He's fired, right?

He's fired, right?

Other urls found in this thread:

youtu.be/N_1F7UrI0PA
youtube.com/watch?time_continue=1&v=QuFG7pw_k50
twitter.com/SFWRedditVideos

POO

He will whine about AMD not being able to attract top engineering talent. He will cry about Radeon supporting AMD to develop Ryzen.

He needs to be shown the door.

Jesus. Don't you shills got better things to do? It's a fucking leak of a workstation card. How about we all wait until the gaming lineup is released before we start shitposting

>Sup Forums
>not shitposting
Choose one.

...

HOW CAN SHE SLAP!?

>have a pajeet as director of your whole gpu division
>wonder he produces shit

L M A O

If it's priced a penny over 499, It's done. Just like his career.

ayyy

It will. HBM2 is expensive. $600 is the most generous price you can expect.

Predicting $400-$450. 1080's already dropped down to $500.

With drivers it will be slightly faster than a 1080.

expect 1070 level performance

after 2 years lmao

As an AMD shareholder I want his ass gone. I don't give a dam if the silicon isn't as good as it should be. I don't expect miracles right now from RTG with the budget they have. BUT this launch has been bungled HARD. The marketing is confusing and misplaced, and I feel I've been misled by RTG management about what the card was supposed to do and when.
I do not trust this Pajeet shitter to bring Navi to market successfully.

It already does that with no drivers.

There is already plenty of proof it's at least as fast as a 1080, this shitposting is just turning into some sort of bizzare misinformation campaign from shills.

no there isn't

Yup. He has repeatedly overpromised, and under delivered with ridiculous benchmarks, and marketing material.

The drivers must be fucked, even that LN2 Fury X at 1600 mhz and 1 ghz hbm was near 1080ti, and fiji is just 2 tongas glued together.

show me the benchmarks

Even the videocardz chart shows it's reaching 1080 levels. With crap day one drivers.

Or the architecture is absolute garbage.

>2015 HBM gen 1
>Massive supply shortages and limited to 4GB, makes the fury GPUs more expensive than they should be
>2017 HBM gen 2
>Massive supply shortages, makes the cards more expensive than they should be again

Was HBM a mistake? Nvidia has been rolling along just fine with GDDR5X.

thats the frontier edition

RX Vega is supposed to perform even better than that, though.

HBM is fine, there are now several HBM makers, and Samsung has even figured out how to make a cheaper, less powerful HBM.

What was Nvidias last big innovation in the gaming space?

Tegras that ended up dumped in Nintendo Switch.

you can't trust this raja guy, obviously

A rebranded shield is not innovation.

Something that the industry has adopted and been used widely by the industry.

It was a joke. Thought "dumped" would give it away.

I already know from the start that that
Mr. Rajeeeeeeeeeeeeet is up to no good

Yes they should've launched a GDDR air cooled version and then the liquid cooled hbm later. They could've been picking binned chips while optimizing drivers and came in with a Vega and Vega "X" versions and >>gamers miners would already have cards in their hands.

Why would he be fired? Radeon is literally selling every single gpu they make all over the world. If was an actual shareholder I would be thrilled.

Miners the world over are giddy.

Serious nd honest question, who exactly "made" Vega? And Polaris?
How much work is credited on Raja?

What's exactly wrong with 1080 performance? Who and when exactly started shilling about Titan and Ti killers? 1080 performance speculations were right since the Doom benchmark and those firestrike benchmarks + this one guy's stream benching Vega FE in games confirm 1080 performance. It's all down to pricing. Even AdoredTV said that shit over a year ago - math is there and you cannot expect a 1080 Ti performance.

375watt with 1080 performance, 2year late

I honestly don't understand AMD shills. This was fucking obvious for years now. But you do have a point. 1080 is much more efficient and I can only imagine Volta stomping Vega even more in efficiency to performance rankings.

The aircooled model is 300 watts, and it's only 2 months later than AMD intended to release it.
NVidia lacks a hardware scheduler, that's how they went from housefire Thermi to efficient Kepler. They completely removed it.

>Workstation class
Not a fire pro
No a gaming card

Is just a shit

why is his hair gray but mustache black? does he dye it or something?

AMD roadmap from 2015 stated that Vega will be released in late 2016

In 2016 Summer (Polaris Launch) they said "You need to wait a bit more for our high end cards"

In 2016 Q4 they stated that Vega will arrive 1H of 2017

The "gaming card" will arrive mid Q3


But wait... Theres more.

