So what's the verdict on R9 290X ?

[Shmee]

Well, I don't see what the big deal is with the noise...just use a headset

Though it seems the reference cooler isn't so good, that should be fixed shortly. I can't wait for OC results on good cooling, on both 290 and 290x, as well as 780 Ti

 

[Enigmoid]

Are you talking about TPU's statement (or the single benchmark that UaVaj included)?

If so, go read their review. I quoted their words above.

The nordic hardware review.

I'm simply saying that in their own results they are using 42 fps as the R290X score while showing that really under sustained gaming load (looks like 15 min+) it only gets around 38-39 is unusual and shows that all their benchmarks are using pre-throttle clocks.

I think Hawaii is a great piece of engineering but am not enthused about the reference R290X. I think a lot of review sites are not looking at post throttle performance (about 5-12% less) as it takes much longer to test (and many sites don't have the time or money to do so) and so would look at the R290X benchmarks as a picture of the best case scenario. I expect aftermarket coolers to fix this though.

I like AMD I just feel that while they (especially the GPU division) produce great hardware they are lacking in vision and marketing. Nvidia is often overpriced and the titan and 780 deliver unimpressive gains vs the 680 and 7970GE.

 

[HisDivineOrder]

The beauty of AMD's going with the cheapest cooler they could (basically the same one as the 7970) is they know most of the reviews, especially the early reviews, will be done on open air benched systems where the airflow is not even an issue. It won't be until users start actually trying to install them in their real systems with cases that close where they might find their airflow insufficient.

This is why test systems should be built around cards in actual cases. Otherwise, companies are going to start building solutions that work fine in open air (even if loud), but fail to reach said performance levels as time passes inside a case. In particular, this card is going to run into this as time goes on and the heat levels rise due to the sheer temperature of the card. The very reason heat will transfer better at high temperatures compared to lower temperatures with the same cooler (a point in the 290X review at anandtech) is the very reason more heat will be transferring directly into your case through the card and the exterior of the cooler. That's the OTHER reason that nVidia and AMD went with lower operating temperatures. They weren't just concerned with what their own ASIC could handle, but also the heat that would passively transfer into your system over time, raising the overall temperature inside the case, which has the end result of raising the overall temperature of the card once your case's ability to flush out the heat runs dry. That high temperature is going to have detrimental effects across the board for a lot of systems if they don't go full water imo. I'm not sure a lot of people have systems up to the task of running it CF.

This was also a problem whenever 7990 CF was tested, a fact that Tomshardware eventually detected and detailed. One card would always get outrageously hot while the other would be at normal operating temperatures, but this was a lot less likely to be detected by a reviewer doing a normal benchmark run outside a case than it would be for an OEM to find in a system they were building.

And I always find it curious how these review boards never seem to have coil whine and yet a great many of the boards users get do. I wish reviewers were doing a better job of figuring out which boards are likely to do it rather than just ignore the issue as if it doesn't exist when it can make a card louder than a loud fan if it's particularly bad.

 

[3DVagabond]

The beauty of AMD's going with the cheapest cooler they could (basically the same one as the 7970) is they know most of the reviews, especially the early reviews, will be done on open air benched systems where the airflow is not even an issue. It won't be until users start actually trying to install them in their real systems with cases that close where they might find their airflow insufficient.

This is why test systems should be built around cards in actual cases. Otherwise, companies are going to start building solutions that work fine in open air (even if loud), but fail to reach said performance levels as time passes inside the system. This was also a problem whenever 7990 CF was tested, a fact that Tomshardware eventually detected and detailed.

And I always find it curious how these review boards never seem to have coil whine and yet a great many of the boards users get do. I wish reviewers were doing a better job of figuring out which boards are likely to do it rather than just ignore the issue as if it doesn't exist when it can make a card louder than a loud fan if it's particularly bad.

That's what hardware.fr does, actually run them in a case. They also use thermal imaging rather than rely on the cards thermal sensors. The pain of course is you have to use Google translate if you don't understand French, and their English site seems to almost be abandoned.

 

[MrK6]

The beauty of AMD's going with the cheapest cooler they could (basically the same one as the 7970) is they know most of the reviews, especially the early reviews, will be done on open air benched systems where the airflow is not even an issue. It won't be until users start actually trying to install them in their real systems with cases that close where they might find their airflow insufficient.

This is why test systems should be built around cards in actual cases. Otherwise, companies are going to start building solutions that work fine in open air (even if loud), but fail to reach said performance levels as time passes inside a case. In particular, this card is going to run into this as time goes on and the heat levels rise due to the sheer temperature of the card. The very reason heat will transfer better at high temperatures compared to lower temperatures with the same cooler (a point in the 290X review at anandtech) is the very reason more heat will be transferring directly into your case through the card and the exterior of the cooler. That's the OTHER reason that nVidia and AMD went with lower operating temperatures. They weren't just concerned with what their own ASIC could handle, but also the heat that would passively transfer into your system over time, raising the overall temperature inside the case, which has the end result of raising the overall temperature of the card once your case's ability to flush out the heat runs dry. That high temperature is going to have detrimental effects across the board for a lot of systems if they don't go full water imo. I'm not sure a lot of people have systems up to the task of running it CF.

