When are we getting any information about this fucking thing? Jow Forums told me to wait for it instead of buying a 2070

That depends on which card you buy and which cooler it has on it, fucktard. Even cards that draw 300W+ like Vega 64 and the 2080 Ti can be near-silent and run cool if you pick a decent model.

videocardz.com/80720/sapphire-preparing-two-radeon-navi-models-for-july-7th-launch

>he doesn't know about navi

>Are you a nigger?
Are you?
>Not even Vega 7 is loud
Okay got it, you are a fucking shill, end of discussion. There has been shit tons of proof posted on this, yet you amd shills will deny this and if I am dealing with a person who thinks vega 7 is not loud, then I have no hope for what ever you have to say about navi not being loud.

>That depends on which card you buy and which cooler it has on it, fucktard.
Yes, but it requires more cooling, more power phases and other shit,which drives up the cost.
So either the price will go up by a lot or, the more likely soltuion is that they will put on the minimum amount of vrm phases, give just enough cooling from keeping it unbearable and keep the cost lower to compete with nvidia.

So a how architecture will always be either louder/hotter, or more expensive.

Yes it will be loud and hot cheapshit. It's GCN with GDDR6.

>While speaking to the Chinese press, Sapphire PR manager revealed that AMD is working on two Navi SKUs: a 399 USD variant to compete against GeForce RTX 2060 and 499 USD model as an alternative to RTX 2070.
LITERALLY same shit, months later, but with no ray tracing tech.
If this is true it sounds like a complete failure.

wait so it's a re re re brand of a fucking 480?

I WANT TO UPGRADE MY RX 580 ITS BEEN YEARS ALREADY AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHH

would you like to buy another 580 thats slightly overclocked? That will be 200$+tax

No. It's a cheap Vega with a configurable memory controller to run HBM or GDDR6. Vega had something similar only with GDDR5.

>Still on my 380
>Still runs everything well enough at 1080

But HBM is more power efficient than GDDR6, so it will consume more power and produce even more heat than the current vega?
Also the bandwidth on GDDR6 is lower than on HBM, so it's also less performance?
Just how much cheaper are they going to make it to compensate for what sounds like a worse performing vega that runs hotter?

>minimum amount of vrm phases, give just enough cooling from keeping it unbearable and keep the cost lower to compete with nvidia
You clearly don't understand amd then. Their reference designs have overkill vrms since the 7970.

ITS BEEN YEARS ALREADY
Two years to be exact.

Precisely, the only reason AMD would have done this and not undercut is if 7nm chip yields are coming in lower and keeping prices expensive. Sort of worries me for Zen 2

>overkill VRMS
It's adequate. Just because AIB partners strip it to the bare bone and sell it to you a bit cheaper doesn't make it acceptable.

my RX580 does well generally but I would like to have 100fps on ultra instead of 60-80. also sick of the fan noise
I got it as a stop-gap till something better. vega was a fucking mess of a launch

techpowerup.com/255768/sapphire-reps-leak-juicy-details-on-amd-radeon-navi

>It's GCN with GDDR6.
pretty much that. those shits won't even do anything to the Arch, won't they...fuck
not that i need a new card or anything but hell AMD, please deliver once again

Are you fucking retarded? It's on a 7nm process compared to Vega's 14nm. How the fuck do you end up thinking it will draw more power? Radeon VII draws the same amount of power as Vega for 30% more performance. Navi is on the same process, has architectural tweaks and is positioned below Radeon VII in the product stack. Of course it doesn't draw more power than Vega with those things considered, you fucking retard. The switch from HBM to GDDR6 is completely negligible in terms of overall power consumption.

AMD are making a big mistake if they're trying to match Nvidia's pricing.

>How the fuck do you end up thinking it will draw more power?

Same way they managed to make 14nm Vega 64 eat more peak power (410w) than the 28nm R9 390 (365W)

>It's on a 7nm process compared to Vega's 14nm
I was talking about the HBM being more power efficient than GDDR, which is why AMD used it in the first place.
And if they clock their new 7nm GPUs it will consume the same amount of power as the current ones. Which they will, look at Radeon 7.
So you have GPU chips consuming the same amount of power and generating same heat, except memory this time around is less efficient and adds to more heat and power consumption.
Sounds like it will consume more than the last vega.
>How the fuck do you end up thinking it will draw more power
Read above amd fanboy.

Indeed I was expecting them to undercut them by 100-200$ easy. If they are just going to match them in price, I have literally zero reason to go for navi and will get a nvidia RTX every time, as it's more efficient, has raytracing and is priced similar at the same performance.

DELETE THIS!

>it doesn't draw more power than Vega
Show me the numbers magic man

The 390 only has 40 compute units and lower clocks

he has faith in AMD, that is enough proof you heathen.

500 USD for 2070 competitor? Isn't thaT v64 for around the same price? What is even the point? AMD just wants to increase margins, numbers be dammed?

if you haven't noticed, AMD have been re-releasing the 480, 580, 590 which is all the same card basically.
So now they will re release vega and call it a day again.

