a GPU down clocking itself

>>61129961
It was likely the small case that did it. Can't dissipate heat well enough.

>>61129945
wow so VEGA=
>shit for mining
>shit for gaymen
>shit for pro

>>61129945
BUTH ETHEREUM IS MOONING AGAIN I NEED MY HASH/SEC GAINS

Hopefully this drives prices down radically for the RX launch. Then after I buy one or two the prices can go back up with gaming drivers released.

>>61129703
That's extremely poor for a card that is 2 years late and costs $1000.

Maybe AMD should call prosumer Navi "Navi FE, a workstation card that can play games, although not as good as full gaming card"

>>61129945
There aren't any mining specs yet.
>>61129993
It's the fucking Founders Edition card, not RX Vega which is for your gaymes.

>>61129988
I personally think they should intentionally sabotage gaming and pro cards to set off a thermite packet and melt the silicon 4 good if it senses it's being used for mining

>>61129767
What about Vega Jw?
Just wait edition fury 2.0

>>61129847
Nvdia just postponed because they know Vega sucks

The only way Vega RX isn't utterly and completely, crushingly DOA is if they price it at 300$

>>61130020
I'm pretty sure that nVidia and AMD couldn't give less of a fuck what their cards are being used for as long as they're selling faster than they can be produced

File: 1485635797490.png (232KB, 718x369px) Image search: [Google]

232KB, 718x369px

>>61129703
>Currytech
>Article literally by Pajeet Sendloo
>Benchmark done on an antique i7, DDR3(!) RAM, and a dollar store PSU
If the issue was the PSU and Vega performs better than it showed (highly probable), AMD straight up has a defamation lawsuit against Currytech

The long game has been over since Nvidia released the 1080ti.

They have no management credibility at this point. I cannot understand people who continue to give Radeon group the benefit of the doubt, because we have seen this same situation play out multiple times now.

>>61130057
It'll be $450-$500.

File: 7ef.jpg (29KB, 600x600px) Image search: [Google]

29KB, 600x600px

>>61129872
> AMD coded GPU firmware to downclock when it notices insufficient power delivery
> Some random fag throws GPU into computer with shit power supply
> GPU downclocks itself to prevent erroneous behavior due to lack of power
> Fag posts numbers showing downclocked Vega running at "only" the performance of a GTX1080
> Intel monkeys flings poo everywhere cause they're still mad X299 and i9 are thousand dollar house fires waiting to happen
> Nvidia XTREME XxX_Gamer_XxX children joins in cause that's how children generally behaves in the sandbox

How about we actually wait until more people test this thing?
Or how about we wait until the actual cards come out and let people test those, rather than having shitflinging contests over a card that's quite literally AMD's version of Nvidia Quadro and thus labeled as "Workstation" everywhere.

>>61130068
it's a 550w corsair platinum

>>61130057
Or if, ya know, putting a 300W workstation card on a 550W power supply that probably needed fucking molex adaptors for the eight pin is the problem

Which do you think is more likely?

File: C7MxQiEV4AAYTh2.jpg (146KB, 1200x780px) Image search: [Google]

146KB, 1200x780px

>>61130072
>It'll be $450-$500.
R I P

>>61129781
in half and double precision it crushes the xp mercilessly.

The problem is with this GPU right now is we don't have gaining drivers for it, I don't believe. This was a professional card, not anything gamer, and professional drivers never work well with gamer loads.

Were also seen day one performance without a real driver update.

And don't think I'm making excuses for Vega, I honestly thought that day one it is going to be at the very worst at 1080, if it was Polaris efficiency it would be halfway between the 1080 and a 1080 TI, closer to the TI than the 1080. And with efficiency as their main focus or at least what they said their main focus was skies the limit on getting close to a Titan XP.

I count this as a preview but not the final version, the actual gamer version will be the final say on if it's good or not, when we can actually take the price that AMD is willing to sell it at into account not the professional use were paying 1000 for the card price.

>>61130092
You know the cheapest 1080 on the market right now is $550, right? Retard.

>>61130074
exactly so AMD should have given out to reveiwers and I dunno published a munual with specs or something so they can point this shit out

Instead they let the first pre orders get into the hands of the braindead

AMD lost the marketing game in the first hours of this debacle

File: de4ed93ec1511e66320b21a4b5de221128e01e8cccd9d8e10e57d970a3155480.jpg (506KB, 2000x1519px) Image search: [Google]

506KB, 2000x1519px

HEY GUIS JUST GONNA TEST THIS $1000 WORKSTATION GPU ON MY "RIG" HERE WHAT COULD POSSIBLY GO WRONG?

>>61130107
>the cheapest 1080 on the market right now is $550

AMD gpus have to be 20% lower price minimum @ same performance to have a chance in real life and not neckbeardland

>>61130074
>AMD coded GPU firmware to downclock when it notices insufficient power delivery
Prove it
I've never encountered a GPU that down clocks because of a power supply.

>>61130117
>ITX case
>shits it up with 10 fans

Dumb niggers

>>61130117
why does the card need 500 watts draw, and a 1500watt psu, raj?

>>61130126
We already know NVidia outsells AMD even when they've got a worse product (Thermi) so AMD really should just drop out of consumer GPUs altogether if this is the shit they have to deal with on a regular basis. Let the gaymurs enjoy their monopoly prices.

>>61130128
> I've never encountered a GPU that down clocks because of a power supply.

Have you ever used a Quadro card or any other workstation card?

>>61130112
AMD lets the engineers name CPUs shit like Epyc, nobody is defending their marketing.

>>61130107
>>61130072
>>61130092
So Vega will be $800+ after all the miners buy them up right?

>>61130117
I'd he kept that pannel off, it's not down clocking due to heat.

>>61130131
It's 300 watts for aircooled and 375 watts for watercooled. Fuck off, troll.

>>61130163
answer me raj u fuk

>all these GPUs
>no games even require a fraction of this shit for 1080p 144hz gaming
>4k+ monitors are still a billion dollars

When will display tech catch up?

>>61130117
If he tested it like that it's no different than an open bench honestly. AMD's blower cards are still kinda shit regardless too.

>>61130126
>In real life and not neckbeard land
The mythical Chad who builds his own PC?

No, admit it, even neckbeard enthusiasts are dumb brand loyalists

>>61130146
>Have you ever used a Quadro card or any other workstation card?
No, but the firmware is very similar to consumer GPUs.

You have yet to prove it would downclock due to power.

File: nigel vaping.jpg (28KB, 472x525px) Image search: [Google]

28KB, 472x525px

>>61130117
looking good there m8

>>61130153
EPYC is good though, it's a direct jab at Intel's abortion.

