• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD to Slash Radeon RX 5700 "Navi" Series Prices Ahead of Launch: $399 & $349

They don't have a choice unless they want most of their cards to languish on the shelves for months and then be forced to discount them even more. That said, I still think prices will drop lower than this before the end of the year.

That didn't stop Nvidia. Their last quarter report has their sales turnaround at 140 days. A 3 day improvement over the last quarter and over 120% of its normal. Their payment window is also up 50%.
 
Nice to see some competition on price. Waiting for reviews to confirm if there will be competition on performance.
 
something I don't understand....
whay this ? and whay now ?
I meen: AMD could have made RDNA on "14nm++++++++" right ? it would be mutch cheeper no ? yields would be high price would be very low becuase even Global Foundaries have it right ? so whay not make RDNA on 14+++++ at 1000mm2 ? it would be (rough estimate) around as effective as a "7nm" 750mm no ? sutch a thing would wipe the florr with nvidea I think :)
nvidea didn't cared that mutch about temperature when they made the fermi (know as thermi)
 
Dear AMD, please DROP prices on already reasonably priced 5700 and 5700XT, so that whole market can revert to the times before GPU cryptomining.

IF we can get a card, that is a only a bit below RTX 2070 Super, but priced 20% lower, that is a really good deal.

Can AMD make sure it will sustain hungry miners if that comes to happen? last time nvidia cards were the only ones left and most nvidia cards were sold out too.
 
something I don't understand....
whay this ? and whay now ?
I meen: AMD could have made RDNA on "14nm++++++++" right ? it would be mutch cheeper no ? yields would be high price would be very low becuase even Global Foundaries have it right ? so whay not make RDNA on 14+++++ at 1000mm2 ? it would be (rough estimate) around as effective as a "7nm" 750mm no ? sutch a thing would wipe the florr with nvidea I think :)

Global Foundries were part of early "real men have fabs" days. That boat sailed a long time ago for AMD.

AMD need 7nm to be competitive at the bare minimum, it's certainly never going to let them wipe the floor with Nvidia, it just helps them keep up at the moment.
 
something I don't understand....
whay this ? and whay now ?
I meen: AMD could have made RDNA on "14nm++++++++" right ? it would be mutch cheeper no ? yields would be high price would be very low becuase even Global Foundaries have it right ? so whay not make RDNA on 14+++++ at 1000mm2 ? it would be (rough estimate) around as effective as a "7nm" 750mm no ? sutch a thing would wipe the florr with nvidea I think :)
nvidea didn't cared that mutch about temperature when they made the fermi (know as thermi)
AMD is so far behind in efficiency they need every bit of help they can get, even if the gains are not huge.
 
Yes, I know Titan came way before, maybe I didn't express myself clearly enough - two weeks later was refering to 780Ti only. But I wouldn't say AMD kept Nvidia on their toes by 2012 at all, since the lackluster performance of the 7970 (only 15% better than more than a year old gtx 580) enabled Nvidia to tackle it with 104 class silicon instead of the top of the line 110. Essentially they could have released what later came to be 780Ti in early 2012 as the 680, but they decided to withhold it and start working on Maxwell early while AMD was forced to keep releasing the best they could come up with right away. This way the green team started to always be one step in front, having an immediate answer to anything AMD launched right away and with years this one step became two and by now probably three. Without some major cataclysmic event on their end I don't expect AMD to ever properly catch up again even if it might temporarily seem that they are closing in.

Yes, the 680 upset the balance by being cheaper and faster than the 7970, and then the 290x upset the balance by being cheaper and faster than the Titan or 780, so Nvidia had to release the 780Ti and Titan Black cards.

And that's not even mentioning dual GPU monsters like the 7990 or the 295X, which remained as the fastest single cards throughout this whole period of time. (Please don't mention the Titan Z, I think we can all agree that abomination of a card should have never been released, especially at $3K}

Look, we can go back and forth grasping at straws about how things went back then, but that's entirely besides the point of my original post, the point being that at least AMD was gunning for the flagship cards back then, and in some cases beating Nvidia and forcing them to innovate and also lower prices.