They Live demoed Vega in 2016 December 12 at the new Horizon Event

And Again at CES

And again at Ryzen Launch.

But it sure worth the wait

Feces

>crap day one drivers.
This meme, this shitty meme over there.

>hardware scheduler
Avx 512 in cpu world

>Drivers for synthetic benchmarks.

Only because Nvidia can't.

Then when you see it working, it makes everything else look like shit.

Clearly AMD carried enough about games to put a gaming mode in drivers for their workstation card.

Hello

All of it, on both.

Any one saying otherwise is on AMDs payroll.

...
I heard this before

youtu.be/N_1F7UrI0PA

3000 behind 1080... Wait what?

Don't trust that faggot. He has no idea what he is doing.

Delusion: the post

He has a shitty system, and doesn't know what tge fuck drivers he is running. It is still fluctuating clock speeds even with the right drivers, and there is no gaming bios or software updates at all for anything he runs.

Or you can keep jerking off your green teeny weeny like a juvenile faggot, makes no difference to me.

>1800x
>shitty system

Fine, it's another random ass internet fag.

>What was Nvidias last big innovation in the gaming space?
Unified shaders? Wait, ATi was first to unified shaders.

...

>It is still fluctuating clock speeds
He try to OC it before...

>Ryzen 1800x is a shitty system
Hello Jewtel

>and doesn't know what tge fuck drivers he is running
He used the latest driver from AMD website

>Fine, it's another random ass internet fag

Stop your damage control

youtube.com/watch?time_continue=1&v=QuFG7pw_k50

Fuck off

I don't give a shit what he runs, he's some know nothing faggot that doesn't know what he is doing.

Thanks for proving my point. He doesn't know what bios he's using, and he's basically trying to crowdsource how to benchmark his computer.

I want it to be good, but even I know day 1 AMD launches mean fucking nothing.

Seriously, what the hell?

What do you want to do?

>Talks about 1080
>Links 1080 Ti and Fury X

Your point?

The problem with vega right now is that it is supposed to roll out a much more refined tile rendering mechanism. One of the elements that helps give nvidia bandwidth efficiency and power usage efficiency over amd. Here we have vega finally rolling out amd's attempt, and it seems just as hot, just as power hungry, and seemingly no more efficient than polaris. That's the real concern. It suggests there's something fundamentally limiting in the arch preventing similar competition with nvidia, not just a lack of feature set. Or a terrible implementation, which calls into question the engineering.

HBM is definitely the way forward but you wont see its true benefits until Navi

The biggest advantage (once production ramps up and costs go down) will be the greater power efficiency of HBM which will be most pronounced in Laptops and Consoles. Consoles have a very restricted power budget so if you can scale HBM at low clocks low heat output you got memory bandwidth that matches or beats GDDR5x and GDDR6 with less of the console power budget going to memory allowing you to push voltages and clocks on the GPU core higher (or put a bigger GPU in at the same clocks)

This will eventually combine with Navi architecture to create an infinitely scaling and ultra-cheap/+/power-efficient graphics solution that can then combine with the new easily scaled Ryzen CPU architecture allowing AMD a greater diversity of applications than Nvidia.

Take the polaris 10 GPU as example. Refine it and upgrade the CU's for better efficieny and clockspeeds then shrink it to 7nm to boost efficiency/speeds even further. Then take the infinity fabric concepts developed from Ryzen and connect four of these tiny ultra efficient GPU dies together like the CCX's on Ryzen. You now have a 10,000 shader behemoth with 32gb of HBM2+ and 2+Tb/s of raw bandwidth. Its matched with a refined Ryzen2 8-core CPU on 7nm. You have made all this fit into the power budget of a Console and can afford to sell it for 500-600$

With the New API's Games developed on console now port seamlessly to PC and run perfectly on AMD desktop hardware. The raw power of the console is so huge that PC gaming is superfluous. Yes the desktop will always be faster but after a certain point it becomes moot.

With Ryzen and Navi AMD will be able to connect and scale tiny dies of Memory(HBM), Graphics(Navi) and Central Processors(Ryzen) together in a seamless mesh fabric (infinity and modified infinity solutions) and scale that shit to infinity and beyond.

Vega is the new Graphics design. Navi is Vega scaled to infinity.

>just wait
yawn

There are both, 1080 and 1080Ti.
Really this shows that 375W is amazingly in spec

Vega is not working. Well, at least the way it was supposed to.

>Vega is not working. Well, at least the way it was supposed to.