You do realize that gddr6 draws 5x more power than hbm2 for the same bandwidth right? We're talking 20w Vs 100w at the same bandwidth.

We already know it will be a replacement for polaris and will deliver performance somewhere between a vega 56 and 64 at somewhere around rx590 pricing.

>be a replacement for polaris
then why does it have vega pricing

zen 2 confirmed 70% rates. zen 1 was ~80%

580 and 590 doesn't pretend like it is new arch.

>vega pricing

Where the fuck did you get that idea? The navi we're getting this year will essentially be a cheaper vega with better power efficiency.

The big vega that is set to compete with the 2080ti will be next year.

because rtg is retarded, nvidia has retarded prices so they must too
they could've undercut the hell out of them and still have 2-4% margins.

It pretends to be a different card, but it's essentially a factory overclock.
The rebranding is strong both in the same arch and from arch to arc.
Ironic if you think about it, this is like what intel does and everybody shits on intel for it.
But the moment AMD does this with GPUS nobody bats an eye, literally "it's okay when amd does it"

It actually has more bandwidth then the 590s GDDR5 while being significantly cheaper then HBM. If you want HBM you'll have to grab an instinct card or a Radeon Pro.

>>Where the fuck did you get that idea?
From this new news

tweaktown.com/news/66003/sapphire-plans-2-radeon-navi-cards-399-499-july-7/index.html

>news is directly from a PR manager for SAPPHIRE during a chat with Chinese press, with the rep saying AMD is working on two Navi SKUs: the first is going to cost $399 while the higher-end Navi will cost $499. The $399 card will reportedly compete with NVIDIA's GeForce RTX 2060 while the $499 model will fight with the RTX 2070

nvidia rebranded maxwell 3 times already same way.

>work on a "Toxic" version of Navi is complete, and it is watercooled
Sold.

>starting with 399$ Navi, source: Sapphire PR manager
>Polaris replacement
nigga wut. i do know that Navi was supposed to be polaris replacement but AMD is going for the ~2070 crowd but will ultimately fail if the 'lower end' 399$ model won't win against that by a good margin. you know that won't ever happen

That doesn't make any fucking sense you can already buy 2060s and Vega 56 for less than $400.

Literally no one would buy the small Navi at that price.

And I really wanted to change back to AMD this year. will keep 1070 I guess.

I mean the only way what that PR manager said makes sense is if the $399 navi has 2070 performance and the $499 navi has 2080 performance.

That's obviously not going to happen so the only explanation is that the PR dude is a retard and misunderstood something.

>Vega 56 for less than $300
ftfy. that is just awesome value but require tweaks for best performance which is fine with me but normies are too stupid if it isn't just plug&play.
too fucking bad i had to buy a 1070 cause mining craze ate up all Vega cards everywhere

>I mean the only way what that PR manager said makes sense is if the $399 navi has 2070 performance and the $499 navi has 2080 performance.
yes that would make very much sense but AMD already has the Vega7 for 2080 performance and the PR stooge actually mentions 2060 & 2070 performance. which is a fucking downer m8

Then the price he is quoting is obviously wrong. They won't release a card with 2060 performance that is more expensive than a 2060.

R7 is failed MI50 stock they made small money off and appeased investors with. They can, in theory, release R7 killer. But they would have to be competent for that.

>Then the price he is quoting is obviously wrong.
> They won't release a card with 2060 performance that is more expensive than a 2060.
hahahah, literal first step is denial my friend

You can buy new Vega 56 Pulses (pretty much the best 56) for €270-299 now, so that's the card to get.

With Chill on I never go above 170w power draw, 55°C

I'll wait and see. I stand by my prediction that it will be polaris replacement. If not and it does turn out to be a more expensive 2060 i'll just get a vega 56.

what's your frame limit at?

>but AMD already has the Vega7 for 2080 performance
AMD honestly should've marketed it as "workstation card that can do gaming" and put greater emphasis on massive 16GB of HBM2 and FP64 performance.
That way they would separate markets. Plus as said AMD barely makes any money with R7, so they wouldn't have any problem changing something thy don't make money with something they do.

>buying pulse
That has
>cut down the VRM phases by HALF
>uses NANO vega PCB
>only 4 heatpipes
The only, ONLY good thing about it is that it uses samsung memory, but sacrificing everything else for it is just not worth it.