The reverse is true too. In systems with great airflow design, the cards will actually run cooler than in open air and definitely within a cramped case. Where do you draw the line? I think having an open air is the fairest way to compare cards, since you're trying to eliminate the environment from the equation, and therefore solely comparing the card, which is the point of the review.

Also, all of these cards throttle if they're running too hot. If a user can't design a system with adequate airflow, well, that's their problem, not the manufacturer's.

And I always find it curious how these review boards never seem to have coil whine and yet a great many of the boards users get do. I wish reviewers were doing a better job of figuring out which boards are likely to do it rather than just ignore the issue as if it doesn't exist when it can make a card louder than a loud fan if it's particularly bad.

They probably don't. Quality hardware and electricity can make all the difference in coil whine. Also, it's harder to hear in open air.

 

[OCGuy]

Unless there is something specific you like about it, it is overpriced. Paying 50 bucks more than the competition and also needing to buy a cooler isn't the best option.

 

[Silverforce11]

Well, regardless of all the fanboyism on this forum, NV just concluded the R290X's verdict: It's a success, they are dropping prices on their high end.

 

[MrK6]

Well, regardless of all the fanboyism on this forum, NV just concluded the R290X's verdict: It's a success, they are dropping prices on their high end.

That's a great point. It shows both companies know price/performance is king.

 

[Will Robinson]

Unless there is something specific you like about it, it is overpriced. Paying 50 bucks more than the competition and also needing to buy a cooler isn't the best option.

Probably being faster than a GTX 780 is specific enough.

 

[Will Robinson]

Well, regardless of all the fanboyism on this forum, NV just concluded the R290X's verdict: It's a success, they are dropping prices on their high end.

+1 Voted by AMD's main competitor.:thumbsup:

The R9 290 release is really going to set the cat amongst the pigeons.
People in the know are saying its a very strong performer.

 

[SolMiester]

deleted

 

[Piklar]

The verdict is that AMD have pulled NVidias pants down in terms of price/performance and now we as consumers are in a far better situation in regards to acquiring seriously powerful graphics than we were two weeks ago.. reference cooler , noise/heat/throttling issues aside AMD are the Heroes of the month imo :thumbsup:

 

[Tempered81]

Unless there is something specific you like about it, it is overpriced.

Yah the slower $1000 nvidia card is a bargain. Unless there is something specific you like about money, get the Titan.

 

[UaVaj]

780 is now $550 + 3 games.
http://www.newegg.com/Product/Produc...82E16814130917
http://www.newegg.com/Product/Produc...82E16814130918

:whiste:

 

[tviceman]

780 is now $550 + 3 games.
http://www.newegg.com/Product/Produc...82E16814130917
http://www.newegg.com/Product/Produc...82E16814130918

:whiste:

$509 with open air cooler, http://www.newegg.com/Product/Produc...82E16814125471
$499 with good blower http://www.newegg.com/Product/Produc...82E16814127740

 

[UaVaj]

$509 with open air cooler, http://www.newegg.com/Product/Produc...82E16814125471
$499 with good blower http://www.newegg.com/Product/Produc...82E16814127740

so much for that 30 seconds of fame.

 

[Silverforce11]

so much for that 30 seconds of fame.

It's the ultimate compliment, your competitor drops prices massively to compete.

Clearly now the 780 is a better deal! Lets see what AMD does and what their AIBs come up with soon.

 

[Gikaseixas]

The verdict is that AMD have pulled NVidias pants down in terms of price/performance and now we as consumers are in a far better situation in regards to acquiring seriously powerful graphics than we were two weeks ago.. reference cooler , noise/heat/throttling issues aside AMD are the Heroes of the month imo :thumbsup:

^ ^
this 1000 times

 

[Granseth]

so much for that 30 seconds of fame.

Thats all you need. Now the reviews are out there and will be used when people are buying a new GPU.

And AMD GPUs usually gets more price cuts than nVidia GPUs, so probably by christmas the 290x will be cheaper or the same as 780.

 

[f1sherman]

so much for that 30 seconds of fame.

Well it's tough to remain the top dog when you are burning 30% more watts against a bigger and older chip

Yes AMD did put a pressure to NVIDIA, and this is it's biggest success.
But the pressure is mostly towards NV margins.
From engineering standpoint, AMD still trails on 28nm because ancient GK110 remains more balanced chip, due to having healthy OC potential vs AMDs 290X which already broke 250W wall.

AMD could have played Gigahertz card vs smaller and overstretched GK104.
Versus bigger, purposely tamed GK110 this is not a physically viable solution as there is simply no TDP room to grow.
So no, "Gigahertz Edition" wont be happening anytime soon, 290X is already "GE".

To see just how much is GK110 ahead... an ancient GK110 does not even need another revision, which more likely than not will be released due to GK110 being more than 18 months old.
All NVIDIA needs vs 290X, is overclocked GTX 780, and then fully maxed GK110 for those $1000 that they had learned to love.

So much for all those complaining about TDP discussions.
Everything revolves and comes back to TDP. TDP is KING!
It's just that AMD did not know how to utilize this fully in those days when they had had the lead.

