GTX 590 reviews

[RSC]

He probably took off points for the card dying. It's a lot harder to win a decathlon if you drop dead during one of the events.

Did the card broke @ stock? Or did he overvolt it like hell, even after he got a very good OC (815MHz) @ 1.0v?

I think you need to read that review once again, more carefully and apply a bit of logic too (to the "all resolutions" part).

He did the test, on his conditions, at the resolutions he chose. This was the final result. End of story.

[XRL8]

Like the unlucky coincidende of his review showing the GTX590 being faster (all resolutions), cooler, quieter, more OCable (22% vs 10% of the 6990) and consuming less power and still it got a final score of... 7.0, versus the 9.1 of the 6990. Oh, and then he blew the card up with 1.2v and... stock clocks.

I, like God, do not play with dice and do not believe in coincidende.

Um.Average of all resolutions show GTX590 winning.Not in all resolutions.At the resolution that is kind of point for this kind of a card.It loses.And in most popular one ATM ,its a tie.So its not winning in all resolutions, but in 1024x up to 1650x.
Performance per watt ,there is only one card thats worse.GTX 480.HD6990 is better.
It (the GTX 590) consumes more power than the 6990 at both idle ,gaming and peak.In furmark its hard clamped by nvidia so it does whooping 4 watts better.But way nvidia handles this clamp.Its no where near real life scenario.
So again, no.
It is Quieter, still loud.
As for temps, in TPU`s review, there isnt any chart with temperatures.However you do understand i hope that if it draws more power, it has (as an entire card, not just GF110 chips) to exhaust this somehow.More power=More heat.
All of above taken from TPU`s article.
So yes.It IS QUIETER.On the rest, youre wrong.

[RSC]

Um.Average of all resolutions show GTX590 winning.Not in all resolutions.At the resolution that is kind of point for this kind of a card.It loses.And in most popular one ATM ,its a tie.So its not winning in all resolutions, but in 1024x up to 1650x.

That would be a very valid point if he had only tested the card at Full HD or more. But he didn't. If you're gonna test something at four resolutions you can't dismiss that in the final results and say "hey, two of the tests don't count".

He tested at four resolutions and the average of those tests showed the GTX590 being 4% ahead of the 6990.

Performance per watt ,there is only one card thats worse.GTX 480.HD6990 is better.

In a dual GPU solution, that costs $700, that is big, hot and power hungry, who really cares about perf. per watt?

It (the GTX 590) consumes more power than the 6990 at both idle ,gaming and peak.In furmark its hard clamped by nvidia so it does whooping 4 watts better.But way nvidia handles this clamp.Its no where near real life scenario.
So again, no.

I'll give you that. Even though i find it very odd that GTX590 Peak power is higher than Maximum power.

It is Quieter, still loud.

It's quieter.

As for temps, in TPU`s review, there isnt any chart with temperatures.

Wrong...

GTX590 @ tPU!


6990 @ tPU!

However you do understand i hope that if it draws more power, it has (as an entire card, not just GF110 chips) to exhaust this somehow.More power=More heat.

You have to have good airflow with both of them.

So yes.It IS QUIETER.On the rest, youre wrong.

Quieter, cooler, more all round performance, higher OC (26% vs 10% of the 6990), same price and... 7.0 score. Go figure.

[DarthShader]

Quieter, cooler, more all round performance, higher OC (26% vs 10% of the 6990), same price and... 7.0 score. Go figure.

Lower core temperatures are not everything, you know?

This has been posted so many times already, you must have seen it.

Oh, and your OC figure is wrong. TPU got 775mhz, which is stock clock for a GF110 chip. So that's a 0% overclock. And not a given on all cards. AMD still managed to get a positive here in any case.

[Kristoferr]

What you want to do is not to compare those two monster cards in idiotic resolutions like 720p. What you should want is compare them FullHD+ and maybe perfomance per watt and perfomance per dollar.




Considering that. GTX 590:

+ Looks nicer than HD6990
+ Runs quieter than HD6990
- Runs hotter than HD6990
- Has a risk of overheating PWM area
- Has no multiple monitor indipendent desktop support
- Is more expensive than HD6990 in europe.

In my eyes HD6990 wins.

[XRL8]

Like the unlucky coincidende of his review showing the GTX590 being faster (all resolutions)

Well, you said its faster at all resolutions.It isnt.
Its faster at low ones, and ties/slower at high ones.So you cant say "its faster".Cos it "depends" on the conditions.And realistically you have to admit that these cards aint gonna be used at 1024x and neither 1650x most probably.I know its hard pill to swallow.ITS NOT FASTER :-).

In a dual GPU solution, that costs $700, that is big, hot and power hungry, who really cares about perf. per watt?