Threadripper less so.

>>61130174
>even neckbeard enthusiasts are dumb brand loyalists
true

>>61130173
So either it's downclocking because it's not getting enough juice or the thing just throttles and tosses up inconsistent results (as seen in the OP) all the damn time because the blower sucks. I'm willing to bet it's column A. All shitposters are going to bet it's column B.

>>61130208
Even Nvidia's blower cards throttle, this won't be any different.
It's unknown if the FW has any power throttling based on PSU current though.

>>61130208
I've seen people run SLI 1080s on a 600w ITX PSU. It's most def the latter. Hell even the Titan Pascal would throttle. Blower coolers are just shit all around. It's going to be even worse with a 300 TDP card.

>>61130112
Why send out a workstation gpu for people to use in non workstation products? also amd lost the gpu race a long time ago and will likely jump back in with navi.

>>61130074
You do realize it don't works that way right?
Also a quality 500w psu can easily deliver 700w for a doable time. Even more, with the system it shouldn't be doing more than 500w.

If you call the problem is with drivers okay, I can accept the answer, but blaming power supply? Come on you have or haven't a stable psu. If there isn't enough juice there is crash and blue screen, not throttling.

>>61130062
amd does because even if they sell the cards, they see no profit from them outside of msrp, and then they don't get market share in a very lucrative market, which will further optimized for nvidia.

Nvidia could give two shits, but amd does.

File: 1492976168995.jpg (39KB, 604x339px) Image search: [Google]

39KB, 604x339px

>>61130267
wait for Navi™

>>61130291
With an i7, watercooling, the Vega and all the fans he has running, he's drawing well over 500 watts peak

>>61130301
well, if amd can do for gpus what they are currently doing for cpus, navi will be able to deliver a fuck load of preformance for an overall lower price than nvidia can put out with monolithic designs.

>>61130315
Thanks the fucking card is not wasting heat like hell.
300w VGA + some shitty fans/pump/see less than 50w. A 4790k don't draw 150w but let's say 150w. That will go 500w well under a sustainable load on a Corsair platinum.

The only real hardware problem I can see is if the case is closed and the GPU is choked.

Just drop this psu meme

File: 1423100875102.png (187KB, 600x339px) Image search: [Google]

187KB, 600x339px

>>61129703
Why's everyone thinking it would beat a titan xp/980ti in gaming ?

Even AMD themselves said it would be between a 1080 and ti.

Don't get me wrong, barely beating a 1080 isn't a good start but vega with mature drivers is looking to be +10% faster that the 1080?

If it comes in at a decent price its a bonus.

>>61129945
where are you seeing hash figures? I'd think it would be pretty good given the compute capability and hbm?

>>61130418
>Even AMD themselves said it would be between a 1080 and ti.
[citation needed]

>>61130172
>what are 1440p high refresh rate screens

>>61130294
I really, really doubt AMD cares who's buying their cards as long as they sell. If anything, they're probably ecstatic about being the crypto miner card of choice. That's a pretty massive market, if it wasn't then /g/ and /v/ wouldn't be crying about card shortages all the time.

>>61130491
> probably ecstatic about being the crypto miner card of choice

every aymd card bought by minurz is one less gaymer with an aymd gpu, meaning a new Nvidia bot that will never switch back

>>61130117
I hope they keep the brushed ally look
>>61130450
I mean at all the quotes from amd the past month

>"will hold it's own"
>"competitive with competition"
>"perf looks nice compared to 1080 and ti"
>"Good value"

That just screams 1080><ti, I dont think they hav ever come out and said it would beat a ti

>>>"will hold it's own"
>"competitive with competition"
>"perf looks nice compared to 1080 and ti"
>"Good value"

That just screams 1080><ti, I dont think they hav ever come out and said it would beat a ti>>61130531
literally all of this was from a MSI guy not AMD themselves

Halfway between 1080 and 1080 ti for $400-$500 would be pretty nice.

File: 1498689415608.png (87KB, 1329x727px) Image search: [Google]

87KB, 1329x727px

>>61130552
>>61130531
Unless you mean pro performance, then it demolishes the entire Geforce lineup and holds its own against a $1800 P5000

>>61130574
That explains the price at least.

>>61130585
If RX Vega is half the price and sits between 1080 and 1080 ti, it won't be that bad. Hopefully $450 instead of $500.

>>61130574
>>61130518

>>61130172
>>61130484

The real question is can you find a monitor with a glass screen, not some "anti glare" piece of plastic garbage.

>>61130574
Why did it took so long to find some WORKSTATION SOFTWARE for a WORKSTATION GPU

>>61130552
it was from that Don guy, works at AMD
>>61130574
is lower better here? not familiar with these benches

>>61130623
>monitor with a glass screen
but y tho

>>61130626
Because the idiots that got FE Vega early are literal manchildren.

>>61130574
I feel less bad about the arch now.
I don't expect RX Vega to be more than 10% faster than a 1080, but eh, it's all in the drivers.

>>61130491
amd does not make over msrp
amd is not selling gamer cards to gamers
nvidia is getting a stronger foothold because of that.

it may not sound like much, but amd has sub 25% gaming market share and if it gets much lower we will see nvidia only games.

>>61130635
Because anti-glare plastic screens make everything look blurry. I have an easier time ignoring reflections (if there are any) than I do looking at a fuzzy "hi res" screens. This is the biggest thing that's kept me using the 1440p led apple cinema display I bought in 2011.

Much like getting your first hi-res (>1080p) monitor, you can't go back to plastic from glass.

File: 9a76104461ff967f30e270f7bd2b45ac27cd790db6e6ba1f0d4c53463f3b7ba6.png (188KB, 299x443px) Image search: [Google]

188KB, 299x443px

>>61130708
>anti-glare plastic screens make everything look blurry

>>61130705
source?

nvidia only games? more like step 1 in killing a platform

>>61130574
How much better is this than a p4000?

>>61130795
when did companies stop supporting dual core only cpus? not 4 thread cpus, 2 threads total.
the market share of pure dual core cpus among gamers was likely higher than amds total market share at the time the first ones came out.
It may never be an exclusive 'only playable on nvidia' scenario, but a what the fuck bother with amd optimizations so cards that are damn near equal, see 480 and 1060 6gb, would be the 1060 being a a good 50-100% more fps then the 480.

Hell, a game could come out and actually use physx as a core gameplay mechanic, effectively killing any non nvidia card.

the less marketshare amd has the more bullshit like that can happen.

Actually, I'm surprised they don't sell "premium" glass LCDs for people with more money than sense.