Look at what we have now, the legacy of Raja, always releasing mid range cards and never really going for the top, and what has that brought us?

Nvidia sat at the top of the stack for almost two years with Pascal, and from the looks of it, history is about to repeat itself with Turing, and Ampere is set continue that dominance, but at what price?

Nothing AMD has done in the past few years has approached the level of competition they had at the top of their game, and so we have to deal with $1200 flagship cards that don't really need any updates in both performance or pricing cuz nothing compares to them. And yes, I know Titan broke the $1K limit years ago, but that's an entirely different market segment.

Can't you agree that a more agresive AMD, one that was actually fighting for the top performance spot, would be a good thing for all of us?

Let's not go in circular arguments about cards that actually competed against each other on equal footing, at least back then we had that, now we don't even have any competition at all at the top performance level...
 
Last edited:
Of course it would be a good thing but it just won't happen, at least not by AMD. Nvidia's years of strategic planning, solid execution and perhaps even more importantly, learning from their mistakes, have brought them into a basically untouchable position where noone can realistically hope to tackle them at the top spot with any measurable success. I mean, let's just look at Navi; yes, in theory AMD could try and produce a "monster" size die that might be able to take on the 2080Ti, but with probably horrible yields, warehouse-heater-like consumption and virtually no profit margin, while still lacking ray traycing, resulting in dismal sales and further reduction of income for the already beleaguered RTG.
 
Of course it would be a good thing but it just won't happen, at least not by AMD. Nvidia's years of strategic planning, solid execution and perhaps even more importantly, learning from their mistakes, have brought them into a basically untouchable position where noone can realistically hope to tackle them at the top spot with any measurable success. I mean, let's just look at Navi; yes, in theory AMD could try and produce a "monster" size die that might be able to take on the 2080Ti, but with probably horrible yields, warehouse-heater-like consumption and virtually no profit margin, while still lacking ray traycing, resulting in dismal sales and further reduction of income for the already beleaguered RTG.

These strategic shifts do take a lot of time to mature, especially for companies with a long history in the business. They have a long chain of distributors and parties to cater to and those have expectations. And chip design just doesn't come quickly or cheaply. I am still missing a long term outlook for AMD that really makes sense, it looks a lot like they will just trail the high end a generation or so for the near future. Beyond that its anyone's guess, really.
 
the problem is by the time yields on "7nm" is high nvidea will also make 7-cards, nvidea know that 7 is expencive and yields bad so they don't wanne waste on it now, especialy that they don't need to, so they do the same thing as they did with the 2000 on 14: waith until it ready and then make cards on it

I think this says something interesting: for AMD to prove the RDNA is good, they will have to comparred to same-size nvidea, but that don't exist yet.
how can we compare this 250mm to the 2000 cards ? or should it be compared to 1660 ti ? or just by number of transistors ? do you think RDNA can beat turing 1:1 ?

Nvidia's years of strategic planning, solid execution and perhaps even more importantly, learning from their mistakes, have brought them into a basically untouchable position where noone can realistically hope to tackle them at the top spot with any measurable success
realy ?
tackle.png


yhe sure no one can tackle them, except us :)
 
Last edited:
Looks like its official

AMD Radeon (Twitter)

D-vYu89W4AAl2xw.jpg

I was about to post this, good i found you posted it. So it's official. Good mive by amd, the super rtx 2060 review I stated my opinion that amd should have kept the $349 and after a day or so amd did it ehhe, here is what I said.

The super rtx 2060 would be the best choice if Nvidia had increased its price to $350 or left at $340, $400 was a bad move. Looking at the competition, the rx 5700 at $380 is supposed to be 10% faster than the rtx 2060, which would fall in line x super rtx 2060, the super rtx 2060 will in theory at moment be 3% faster than the rx 5700 and will cost 6% more money.

On the other side, the super rtx 2070 at $499 is supposed to be 10% faster than the rx 5700xt and cost 12% more money.

What I see here is, AMD x Nvidia has never been so close in price performance market segmentation like this in years. It will be your choice to decide what to get, both are pretty much equally priced and matched.