Attached: vega 56 saphire.jpg (1682x642, 376K)

40-60, which is the freesync range of my monitor - LG 24UD58.
Usually play at 1440p, 2160 or somewhere in between if there is a resolution scale, depending of course on how demanding the game is and what it takes to stay in the 40-60 range.
On AC Origins I cap at 45 because my CPU can't keep up with Denuvo (2200g while waiting ™). If I let go of the controller GPU power draw will go as low as 45w.
Dota 2 at 2160 with everything maxed doesn't draw more than 100w (still capped at 60fps). The game draws power from my soul thought

>listening to Jow Forums

You obviously have no idea what you're talking about. Try owning one first

fair enough, though the reason you're getting good temps and power draw is because you are heavily underutilising that card.

son, you DO know what everyone is doing with their AMD cards for ages? lower voltages while the clocks remain the same, literally cutting the power draw by a huge margin depending on the card. you don't really OC a Vega for example at all, maybe a mild +20-+50Mhz on top if the card manages to do that at a lower voltage. heat and power draw is way more important

Attached: 1552493286350.png (765x624, 171K)

>$399
>$499

AYYMD IS FINISHED & BANKRUPT

AYYMDPOORFAGS CONFIRMED ON SUICIDE WATCH

TOLD YOU DUMBFAGS 7NM IS EXPENSIVE

Forgot to mention my P7 state is undervolted to 0.950mV, going down by 0.005 for each lower state.

I agree that I may be under-utilizing the card, but I'm still gaming at 1440p and above at the frames that matter to me, with very good temps and power usage. With this level of undervolt it still clocks a bit above 1500mhz while not reaching 70°C if I turn Chill off and uncap. Power tends to average 180-200w then.

These temps are in a sff Raijintek Ophion, so I guess they would be better in something like a Meshify

>you obviously don't even own it
>you obviously don't know what you are even doing, it's the users fault
>you obviously got a unlucky gpu

compare this to pic related

>son, you DO know what everyone is doing with their AMD cards for ages?
Yes, undervolt heavily, then limit the max FPS to keep the heat down some more, and then using radeon chill that reduces your FPS some more to keep the card cool.
It's truly pathetic how much performance is being thrown out just to.
It's like that time when intel said that to avoid overheating their cpus you shouldn't overclock them and everybody laughted at them.
Now amd gpus come with optional gimping and people thank them for it because it makes their hot and loud cards less hot and loud at the cost of performance.

Attached: vega 56 reference and strix on bottom.jpg (1836x1376, 680K)

Ah..yes..the performance advantages of empty space..

>optional
eh yes

...

>performance advantages of having more VRM phases, spread over a bigger part of the PCB
>not only is less phases less efficient so every one of them runs hotter, but they also concentrate the heat in a smaller space
>they are also located close to the capacitors which when overheated degrade very fast and violently explode
>who cares about annnny of this I can FPS cap my games to 60fps and undervolt while praying for silicon lottery to save me

God you god damn retards.
I bet you would recommend a milled aluminum block for 2060 with less than 5 phases as well.

Yeah, some amd fanboys actually said they should force it on a hardware level to make cards less hot.
Imagine that, actually suggesting that cards be gimped to get rid of the hot and loud stigma which they rightfully hold right now.

>How the fuck do you end up thinking it will draw more power?
Because GCN is already at its limits and squeezed for all it's worth, not much else to do than go for the moar clocks, which WILL ramp consumption up, even with a smaller node.
At least this is supposed to be GCN's last iteration. AMD just cannot keep going with it, it was good when it came out, but they need something new fast to keep up.

>click source
>404 chink runes
Do you really think a "product manager and PR director" would just casually leak this info?

Attached: Capture.png (923x358, 28K)

Why GCN is bad at its gaming performance?

But can it be that 6 very high quality, overengineered vrms beat 8-10 medium quality ones? Genuine question

>he only, ONLY good thing about it
It's also quieter than the 2060 FE while performing the same for over $50 cheaper.

You're a dumb shill.

dunno i remember meeting a guy working in 3d tech saying that it can't do basic things that nvidia can since 10 years.

Wow... are you 12?

isn't it ironic that the logo is a partially rendered cube

it can't draw polygons as fast as nvidia

Drivers.

i'm 26

Attached: 988541.jpg (660x262, 32K)

>amd gpus in 2019
the only good combo is nvidia & zen 2

>But can it be that 6 very high quality, overengineered vrms beat 8-10 medium quality ones? Genuine question
Not really. Unless you are talking about top quality VRM components vs bottom tier trash.
You have to keep in mind it's not just being more efficient and producing less waste heat, it's also spreading out the head over more components and a larger area that's easier to cool.

>It's also quieter than the 2060 FE while performing the same for over $50 cheaper.
>no proof or sources posted
>calls others shills while trying to damage control downsides of a specific AMD card

and whos wet dreams are these?

Vega56 was/is pretty based

it's not, 10 year old card runs ps4 and it's fidelity is pretty high
if all games could be optimized on PC for each variation AMD would be x5 faster than nvidia, that's why they tried vulkan push

nvidiots HOMOSEXUALS crying all time, all days

the dream. would buy 3090 in an instant if it was true.

30-45% improvement for the same price is not outlandish. It's what you would expect.

This is why all these rumors are bullshit. Too easy to just make up some numbers.

I just hope it's a good upgrade from my 980ti, but I don't mind waiting for whatever comes after gcn. I hope they figure out how to mcm gpu dies. Sadly I prob need to change to Ryzen. Dunno how much my 5820k is affected by all those vulnerabilities etc.

>980ti
how are you holding up man?