>>61129703
intel pajeet slide thread, sage in all fields

>>61131102
>gpu
>intel shill

kys

>>61130737
t.plastic pleb

Seriously though, go find a glass screened monitor and you'll understand.

>>61131115
>IGNORE THAT INTEL IS GETTING BTFO BEYOND REPAIR LOOK AT THIS SHITTY GPU AMD BTFO
(you) need to go back to the designated, pajeet

File: 1403224093650.gif (3MB, 286x258px) Image search: [Google]

3MB, 286x258px

>>61129976
This damage control is the most pathetic thing I've ever seen on 4chan.

>>61129767
FE isn't a workstation card

>>61129703
The primitive shaders aren't enabled, so I'll wait and see.

File: Screenshot_20170629_032844.png (136KB, 1515x585px) Image search: [Google]

136KB, 1515x585px

>>61131390
http://www.anandtech.com/show/11583/amds-radeon-vega-frontier-edition-formally-launches-air-cooled-for-999-liquid-cooled-for-1499


>b-but
>that..
>it...

No.

File: itsstilladisappointmentlol.png (131KB, 1040x742px) Image search: [Google]

131KB, 1040x742px

>>61131390
I dunno man

>>61131355

Happens with every AMD launch.

You should have seen /g/ when Bulldozer launched (I wonder if archices go back that far).

>>61129703

The original GTX Titans back in the day weren't great performers at gaming either. They were weren't that much faster than 7970 and 680 at the time.

Their own gimmick was general compute mainly FP32 perfomance.

The gamer-orientated refreshes of GK110 a.k.a 780 and 780Ti spank the GTX Titan at gaming for about 1/2 of the MSRP cost.

File: Quadro_P6000_Hitman.png (15KB, 622x581px) Image search: [Google]

15KB, 622x581px

>>61131432
>>61131437

Not this guy but AMD claim Radeon Vega Frontier is a ''hybrid'' card and Firepros are workstation cards. Also Quadros are workstation cards and they perform good on games

>>61131437
>$500 for liquid cooling

>>61131432
>>61131437
If the FE is a workstation card why does the Radeon Pro Vega exist, why will the WX 9100 exist and why do AMD compare it to the Titan Xp and not a Quadro?

Because it's Prosumer shit like the Radeon Pro Duo, hence the special hybrid gaming&pro drivers.

Now fuck off you clueless redditors.

>>61131494
You know the P6000 costs $5500-$6000 right?

>>61131441
https://warosu.org/g/thread/S12116229

From 2010

It seems like RX Vega is going to be find itself within 1080 and 1080Ti territory which is what most level-headed people were expecting.

The days of massive jumps in GPU performance for every generation have over over since Kepler/Tahiti.

I don't expect Volta to offer anything significant over Pascal in terms of gaming performance. Volta is geared more towards general compute since that's where all the big money is for large GPU designs.

>>61131355
>>61131441
I'm going to wait until the card is in the hands of some actual competent reviewers and not complete idiots before making any judgements.

And those judgements won't be based off gaymur benchmarks anyway. They'll be based off how well it performs in workstation tasks.

>>61129462

>>61129703
>Ryzen comes out
>AMD could be on to something
>Vega comes out
>AMD back in fashion with the latest disappointment
Surely they can't fuck up Threadripper this much? Epyc is already good.

>>61131507
Because "workstation" is loosely defined, it literally means a static machine for work you, difference between WX Vega and FE is that WX will be getting certification like WX Polaris, FE doesn't but still has decent compute/WS perf

WX Vega will probably also cost well over $2000

>>61131539
How many levels of autism do you have to be on to be "disappointed" by some random idiots running gayming shit on a prosumer card with day one drivers?

>>61131510
My point is than Vega Frontier was supposed blow the 1080ti even if it's a workstation card...and I don't talk about the P4000 or P5000

>>61131494
Firepro is dead, FE fills the spot the OG Titan left, WX is a direct Quadro competitor, but FE does decently against them as well.

>>61131539
You won't be getting your hands on FE Vega anyway, so just calm your nuts and Just Wait(tm) for RX Vega in a month or so.
>>61131557
P5000 costs $2000. That is still twice the price of FE Vega.

>>61131579
I'm agree but why at least the FE Vega can't compete with the 1080ti?

>>61130020
How can you be this retarded?

>>61131607
By the time RX Vega launches the gaming side of the drivers should be in better shape. It's still a hybrid card, though.

>>61131655
I hope

>>61131676
We haven't seen the performance of the watercooled version, either. It's more expensive, takes more power but also more likely to match the 1080 ti.

>>61131642
I would honestly pay more for it to become common technology tbqhwy

File: nvidia-gtx1080-specviewperf-scores.jpg (75KB, 667x761px) Image search: [Google]

75KB, 667x761px

>>61131607
The shit are you smoking? a the entire Geforce lineup gets squashed by $400 quadro in work, and FE is close to a P5000 >>61129462

1080ti isn't even on FE's radar

Guys, please stop.

You always do the same thing and shill for AMD, only to get fucked in the ass by them on release day.

>>61131687
Should mention it's $1500 instead of $1000 for the aircooled version.

>>61131693
Will RX Vega have this performance?

>>61131733
No, because why the fuck would someone buy the 3 times more expensive FE then.
RX won't be able to run pro drivers.

You people ask the stupidest fucking shit yes I mad

>>61131733
No, that is the workstation stuff. RX Vega is going to be priced accordingly for a gaming GPU. Meaning it won't cost anywhere near $1000.

>>61131693
Gaming is way more important than this though

>>61131766
When we're talking about a prosumer card, not really.

>>61131541
>Because "workstation" is loosely defined
Basically it's just a really fucking terrible excuse for the bad gaming performance.

>>61131687
>375w

No thanks

>>61131796
Or simple shitposting.
Now you can pretend this is a gamer card all you want, it won't stop it from doing its intended job, which it seems to have fulfilled.

>>61131796
>1080 is now bad gaming performance
>day one drivers will never get better
Ok.

>>61131857
1080 is a year old and a $550 card now, unless Vega is 1070 pricing it's DOA.

>>61131838
>it won't stop it from doing its intended job
Quadros perform well in games too, there's only so much difference between a workstation GPU and a consumer one, there is no reason why the consumer Vega will perform significantly better than the FE in games unless it's clocked at 1800MHz or higher. Driver optimizations don't make up for 30% of the game performance unless it's broken in literally every single video game out there.

>>61131908
Dude, do i look like I give a damn? It gives close to P5000 performance on day1 for $1000, that's it, everything else is a bonus.
I couldn't even begin to give a shit about RX Vega, let it burn

>>61131908
Driver optimizations should at least push it higher than 1080 levels, but not above 1080 ti.