Nvidia will have to cut prices too and that will be placing the super rtx 2060 to where the rtx 2060 belonged to $349 or so.
 
I think it is more likely AMD pushed Nvidia to release the super series with there current lineup. So Super is out, and AMD can get some positive news when they lower the prices just before release. Remember, a lot of people were not happy with the AMD prices when they were first shown.

On the latest Gamer Nexus video, it was said that the Super Cards were expected be released after the AMD GPU launches.

Just some interesting strategy playing behind the scenes, which is good for the customer :)

On the 7nm Process for AMD. AMD historically was faster to transition to a newer node, with all the risks and benefits involved. Nvidia usually was slower to transition, because they are very good at optimizing on the current process node. Both strategies work. Both come with advantages and disadvantage.
 
I never really understood why AMD is always mandated to put a lower price on their cards compared to the equivalent products from Nvidia.
We noticed!
 
Actually the RX5700 seem to be performing very close to the RTX 2060 Super, which would be great for $50 less. The $100 cheaper RX5700 XT also seems a much better price/performance option. But I agree, a 1 or 2 game bundle would be so much better than a 3 month Game Pass.
I meant all AMD hardware bundle: Ryzen 3xxxx + RX 5700 + (maybe) some mobo.

No, bundles do not make up for higher MSRP's. If you want AMD to go out of business in GPU's, keep arguing for them not to beat the price of the competition that's ahead in the gimmicks and close enough to the same in performance to make the gimmicks matter again. AMD's the guy coming late to the party. Of course, he's got to lose some money to get back in the groove...
That is exactly the point.
AMD can reduce price for people actually buying AMD hardware, without engaging in price wars .
 
Last edited:
With these new prices, Navi's are the better buy if you don't want ray tracing (to tank your performance for negligible graphical enhancement).
 
the problem is by the time yields on "7nm" is high nvidea will also make 7-cards, nvidea know that 7 is expencive and yields bad so they don't wanne waste on it now, especialy that they don't need to, so they do the same thing as they did with the 2000 on 14: waith until it ready and then make cards on it

I think this says something interesting: for AMD to prove the RDNA is good, they will have to comparred to same-size nvidea, but that don't exist yet.
how can we compare this 250mm to the 2000 cards ? or should it be compared to 1660 ti ? or just by number of transistors ? do you think RDNA can beat turing 1:1 ?


realy ?
View attachment 126271

yhe sure no one can tackle them, except us :)

You did miss a good part of that graph though

I'm more of a fan of the big picture and perspective

126292


Recovery is setting in and a new trend upwards can be detected. This also nicely shows that the stock price was in desperate need of correction, look at how it compares to just a few years ago.
 
Question - I haven't ran AMD cards for probably 6-8 years, so I am curious - do all AMD cards support multi gpu these days? I haven't kept up on their architecture with regard to multi gpu support.

Thanks.
Yes they Cross Fire but it tends to bring on frustration in gaming - it really doesn't pay to fool with it. Sorry to see it fade been doing Crossfire since the ATI 850s even had a triple Fury X.
 
I never really understood why AMD is always mandated to put a lower price on their cards compared to the equivalent products from Nvidia.

All things equal people would just buy Nvidia anyway. To justify their purchase they'd dredge up an excuse from the bottom of a barrel of silly reasons. E.g. AMD has terrible drivers, SLI is better though no intention to do so but wants the option available just in case. Insert gimmick here.

You did miss a good part of that graph though

I'm more of a fan of the big picture and perspective

View attachment 126292

Recovery is setting in and a new trend upwards can be detected. This also nicely shows that the stock price was in desperate need of correction, look at how it compares to just a few years ago.

S&P is like that as a whole.
 
All things equal people would just buy Nvidia anyway. To justify their purchase they'd dredge up an excuse from the bottom of a barrel of silly reasons. E.g. AMD has terrible drivers, SLI is better though no intention to do so but wants the option available just in case. Insert gimmick here.



S&P is like that as a whole.
if not that (seriously, not that) more power consumption.... more noise out of non aib cards.... no RT capability (for what little that is worth.. it is worth something).

Be fair! ;)
 
Back
Top