 

[Rvenger]

TDP is king now eh? What will it be when AMD wins over TDP?

GK110 isn't ancient and plus 290x is faster than Titan. We will see how much power the 780Ti consumes when its released.

 

[MeldarthX]

Well it's tough to remain the top dog when you are burning 30% more watts against a bigger and older chip

Yes AMD did put a pressure to NVIDIA, and this is it's biggest success.
But the pressure is mostly towards NV margins.
From engineering standpoint, AMD still trails on 28nm because ancient GK110 remains more balanced chip, due to having healthy OC potential vs AMDs 290X which already broke 250W wall.

AMD could have played Gigahertz card vs smaller and overstretched GK104.
Versus bigger, purposely tamed GK110 this is not a physically viable solution as there is simply no TDP room to grow.
So no, "Gigahertz Edition" wont be happening anytime soon, 290X is already "GE".

To see just how much is GK110 ahead... an ancient GK110 does not even need another revision, which more likely than not will be released due to GK110 being more than 18 months old.
All NVIDIA needs vs 290X, is overclocked GTX 780, and then fully maxed GK110 for those $1000 that they had learned to love.

So much for all those complaining about TDP discussions.
Everything revolves and comes back to TDP. TDP is KING!
It's just that AMD did not know how to utilize this fully in those days when they had had the lead.

OC forums - Gibbo has proven it 290X has room to oc and is a very good chip along with all the actual owners of the card.

Is the ref cooler crap? yea in many ways; but it does its job - Once people figure out how to tweak this card everyone's come back to say its a hell of a card and we're just starting to see what it can do.

With more robust cooler; we very well could see a GE version; as it looks like standard op for card is 800mhz and it boost to 1 ghz....better cooling and settings it runs easily at 1200.......its damn fast; I believe its only going to get faster....

http://forums.overclockers.co.uk/showthread.php?t=18551717

 

[MrK6]

Well it's tough to remain the top dog when you are burning 30% more watts against a bigger and older chip

I'm quite honestly not sure how some of you passed elementary school: http://www.techpowerup.com/reviews/AMD/R9_290X/25.html
It's 13.8%, heck round up to 15% so it makes it easier for you to work with. Far cry from 30%. Also note that the performance/watt metric is even closer (~0-10% depending on resolution) since the 290X is also faster than the 780 or Titan. Not sure what point you're attempting to make here.

Yes AMD did put a pressure to NVIDIA, and this is it's biggest success.
But the pressure is mostly towards NV margins.
From engineering standpoint, AMD still trails on 28nm because ancient GK110 remains more balanced chip, due to having healthy OC potential vs AMDs 290X which already broke 250W wall.

What is this "250W wall" you speak, please explain to us why you feel it is significant.

AMD could have played Gigahertz card vs smaller and overstretched GK104.
Versus bigger, purposely tamed GK110 this is not a physically viable solution as there is simply no TDP room to grow.
So no, "Gigahertz Edition" wont be happening anytime soon, 290X is already "GE".

To see just how much is GK110 ahead... an ancient GK110 does not even need another revision, which more likely than not will be released due to GK110 being more than 18 months old.
All NVIDIA needs vs 290X, is overclocked GTX 780, and then fully maxed GK110 for those $1000 that they had learned to love.

Little early to call that when we don't even have aftermarket solutions for the 290X out yet. Also, since the 290X is already a good deal faster than the 780 and many times Titan, I'm not sure where AMD is really concerned. Nvidia had to undercut them on price and with gaming bundles because they knew they couldn't compete with performance. If the 780 Ti is still a cut-down GK110 that can finally compete with the 290X, anybody with half a brain will realize it's because they pushed the clocks. There was nothing special about it when AMD did it with the 7970 GE, and there will be nothing special about it if nvidia does it with the 780 Ti. Where enthusiasts here, not kids with too much money.

So much for all those complaining about TDP discussions.
Everything revolves and comes back to TDP. TDP is KING!
It's just that AMD did not know how to utilize this fully in those days when they had had the lead.

Again, explain how this has any relevance to us, as so far it just seems like you don't understand basic physics in the least.

 

[R0H1T]

Little early to call that when we don't even have aftermarket solutions for the 290X out yet. Also, since the 290X is already a good deal faster than the 780 and many times Titan, I'm not sure where AMD is really concerned. Nvidia had to undercut them on price and with gaming bundles because they knew they couldn't compete with performance. If the 780 Ti is still a cut-down GK110 that can finally compete with the 290X, anybody with half a brain will realize it's because they pushed the clocks. There was nothing special about it when AMD did it with the 7970 GE, and there will be nothing special about it if nvidia does it with the 780 Ti. Where enthusiasts here, not kids with too much money.

Don't count on that, I bet there's still a bunch of kiddies here with a ton of money to burn on the next titan err 780 Ti

But the thing that'll literally light a fire under Nvidia's arse is the 290 at 400~450$
Radeon R9 290 Probable Pricing Surfaces

 

[omeds]

I think you guys are missing fishermans point. It's like 7970 vs 680 all over again, but in reverse.