>>61131510
P4000 is $900. Is Vega Fe outperforming that by a huge margin?

>>61131949
It's closer to a $2000 P5000, so yes.

File: Benchmark.png (302KB, 659x1621px)

302KB, 659x1621px

>>61131949

>>61131927
P5000 is $1900 and has considerably better FP64

>>61131979
All non GP100 silicon is 1/32 FP64 you fucking idiot, and another thing, Vega is 2:1 FP16, while again,, all non-GP100 silicon has laughable FP16, it's not even an option for GP102, much less 104.

You seem rather fucking clueless on this subject, go to bed.

>>61132004
You know Volta will change that.

>>61129815
Yes, many people who buy brand new flagship hardware are gamers, what's your point?

is AMD on suicide watch? are they bankrupt and finished yet?

>>61132030
lel

>>61132030
>muh big dies
>>61132036
They're again comparing it to a completely different product than the one it's actually competing against.

>>61132039
Yes, AMD will declare bankruptcy in exactly 10 minutes 39 seconds from now.

>>61132039
It looks like they're gonna be kicking ass for the next few years in the CPU market, notably in servers.
GPU's on the other hand seem to be heading the same direction as "great value", but I doubt miners gobbling up every single AMD GPU is hurting their profits.

>>61132039
They have too many partners lined up in servers to go bankrupt. Zen architecture is a win any way you look at it. I don't think they'll ever beat Nvidia again though, either in raw performance or in professional applications. Nvidia is way more aggressive than Intel.

>>61132036
it's not destinated for gaming you dumb shit

and AMD did a "gaming mode" even though it's not for gaming, just because retards like you think the most expensive thing is the best one for everything

>>61132174
Maybe if they build up some R&D dosh and put their CPU advantage to use, but that's years in advance.

Though if they can do the same they did with EPYC on GPUs, but I doubt it, not even Nvidia can fight off that much raw cheap silicon

>>61132174
Anon people were saying the same thing when faildozer was their only option for 5 years, even amidst the knowledge of Zen.

Now that AMD has the biggest CPU development hurdles taken care of they can afford to spread more R&D around to other departments. Maybe in another couple gens, heck maybe Navi really is a new "ground up" game changer like Zen.

>>61132501
My point was Nvidia isn't the same company as Intel. AMD caught Intel off guard by sandbagging Ryzen, and Intel had virtually abandoned desktop in focus of enterprise, which succeeded, and mobile, which failed. Nvidia isn't just waiting around for AMD though, they're aggressively promoting GPGPU, trying to create a walled garden of CUDA processing. They're not just taking on AMD, they're trying to take on Intel as well. Huang is an absolute madman trying to own literally every market space he can.

>>61129803
this
>>61129961
and
>>61129976
Are on the right track I believe. Someone tell this dipshit to set gpu fan to 100% and take the side panel off.

>>61132562
The case doesn't have a panel on it. It's either insufficient power or the blower style cooler just sucks.

>>61132588
I just saw the pic, shouldve finished the whole thread, I'm dumb.

Is the gpu at 100% fan? inb4 loud, just curious if this is really it. * grabs rope

>>61132609
Don't know, the guy running the tests was still pretty derp. Combination of day one drivers, crap setup could still be driving it down. Wait for reviews from actual sources, not some random guy who got the thing early.

>>61132187
>I-it's not for gaming
>I-ignore the gaming mode
>w-we just need more gaming drivers
>j-just wait
Just let it go guys
Wait for Navi

>>61132623
*sets down rope

Thanks man, I'll wait™ for reviews.

>>61129703
I have a feeling they are just going to release this shitstain to the market, say it's not for gaming despite it being for gaming, then wait a few months to let everything blow over, release the same cards, call them RX Vega, and say they are for gaming now that drivers have matured

Screen cap this

>>61132637
It's likely that RX Vega will set somewhere between 1080 and 1080 ti, so if you were hoping for it to BTFO the 1080 ti, might want to pick up that rope again.

File: AMD Vega Frontier Edition Live Bench Testing.webm (3MB, 1280x720px) Image search: [Google]

3MB, 1280x720px

Sup

>>61132762
Is that .. good? I'm retarded.

>>61132817
Nop.

>>61132817
In firestrike ultra that is about 400 points higher than my 1.48ghz 980ti

>>61132762
DELETE THIS

>>61132039
No but RTG is

>>61132636
how about, wait for RX Vega? Something AMD was telling everyone to do for a month now.

Hey anons, do you smell what I smell? I think it's a bit of a /v/ spillover.

>>61131355
Well fuck you too, the vega launch isn't supposed to be the second coming of christ, but it builds a new basis for awesome GPUs. So you'll only see the big numbers on generation later.

You delusional nvidiots should already kys because the generation after the vega launch will kill nvidia

>>61131978
>>61131693
>>61130574
The fuck? Is everyone too focused on /v/idya to not see this?

File: 7999ceead3c47f872878ffc408b76410.jpg (121KB, 800x800px) Image search: [Google]

121KB, 800x800px

>>61133276
Yeah like seriously
Fuck these people benchmarking GPUs with games

>>61133335
>Brand new GPU beating quadros from last year
sign me up

>>61133356
>While costing fraction of the cost
Wow, it's like you are using exactly the same Ryzen argument every Good Goy did with 6900x

File: 1445294603347.jpg (38KB, 450x495px) Image search: [Google]

38KB, 450x495px

>>61129703
>playing games with graphics cards

>>61133335
Unfortunately yes. It offers some pretty damn great price/performance ratio as a workstation card. But it is obviously shit because it can't game gud.

>>61133366
the fuck is a 6900x?
Congratulations
your 1000 dollar GPU is barely beating a 2000 dollar GPU from last year, that's how progression works

Interesting though how the p5000 and the p6000 is still better at gaming too despite also not being gaming cards

>>61133299
>WAIT NAVI

>>61133373
>>61133366
>>61133335
>ever buying a non radeon pro "workstation" card when everything supports CUDA
who the fuck is this being sold to honestly? Like who is this being marketed towards

>>61133384
6900x is Intels CPU that was direct competition of 1800x and guess what, new 8/16 intels CPU, turned out to be worse than 6900x

Congratulations, This is how progression works.

>>61133366
Except that $2000 Quadro is still beating Vega in any workstation related tasks.

Vega FE is so fucking bad that it has to compare itself to 1080 and Titan XP on AMD's official website when both of those cards aren't even workstation cards.

>>61133396
This card is aimed at game devs, people who do modelling and then want to quickly check how their models work, or maybe even solidworks users.

