This kills nvidia

>this kills nvidia

Other urls found in this thread:

nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-third-quarter-fiscal-2018
devblogs.nvidia.com/parallelforall/inside-volta/
twitter.com/NSFWRedditGif

I wish
it won't though

>half the board is voltage regulators

lol @ OP

It's a development board user. The final one will end up in a laptop little bigger than Ultrabooks

is that the new raspberry pi?
I bet that will hold way more dust than my last one

Really good perf for a laptop part if this is true.

ffffffuuuaark I want one in my laptop

why i think that intel will use avx on gpu's now?

oh my boner is getting ready if this happens using avx on a compute monster such as vega will totally kill nvidia tho

nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-third-quarter-fiscal-2018

>NVIDIA today reported record revenue for the third quarter ended October 29, 2017, of $2.64 billion, up 32 percent from $2.00 billion a year earlier, and up 18 percent from $2.23 billion in the previous quarter, with growth across all its platforms.

THIS KILLS THE FAGGOT OP

so its a 570 ?
thats weird

PLZ ESPLAIN WHAT THIS IS MISTER

a prototype of the new intel and amd gpu GLUED together

Oh i was hoping it was like a raspberry pi sized board I could stick in a old laptop

well i can imagine it coming out on SFF form also at some point

>small form factor form
(sorry I had to)

But at what tdp? One does not simply put an 150W APU (100W Polaris10 + 50w i5) into some small enclosure with shit-tier cooling.

What the fuck?
Do you know what AVX is?

Can't wait to see how this will perform with 7nm lithography

no user please tell me inform us that a vectoring ISA cant be run on gpu's if intel decide to..

GPUs are already running VEEEEEEEEERY wide SIMDs.

point proven

It won't get any hotter or powerlhungry from running wide SIMDs.
It's already doing it.

i never said it will run hot

thus why they probably gonna make some gpu's to be able to run them

its not like they are gaining anything atm with phi

You are suggesting them to run an x86 ISA extension on GPUs.
It makes no sense.

So you just add the processing power of each CPU?
3+3=6 Ghz? Im not familiar with clusters thats why Im asking

That's the theory, but in practice is very hard to make programs that use 100% of the power of all the CPUs at once.

you probably forgot what was the original plan of HSA
also avx is just vectors nothing more atm very few use them thus why its not going to hurt anyone if they use a bigger portion of the gpu market to do it instead of relying on a failed gpu that became a compute monster that serves only avx

>you probably forgot what was the original plan of HSA
Unified memory with intermidiate layer?
Also plese format your posts properly.

>kills nvidia
It absolutely totally kills nvidia in the laptop market, it will kill their profit margin there for sure.

Not many people are aware that prices on MXM module NVidia cards are around twice - yes, 2x - what they charge for similar desktop parts. That is why this new Intel+AMD partnership is a huge game-changer and it's probably why Intel went this route.

Nvidia has pretty much had a monopoly on laptop GPU parts for quite a while, prices indicate that they are abusing that position somewhat.

Keep this in mind when you compare the performance of these new AMD/Intel parts to Nvidia parts: They absolutely crush nvidia's offerings on price/performance, everything nvidia has in that market is totally blown out of the water.

Past performance does not say that much about future performance. Nvidia will remain very profitable and still own quite a few market segments but the small form factor and laptop segments will take a big hit for sure.

Keep on dreaming, this Poolaris garbage is not power efficient and does not support Feature Level 12_1, so it won't run future games that require Feature Level 12_1 support, retards that buy this garbage will have massive buyer's remorse

Volta is 50% more power efficient than Pascal and will crush this garbage in performance and power efficiency

devblogs.nvidia.com/parallelforall/inside-volta/

>The new Volta SM is 50% more energy efficient than the previous generation Pascal design, enabling major boosts in FP32 and FP64 performance in the same power envelope.

Look, the nVidia's damage control brigade has arrived.

A 570 in a ultrabook that will undercut every laptop with a 1060 in size

>that
>ultrabook
nigga you high

I thought the biggest problem was going to be to equalize CPU usage between both processors, but thats better

If you clock it low enough for a 15W (GPU well under 600MHz) it could fit.

Just published their quarterly figures and record sales number. OP BTFO.