Its on par with Quadros that cost 2x more what the fuck do you mean its not working?

>What was Nvidias last big innovation in the gaming space?
Innovation is pointless if it's crap. Nvidia are doing the smart choice and sticking with technology that is a proven performer.

It still handles 4triangles/clock.

>It still handles 4triangles/clock.
proofs?

Providing top of the line performance, which is what consumers want, and being untouched by competition for way over a year. But you do keep clinging on to your meme memory and see where does that take you.

...

Polaris was already in the pipeline when Raja was hired, Vega is the first design completely born under Raja's leadership

Thank you, happy merchant.

Its working exactly the way its supposed to. Vega was never intented to compate Huge die against huge die with nvidia pascal/volta on the desktop.

The entire purpose of the Vega architecture is to redesigned every part of GCN down the shaders to maximize efficiency.

AMD is building a single design that is used in both professional markets (servers/workstations) and consumer markets (consoles/PC) because they cannot afford to make two designs optimized for each specific market.

Nvidia builds two separate architectures optimized for each segment of the market. Their gaming design (maxwell) which is literally what Pascal and Volta are. They are just tweaked and refined Maxwell on better processes. Dont get me wrong it works exactly like its supposed to for gaming so why would they change it?

AMD has no interest in compete in the high end gaming PC market with nvidia. Its fucking pointless. Even if they did build an ober GPU that beat Nvidia in high end gaming the nvidia fanboys would still buy team green whichis exactly what happened the last time AMD build a GPU that destroyed the godawful peice of shit that was Nvidia Fermi. Fanboys still bought fermi anyway even though it lost in every single metric and was more expensive. AMD cant win PC faggots over which is why they have stopped trying.

This architecture will be good overall because the baseline for the quality of PC games is limited by the hardware capabilities of consoles. So if AMD who makes Console SOC's can now cram more raw power into them on the cheap then the baseline for games will go up which will improve quality of all games across the board including on PC.

I expect Volta to smash vega simply because its gonna be pascal with a price cut and a new massive 5400-cuda core die on the top end.

Vega is the design. Navi is that design scaled to infinity.

Vega was all about pushing pixels. Its bad at pushing pixels. The design is a complete failure.

>never mean to compete
>just wait for navi now
>infinity performance

Can you stop your infinity shilling?

What I don't get is why they halved the memory bus width from 4096 bits to 2048 bits and made the memory bandwidth less compared to the Fury.

Like how can they be hitting any limits with HBM2 to the point that AMD would design Vega like this?

Two stacks versus four stacks.

Dumb fucks. This was never meant to be a high end desktop gaming GPU.

Its literally just for servers and Consoles. Thats it.

also the sales will be good simply due to crypto-coin mining.

They had 480/580's selling for 500$ each because of those miners. You think AMD gives a fuck if you faggots got a better gaming gpu than the gtx 1060 for 200$?

>Sup Forums
>video games

HBM1 on Fury had stability issues and the actually bandwidth wasn't and big as advertised.

Also Bandwidth is less of an issue due to advances in compression.

Finally that larger bus would mean putting 4 HBM packages on die instead of two which means higher cost and more power consumption.

This HBM tech is still going through its growing pains whereas GDDR memory has been around for a long time and is simply being refined.

The advantage of HBM and the reason AMD is banking on it is that eventually down the road the cost to make it and the power efficiency it has over GDDR will yield a major advantage for the next 10-15 years.

You forgot to change the y-axis title. Wouldn't want your graphs representing something silly like infrastructure development!

Did he really?

D E S I G N A T E D
S H I T T I N G
S T R E E T S

DESIGNATED

HBM2 doesn't seem like a big power advantage over HBM1 or GDDR5/X seeing as it seems like it's still hitting thermal limits and throttling according to early benchmarks. I expecting that some credible experienced tech reviewer will take some time but will be able to to post benchmarks on this later that I can trust more than some random smucks on the internet. PCper did say they ordered one so will probably wait for them to get their take on the card.

And no one cares

Or the shitty 500w psu wasn't enough

OH FUCK MINERS aretalking about buy it.

NO.... DETELE this FUCKING NOW.

It was explicitly meant to be a high-end desktop GPU.

This guy here is the actual Shill. It's in nvidias best interest for AMD to keep raja. He is fucking up every one of AMD's gpu products.

SOMEONE STOP THEM

>non stop shilling about Intel's TDP on their cpu's
>AMD releases a card that a 550w psu might not even be able to support
>TDP is no longer an issue anymore