You know, visualisation, where CUDA is shit and was for the past 10 years.

>>61133401
Do you mean the 6900k? or the 6950x?
Cause the 7820x or whatever the fuck it's called beats the 6900k

>>61133414
Looks to me, like it's great value for performance it gives and people who would need workstation card would be happy to grab Vega.

1800x also isn't better than 6900x in every aspect, but compromise of price/performance makes it a better deal.

>>61133418
You guys realize that game devs are just going to buy the best cards available and don't give a shit about pricing because they are funded
Or do you mean the self proclaimed game devs on /v/

>>61133418
>maybe even solidworks users
Even p4000 shits on Vega FE in solidworks I think even p2000 outperforms it.

>>61133401
>>61133433
what is the 6900X and why do you keep mentioning it
I am looking online and I literally cannot find any information on it, I keep getting Sears faucets

>>61133396
Although I'm impressed by the performance, yeah, this launch also confuses the hell out of me.
It's not exactly aimed at gaming card, but it can do some gaymen, but it can also render stuff. And it's in a price range that's too cheap for big companies but fairly expensive for an individual.
Maybe it's aimed at small studios?

>>61133455
Pretty sure anon is talking about the 7900X

>>61133463
I think so, I'm literally not trying to start a flamewar I'm just baffled at the marketing of this thing
It can't really play games, and it can render stuff okie dokie but not as good as the type of cards actual companies buy. Maybe it's like an "At home" sort of card so you can take your work home with you? But still doesn't make munch sense because my job lets me take my PCs GPU out when I go home so I can get shit done at home too

>>61133427
7820x or evemn 7900x has one major flaw, L3 cache, performance is abysmal to say the least in software where L3 matters, when L3 doesn't it's beats everything on the market.

Problem is inconstancy performance that even lands it behind 1800x in some software and worst part is, there is no cure for it.

>>61133487
I wasn't even aware anyone is buying the new skylake and kabylake stuff
Nor do I know why people are bringing intel up in the conversations of GPU progression, Nvidia and intel are completely different companies, huang fucking hates intel and wants to steal their market

>>61132030
>Wait for Volta!
My sides.

>>61133508
Because arguments are too similar to ignore.

>>61133513
No they aren't, not even close, at all.
CPU and GPU progression are completely different, the fabrication and architectures are entirely different and so are intel and Nvidia

>>61133528
Yea, but that is not the argument, argument is that
>it's not better than higher priced products therefor it's useless
with is exactly the argument was used against Ryzen.

>>61133338
I would like to see someone test frontier with this http://pro.radeon.com/en-us/radeon-prorender-for-blender-and-solidworks-now-available/

>>61129703
Why the fuck did they compare it with Titan? Now everyone wants to gayme on it too, they should have compared to quadros and showed price/performance ratio. Are people on their marketing team retarded?

>>61133486
Hell, I think a lot of people who aren't solely interested in the gaming aspects would be as confused as you.
I don't know about the at home part though. It's not often you see people transporting desktops from work. But maybe someone who does work at home could benefit from this card.
It seems like a product aimed at a very sparse and diffuse set of buyers, and it does not make much sense to me, but hey I don't know about how the market works.

This meme card uses 300-375W
Are we going to be seeing similar numbers for the RX Vega cards?

>>61129703
>phonepost
>talk about gayming

lel

>>61133608
no because the FE is a work station card.
also workstation users don't care about power drawn. Just a reminder that Radeon pro duo version 1 2015 edition pulled 500W

File: 1445297301531.jpg (78KB, 816x840px) Image search: [Google]

78KB, 816x840px

>>61133906
>Kabylake X draws large amounts of power
>LMAO HOUSEFIRE YOU NEED NUCLEAR POWER TO KEEP THAT RUNNING LMAO
>Vega pulls large amounts of power
>I see no issue here workstation users don't care it doesn't matter

File: DOOM_1920.jpg (110KB, 548x486px) Image search: [Google]

110KB, 548x486px

>>61133537

>with is exactly the argument was used against Ryzen.

I primarily because places like /g/ (i.e /v/) and OCN all talk a big game how they only buy the tippest toppest model and anything below these halos products in shit when the reality is 1) these people don't and 2) just because a company doesn't have the aboslute best (or in this case, fastest) product doesn't mean their entire product stack is worthless.

That is what this vega stuff is all about - AMD havem ade it quite clear that for gaming RX vega is the product you want (and its performance is still unknown) but /g/ is taking the FE's gaming performance as gospel despite the numerous factors that can - and in fact do - lead to lower performance its gaming branded equivalent.

/g/ simply does not realise quite how much of a modern gpu's gaming performance is on the software side (as in, specific game profiles in the driver) that the FE doesn't use. Its why even on the Nvidia side without their game ready branded drivers performance is often shockingly bad until Nvidia go and hand tune a profile for any given game (pic related).

tl;dr /g/ is being retarded as is the norm.

>>61133920
>A CPU drawing 300W is fine

>>61133920

Implying said users are not pairing a Vega FE with Skylake X

>Implying kabylake-X sould exist.

>>61133936

When the inventor of the Intel circumsion tool is calling x299 shit you know something is up.

>>61133920
It's relative. Kabylake X draws substantially more power than its rival, and that's why it's housefires.

Vega pulls about the same amount of power as its rival.

>>61134006
It pulls 100 more watts than a 1080ti at a lower clockspeed

>>61134034
Vega also has a 945Mhz clock speed beating a 2000Mhz OC'D 1070 and getting really close to 1080ti

really gets the primal brain fired up.

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

36KB, 413x395px

>>61129703
Oh boy, another product hyped to shit by /g/ AMD fanboys that turns out to be crap.

It's your own fault you idiots, AMD already released several gaming benchmarks of Vega in Doom and Sniper 3, and all showed it around 1080 level, or slightly above.

>hurr no it will beat titan x just wait!

This is an exact fucking repeat of Fiji

What makes it more dissapointing this time however, is that AMD is a year late to the show and still can't beat Nvidia's top cards

>>61134090
>getting really close to 1080ti
It's not, it's much closer to the 1080 than it is to the 1080ti

>>61130128

Not him.
Not a gpu, but I already saw a cpu do this, on the desktop I traded my old laptop for. Granted, the guy had a combo of both a shit psu and shit board feeding a 4790, but point is it was still throttling. Gpu's otoh, not sure how that would affect them.

>>61130737
they do. it's gotten much better to the point where it's arguably ignore-able, but an AG layer will always diffuse light on either side to some degree, this not being a true representation of the underlying pixels on the panel. AG was bad enough on otherwise good monitors that people removed the coating themselves to get "glossy" versions of the best dell IPS monitors.