Why bother?
Just use Raven Ridge.
A bubble is still a bubble.

Maybe Intel wants actually working 3D drivers and features with their shit instead of their joke iGPUs that can only do 2D

>Posting jewish bingo results

Irrelevant.

No, why would OEMs bother with this pricey SKUs when AMD will sell RRs for peanuts should they even want to?
Like, WHY?

o wow it places right between the two worst nvidia cards of the generation
good jahb
it's really funny the head poo of amd went to intel to produce this

Meanwhile Nvidiots 90W from the GPU alone just to run at that performance :)
While the whole AMD + Intel package is under 50W

>he cares about 2 cents of electrcitiy

I hate GPUs REEEEEEEEEEE

This isn't desktop hardware you fucking manchild nigger, at most it could end up in SOME desktops if it gets put a mATX/ITX mobo.

He doesn't.
Laptop OEMs do.

Poojet is mad

>says something retarded
>gets called out on it
>"umad"

That's not how it works you favela monkey, get back to Sup Forums

You seem to have forgot to actually read what you replied to before spewing out a ton of shill fag-talk.

The fact remains that Nvidia MXM laptop and small form-factor parts are priced around twice their desktop counterparts.

Direct3D 12.1 doesn't matter anyway, it's all about Vulkan games and those who do use Direct3D develop for XBox and PC at the same time which means they won't be using it.

And Volta won't change the fact that nvidia's laptop parts are horribly overpriced. They will have to bring the price of those parts down to desktop-card levels if they want to compete. In case this is too advanced for you: A price/performance ratio involved both Price and Performance. If a slightly faster part costs a OEM three times as much as another part that's slightly slower and uses a bit more power then the choice is pretty obvious.

>Volta

Imaginary consumer hardware in mid to late 2018? Might as well WaitTM for Navi at that stage

>Imaginary consumer hardware in mid to late 2018?
Assuming it even yields and nVidia is finally ready to sacrifice the margins.
Besides Vega will be unfucked by then.

I don't really care if electricity costs 2 or 10 cents somewhere when I'm using a laptop and there's not power outlet nearby. How long the battery lasts is a far more important concern...

I also care about the TDP on my desktop and my HTPC, noise does matter. A 65W TDP CPU with a 140W TDP rated heatsink is awesome in that respect and something that really is entirely different from a 95W TDP CPU with some small man-let cooler.

REEEEE YOU NIGGER FORGOT TO CUT OUT THE BLUE SKY BEHIND HIS HEAD REEEEEEEEEEEEEE

>he didn't watch the announcement vid
Lmao@u

You know they have no problem stuffing this in a ultrabook that usually has a 15W chip, right? It's called firmware and throttling.

15W chip that can pull up over 40W because Intel's definition of TDP is different for each market and changes every year.

They've already stated this will use H series chips, which are 35-45W. This is not aimed at ultra books at all, it's aimed at thin and light desktop replacements (Dell XPS, MBP).

It's almost a certainty that Apple pressured Intel into this behind closed doors. They needed a better solution than Iris and refuse to deal with Nvidia, so they told Intel to do this or they drop Intel altogether and port their entire Mac lineup to ARM instead (something that's a lot closer to reality than you may think). AMD is just happy to take the money, they don't even have to do much retooling as it's using the same semi-custom IP as the console GPU's.

This is likely the 14" market, but I wouldn't be surprised seeing a 13" with around 1.5-1.7kg with this.

Meant 14" and up

spergo

This thread provided enough shitposts for a decade

because this is much faster than RR on the GPU side. RR is targeting light gamers. This part is targeting gamers who want a laptop they can game on, but don't want to pay for a laptop that has an expensive MXM card. This technology may also enable Steam boxes to become a reality, since a Rocket League machine for your living room should be much cheaper than previously. And remember, these machines are going to be as fast as the consoles are. That means these laptops will play all the console titles just fine and for a price comparable to the consoles. That's the average gamer's wet dream.

>Putting AMD GPUs in laptops
Hasn't the RX470M thought you anything...

>6 phase + 5+1 vrms
are they trying to power an oced 7700k and a 1080?

why so many VRMs?

this isn't an apu moron. Its a 45-55w part. did you not read the press release?
Raven ridge is almost half that