>>61130484
>what are 1440p high refresh rate screens
still $500-900 for the good ones, and the cheap ones are TN and/or have refresh rate limitations.

File: 1498186555225.jpg (168KB, 640x960px) Image search: [Google]

168KB, 640x960px

>>61130397
>The PSU "wastes heat like hell"
>It produces 300 watts of heat output with 50 watts of power input
>capture heat output
>use it to power GPU and generate excess
>???
AMD INVENTED A PERPETUAL MOTION MACHINE NVIDIA BTFO

File: 1483412012125.jpg (15KB, 213x237px) Image search: [Google]

15KB, 213x237px

>>61129703
AAHAHAHAHAHAHAHAHAHAHAHAH

>>61130418
Landing in between the 1080 and the 1080ti, a year later than Nvidia released their cards is kind of pathetic.

It shows that AMD can't compete anymore in the high end gpu segment

>>61134332
hey, as long as I can switch from my fury x to another amd product that performs better than a 1080 and is still cheaper - I will be happy

>>61134483
What makes you think it will be cheaper?

The die size is much bigger than the 1080's, and it has HBM2 which is way more expensive than GDDR5X

>>61129767
FE is a developers card for rx cards i assume given what i gathered from beyond3d.com

apparently the game mode is there to give a very basic understanding of how vega works and nothing more
no wonder why amd was deleting everything they are sandbagging really really hard

>>61134332
amd release a workstation card that sits in the high end gaming cards

ITS A FAILURE

amd later one release the actualy gaming cards
btfo nvidia

oh amd is DOA

>>61134657
It has all the streaming processors the gaming gpu will have, it has HBM2 and more of it than the gaming gpu will likely have.

The only difference with the gaming card will be that they might solder off some useless shit or disable it. Not add anything to it.

Also they might clock it a bit higher , but with these kinds of tdp I doubt there is much room.

So yes this is a good indication of what the 'gaming' card will be capable of.

And that means it will most likely not beat the 1080ti, but will beat the 1080. A year later, and at a significantly higher price.

>>61134689
they will solder off quite a lot
they will have another microcode
they will have a non ecc hbm2
they will obviously have drivers tailored to games
and obviously they will have a bigger room for oc and not being actively throttled down the engineers on beyond3d said they are expecting a maximum of 30% increase (assuming the drivers are top notch for it) on perfomance

>>61134713
HBM2 supports ECC by default and there's no OC headroom. RX Vega is DOA.

>>61134713
My point still stands, the actual performance won't be miles off from what we are seeing from the frontier edition, although it might improve somewhat.

That still puts it in between the 1080 and the 1080ti which is disappointing, considering it's a year late to the party and Volta is on its way

>>61134746
>Volta
You mean Pascal Refresh?

>>61134746
ye i gonna get a volta for sure, vega really is DOA.
I rooted for AMD with vega but anything short than being better than a 1080ti is just embarrassing

>>61134775
Pascal refresh on a 12nm node yes.

Your point being?

>>61134738
no shit sherlock ecc add quite a lot of interleaved lag so turning off the feature will help a bit more

>>61134783
It's "12nm" node.

>>61130128
>I've never encountered a card that hasn't even been released yet that downclocks itself, therefore it doesn't exist

These retards are allowed opinions

>>61134788
That's exactly what I wrote

>>61130102
Newsflash: they regressed performance from Polaris, or Vega is literally just a big die Polaris. Vega is dead on arrival.

>>61130267
Because of exactly what happened: everyone looked at it as evidence of rx Vegas performance. Even a toddler could have predicted this.

>>61130574
Amd fucked up keeping the same name for workstation and gaming.

>>61134713
But what about the 300w????????

AYYMD IS FINISHED & BANKRUPT

AYYMDPOORFAGS CONFIRMED ON SUICIDE WATCH

>>61134804
It's not an actual 12nm node. Just a refined 16nm FF+.

>>61133920
they didn't lie and tell us it's a 140w part

>>61130574
>performance worse than a cut-down GP104
>demolishes
AMDkeks are the funniest.

>all this screeching about 945mhz clock
isnt't that its memory clock?

So how long till the gaming version of this gets released to the public? Seems like we've been waiting forever

One thing is certain, heads will roll over at RTG.
Lowering the announced pricing right at release was already a red flag that something was off. To top it off they're not even controlling the reviews that come out, and instead we have to rely on random streamers to get every little bit of info. If this tells me anything, is that they knew it was a lost cause for quite some time, hell so far we're not even seeing ipc improvements from fiji.
After so much bark and little bite, it really is time for Raja to go, and admit hbm isn't ready for consumer products yet.

File: 1317726561959.jpg (11KB, 250x250px) Image search: [Google]

11KB, 250x250px

>>61129703
Wait what, a 550w psu isn't enough to power this thing?

The power draw must be gigantic

>>61135493
As long as it brings performance increases, it doesn't really matter, seeing as even a year later even Pascal is faster than Vega

I don't care about 1K GPU semi professional gayming performance, especially when there is not a single adequate benchmark. Im interested in Vega RX that designed for gaymers. I don't understand how can people be so retarded and stupid to make some conclusion on info we have so far. Must be very easy to live with such a tiny brain.

>>61135795
On AYYMD's website, they recommend a 850W PSU, you know it's inefficient as fuck

>>61135794
That's a Reddit post. Go away shittet.

>>61135811
>efficient
>GPU
Oh come on.

>>61135807
The gaming card will be the same as the workstation card with some shit soldered off, higher clocks and less HBM2

If you are expecting huge performance gains over this think again

>>61135837
I don't need to think or speculate. When there will be RX Vega and benchmarks I will make a conclusion whether it's shit or not. For now no one knows for certain.

>>61135837
There's no difference in performance between Pro and "gaming" drivers, which makes me think this is all horseshit

>>61135815

>Everything I don't like is reddit

Unlike you, I can see past the red tinted glasses and tell when a product from a company I support is a failure. Fuck off shill

>>61135869
AMD already showed demos of the Vega gaming card in Doom and Sniper 3

They were all in between the 1080 and 1080ti, this leak is right in line with that

This topic is gay as shit

Bottom line, we want the benches, we want the drivers, and we want the actual prices. Months away

>>61132762
how much is the frontier?

I already got my 1800x but i'm using an old-ass GTX card and i'm deciding between frontier/ an instinct card(assuming i would decide to game with the current one).

If it's less than 1000$ then it's so worth it, the bad thing is that it would probably cost like 400$ where i live because taxes n shit

>>61135881
It was Vega FE not Vega RX.

>>61135837

This, there won't be any miracle performance gains, I'm expecting 10% more at most.
Which means amd will have to rethink their strategy marketing the mainstream cards, namely portrait it as a 1080 competitor with a very aggressive pricing. Problem is this now way too late, either the price is incredibly low, or they had better made an incredible mining card to have them even get off the shelves these months.

>>61135948
No it was the Vega gaming card

>>61135968
Proof?

>>61135978
Proof that it wasn't?

>>61129815
>before amd's product launches
>IT WILL CRUSH X IN GAMES LOOK AT THESE TOTALLY LEGIT NUMBERS
>after amd's product launches
>B-BUT IT'S FOR WORK AND NOT FOR GAMING
how many times did this already happen?

>>61135985
It was Vega before there was a FE or RX and the FE came out first

>>61133414
Good job on being vague as fuck: "workstation related tasks"

Unless we're talking about CUDA,framework which AMD has already come up with an alternative(although not remotely close to being so mature) then i doubt a quadro card with the same price-point beats the FE.

That being said, FE isn't even that much destinated to the workloads which could be comparable to using CUDA cores(ML,scientific,rendering,etc);It's more destinated towards game-dev,which is way more "light" than the one mentioned before.(That's why probably it also has a game mode)

If you wanna compare apples with apples, compare Quadro(even Tesla but that's kinda different) with Instinct cards, those are destinated to the same tasks the quadro cards are

>>61135985
You were the first to say it was RX card. The burden of proof is on you.

>>61135991
Why would they show game performance on their supposedly workstation card?

>>61135991
The RX version is only a few weeks away, so that literally proves nothing.

They might well have been showing an early sample of RX, which makes sense seeing as they were showing gaming performance, in which it would obviously fair better than the FE

I don't give a fuck. There are no consumer cards and no benchmarks yet.

>>61135986
It's their own stupid fault.

AMD has been very clear about the gaming performance, both demo's they have shown pit it in between the 1080 and the 1080ti.

But ofcourse fanboys Always overhype the product.

>>61136013
>end of July
>few weeks

>>61136036
>source:my ass

>>61136045
source:AMD

ITT /v/ermin cross posters think Vega FE is the same shit as novidia's Founders Edition. LoL.

File: 123.png (36KB, 406x420px) Image search: [Google]

36KB, 406x420px

>>61135986
>>IT WILL CRUSH X IN GAMES LOOK AT THESE TOTALLY LEGIT NUMBERS
citation needed faggot, only people who got over-hyped claimed those things, way to stereotype so many people by judging few pajeets on /g/

It was fucking obvious that Intel would still beat AMD in games since the games today are favoring clock speeds and not the number of cores, just proves that 90% of the game studios are filled with indians who can't into parallelism

Again, R7 at least wasn't even marketed for SOLELY gaming.But the bad thing is that to attract customers most the tests they and the reviewers did were done with games instead of showing real-world multi-tasking/workloads power.

Go compile the linux kernel in 90 seconds or run 4 VMs with 7700k, see how that works for you

>>61136045
are you retarded? it launches at siggraph.

File: 1241231211.jpg (5KB, 117x159px) Image search: [Google]

5KB, 117x159px

>>61136036
>end of july
>not a few weeks

what happened to the live benchmark stream from the guy who got his vega yesterday

>>61136101
his house burned down

>>61136101
He got the power bill

>>61136101
He got heatstroke.

>>61136079

The thing is Ryzen is a legitimate good product. Even if it doesn't sport the ultimate single core performance, it is a fantastic jack of all trades. Vega on the other hand looks sketchy at best. What we have so far puts it little ahead of gp104, which is a disappointment for a supposed flagship with a chip of this size, coming this late. Add to that even the signals from amd suggest they don't hold much faith in it either.

Still, ignore him, he's just memeing

>>61136140
>>61136123
>>61136106
Samefag.

>>61136186
nope

>>61136186
nope

>>61136186
nope

still conveniently ignoring comparison within price brackets in workstation applications and dual purpose of this GPU

>>61136175
for $999 its amazing value though
the thing that beats it is $2000
right in the middle between q4000 and q5000

you all forget it's got ECC memory it's slower than gaming card by default

>>61136175
Of course it will do shitty in frameworks & software designed for CUDA/ ported for OpenCL at best.

AMD legitimately took a big, worthy risk by making their own alternative of CUDA,RocM, which is only like 1 year old.By making such interesting hardware(FE,though Instinct is the goods) they hope to attract more of this market to developing RocM since their hardware is cheaper for the horsepower they give.

It's kind of like the gaming market with the exclusive titles for AMD/Nvidia, but it's just for another industry(with the exception of FE which isn't exactly for AI/scientific/rendering/etc)

File: 1337045315355.png (164KB, 500x500px) Image search: [Google]

164KB, 500x500px

>>61129703
Had the misfortune of reading the reddit threads. Misinformation and poo-flinging everywhere.

I'm still waiting on real results form someone who's clearly not a retard with a potato setup.

I don't understand why everyone is so upset about it under-performing the 1080-ti Xp, It was known long ago that the Vega cards would possibly sit between the 1080 and 1080-ti. AMD seems to have delivered.

>It's a year behind and Volta is right around the corner.
I saw this being said so many times in those threads. Most people hold onto their cards for about 3 years anyway. If the price is right, most will be happy with the RX. If Volta really is all it's promised to be, it'll get picked up when the price drops sometime later.

>>61136261
>with the exception of FE which isn't exactly for AI/scientific/rendering/etc)

it is for rendering, I'd guess small studios will jump on it for content creation since it's twice cheaper than quadro 5000 and not much slower

instinct is for AI and teraflopping

>>61136287
>If Volta really is all it's promised to be
10% bigger dies? Wow.

>>61136287
>If Volta really is all it's promised to be
nothing good for now, they boasted giant die which is cool but raw power barely increased even pascal was more promising

>>61136255

And the thing that matches it is $500.
We still lack a lot of workstation benchmarks, yes, but after this show, Vega has a lot of ground to claw back to become a worthwhile alternative. And when with 1000€ in hand I can immediately grab a P4000, which beats the FE in the little WS results we have so far, Vega's value is deflating fast.

File: question.png (863KB, 1366x768px) Image search: [Google]

863KB, 1366x768px

can someone explain why vega's PCB is so much bigger than Fury's?

File: AMD-Vega-10-GPU-Radeon-Vega-Graphics-Card_12.jpg (375KB, 2560x1440px) Image search: [Google]

375KB, 2560x1440px

>>61136005
Does this look like the Vega RX?
Will the RX have a USB and various headers on the back of the card?
No it's a Frankensteined Vega chip and the first Vega chip to be produced was the FE which means it was either mostly an FE or something so beta its unlike the FE or RX

>>61136290
Well yeah rendering is both in content creation & for scientific purposes(obviously it's not only used in game-dev) so i guess i was referring more to the latter.

Instinct is probably more comparable to Tesla to be honest, and FE is more of an in-between

>>61136343
We don't know.

>>61136346
>Does this look like the Vega RX?
Since we haven't seen what a Vega RX looks like, the only right answer would be, we don't know

>>61136349
>Instinct is probably more comparable to Tesla to be honest

Instinct is amd's response to tesla

>>61136342
are you sure?

see >>61131978
see >>61131693
and >>61130574

File: 1475018515057.jpg (645KB, 1376x1729px) Image search: [Google]

645KB, 1376x1729px

>>61129703

how does vega pump those results with specs like this

drivers and workstation aside, still quite awkward

>>61136413
D R I V E R S
Also it still has 4triangles/clock geometry throughtput aka it's shit.

>>61136362
Seeing as Raja said it is not, it is not.

>>61136384

Haven't seen those yet, I stand corrected.
Still, what's the source of that? It's not something I'd recognize from a glance.

>>61136437
Yeah lets just resort to blatant lies why not

>>61136441
same source as those 3dmark benches
everything is half truths before someone does real benchmarking

File: Screenshot_2017-06-29-14-03-43.png (276KB, 1080x1920px) Image search: [Google]

276KB, 1080x1920px

>>61136384
You can still see p4000 beating Vega in solid works and creo and remember p4000 wasn't tested on the same hardware.

>>61136349
Instinct is the new Tesla competitor, yes, and unlike the previous FirePros, they run completely headless and are designed to be run for DL/ML/HPC/Virtualization in enterprise.
WX Pro series is their professional cards with viewports meant to segment the FirePro market into HPC and content creation, FE is more or less an early adopter Vega card to further bridge WX Pro and consumer Vega. Considering how late it is, I think the AMD software team are probably scrambling to get the drivers together for "Day 1" for RX Vega, hence why nothing happens when you switch from non-gaming to gaming mode.

Ok dude you got him. It was RX Vega with gaym ready drivers(tm). I think there is no point to release the Vega then. RIP in piece. Everyone go home. BTW I just ordered GTX1090Ti. Pic related is my order receip! :D

File: 1489534825515.png (172KB, 1424x320px) Image search: [Google]

172KB, 1424x320px

>>61136450
>picture from youtube
>from a video about Vega running games
>with Raja on hand talking about that card in that machine
>It's all lies

>>61136502
Then show it you snake, show the evidence where he says that card is the FE edition

sadly nobody ran any opencl benchmarks, but from the mining I think it's safe to say it's even shit at compute
how the hell can they mess something up so baldy
seem like it's worse than fiji clock for clock
I was expecting to get the wc version as a compute card, but even that appears to be shit.
at least it has fp16 so I might still consider it if they release some magical fixes

>>61132636
The fact that the gaming mode isn't the default one means it wasn't made for gaming, but rather a second thought.

>>61135213
Remember vega has a 2.5x improvement in watts/performance against polaris

>>61136634
LOL no

ITT: Nvidia gaymur xxxExtremetessulationMonsterEnergyDrinkxxx sperglord turboautists equate a workstation cards performance (which is the same as $4,000 quadros in workstation applications) in games with experimental drivers and on test benches running 550w psus and mitx boards to amd being doa...

>>61136644
>not being able to run a single gpu setup with a 550w powersupply

Good stuff AMD

>>61136593
>Basing OpenCL results on MUH COINS
Fuck off.
>>61136655
You know some vendor 1080 ti boards run at 275-300 watts, too. Would you want to pair one of those with a 550 watt PSU? Yeah, I didn't think so.

>>61136655
1080Ti recommends 600W

>>61136655
>1080 recommends 500W

>>61131437
Wow, they're gonna nickle-n-dime you on shipping a $1k video card?

>>61129781
it is faster than the titan in workstation workloads, it's crushing things like solidworks and autoCAD. Gaming wise it doesn't have the drivers for games, so games just get generic adapter performance. the FE is actually closer to a quaddro in terms of optimization and performance, and a quaddro does way worse than the same chip on a GTX when it comes to gaming.

File: a498e19ea036d0243de1b40697ca62b9.png (393KB, 448x567px) Image search: [Google]

393KB, 448x567px

Why is everyone so upset that there is going to be a gtx 1080 competitor that isn't bios and volt locked with gpu boost fuckery giving you only the illusion of being able to overclock?
Oh yeah and the competitor will let you use adapative sync tech with out paying the insane fucking g-sync price.

If it performs like the gtx 1080 but costs $100 more I'm still buying it because I'll save $2-300 on gsync and be able to actually control and overclock my card.
No, the 2150 pascal "overclock" isn't enough. It's exactly 10% on every single pascal card regardless of silicon quality. Pic related can do 2150 on a 1060 6gb. Unlocked 1080s have been able to reach 2300. Volta will be bigger pascal with gddr6 and more voltage so you can clock higher.

>>61137467
It's just shitposters. They flail their spastic tard arms every time something like this happens. They were never going to buy Vega anyway.

File: 64127b2065263d0b8133da67e819bcbf.jpg (562KB, 1920x1080px) Image search: [Google]

562KB, 1920x1080px

>>61137576
I get most people don't overclock like I do and so the nvidia thing with gpu boost 3.0 doesn't bother them but it SHOULD just on principle alone. There are two key videos on youtube showing how sinister it is.
It doesn't get in the way of LN2 overclocking or the normie who just google a guide and copy someone elses settings. It only gets in the way of people who use good cooling to try and get the most out of the GPU for 24/7 use, IE having a 1070 pushing a monitor that nvidia would rather you buy a 1080 for.
Those videos:
https://www.youtube.com/watch?v=bflLDenKirQ

https://www.youtube.com/watch?v=T1YCoTXBNJY

That totally aside, you don't overclock and if you do the fake 10% "overclocks" on keep you happy, freesync is a HUGE deal.
It adds almost nothing to the cost of the monitor and is such a great feature especially if you are a high refresh fag like me or even a 4k user.
Lots of game engines don't like running at 144 fps, you need insane ram and cpu to do it. So going from 144 fps to 60 fps is really jaring when it happenss, adaptive sync fixes this.
it also makes 4k gaming much more realistic because going from 45 to 60 regularly is tolerable.
If nvidia let me use freesync I would either keep my 980ti(a card you can actually OC) or maybe buy a 1080ti or volta.

>>61135025
Honestly I am still wait and see till the consumer version comes out.