Anybody who cares about heat output and noise of their system.For example.
You put waterblock on a card A that draws 100W and card B that draws 200W, in the latter example you will have to turn up fan on your radiators.
Same goes for any decent aftermarket cooling.Idle wattage is important also, low enough idle gives you silent system in non gaming scenarios.
I know i know, nvidia has put bigger fan on their reference cooling.But im gonna point you to the 700$ bit, if people arent concerned with power consumption due to the electricity bills, they most probably can afford arctic cooling silencer or a waterblock ;-).

As to the temperature, i honestly missed that bit.Yes.Nvidia wins, by one celsius on one of the GPU`s which isnt in boundaries of a statistical error at all...
Besides if you dont use same cooling, results arent really comparable.
Yes in idle it looks better for nv, but higher idle temp(still lowish) while consuming less power on the ati card clearly indicates different fan/and its temp curve.

Quieter, cooler, more all round performance, higher OC (26% vs 10% of the 6990), same price and... 7.0 score. Go figure.

The CARD isnt cooler, it exhausts more HEAT into the system than the 6990, it was proved in anands test for example.Besides its pure physics.Uses more power,generates more heat.Then there are of course those temp shots on this french website which clearly show that the card is hotter (while the GF chips are a bit colder,bigger fan).
TPU is european website, price isnt "same" also here.
As for the OC capability, it gets dangerous on the 590 ,a few of them blew up, and not only at 1.2v, 1.025v was one of them.Besides for ocing, GTX580 or HD6970 are much better suited.
And dont get me wrong, i wouldnt buy any of them.
580 or a pair of 6950`s .Thats the only thing that makes real sense.

[bamtan2]

Considering that. GTX 590:

+ Looks nicer than HD6990
+ Runs quieter than HD6990
- Runs hotter than HD6990
- Has a risk of overheating PWM area
- Has no multiple monitor indipendent desktop support
- Is more expensive than HD6990 in europe.

In my eyes HD6990 wins.

It wins, but are you buying it?

[RSC]

Lower core temperatures are not everything, you know?

This has been posted so many times already, you must have seen it.

Oh, and your OC figure is wrong. TPU got 775mhz, which is stock clock for a GF110 chip. So that's a 0% overclock. And not a given on all cards. AMD still managed to get a positive here in any case.

1- Is it that hard to understand that I was only referring to the tPU! review? On tPU! he doesn't test system temp, nor mosfets or VRMs temp. He tests core temp. And on that matter, as I said and as it is written on both reviews, the GTX590 got... better temps.

2- The card comes stock @ 613MHz core clock, he got it to 775MHz core clock. That's 26% OC. Don't try to spin it...

What you want to do is not to compare those two monster cards in idiotic resolutions like 720p. What you should want is compare them FullHD+ and maybe perfomance per watt and perfomance per dollar.

(...)

Considering that. GTX 590:

+ Looks nicer than HD6990
+ Runs quieter than HD6990
- Runs hotter than HD6990
- Has a risk of overheating PWM area
- Has no multiple monitor indipendent desktop support
- Is more expensive than HD6990 in europe.

In my eyes HD6990 wins.

My point is not what I should be comparing. My point is what the tPU! review showed in their tests and how, independently of that, the GTX590 got a 7.0 score when the 6990, that had worse all round results, got 9.1.

Well, you said its faster at all resolutions.It isnt.
Its faster at low ones, and ties/slower at high ones.So you cant say "its faster".Cos it "depends" on the conditions.And realistically you have to admit that these cards aint gonna be used at 1024x and neither 1650x most probably.I know its hard pill to swallow.ITS NOT FASTER :-).

1- Mate, I was talking, as you understood by now, about the "Relative performance (All resolutions)" chart. So don't try to change the subject. Cause on that chart, which is the global performance chart for the tPU! tests, the GTX590 is FASTER than the 6990.

2- Do you have any stats on the resolution at which dual gpu cards are used? Any study, etc? No, you don't. So you can't say that they will be used at this or at that resolution. It is like saying all Asus Rampage boards were bought by people who overclock. This is computer hardware and lots of people buy stuff just cause it's expensive.

As to the temperature, i honestly missed that bit.Yes.Nvidia wins, by one celsius on one of the GPU`s which isnt in boundaries of a statistical error at all...

You missed the idle temps also, didn't you? And the OC temps as well...

Besides if you dont use same cooling, results arent really comparable.
Yes in idle it looks better for nv, but higher idle temp(still lowish) while consuming less power on the ati card clearly indicates different fan/and its temp curve.

That is irrelevant to me. Again, my point is just the difference in score on the tPU! reviews. GTX590 got better all round results and lower score. And that, in my book, is shady. And I won't even start on the EPIC FAIL anagram...

The CARD isnt cooler, it exhausts more HEAT into the system than the 6990, it was proved in anands test for example.

Take N: I was talking about tPU! review only. And they measure only core temp.

TPU is european website, price isnt "same" also here.

Price is the same on their review. $699 each.

As for the OC capability, it gets dangerous on the 590 ,a few of them blew up, and not only at 1.2v, 1.025v was one of them.

Have you read why? Go search this thread about SweClockers and drivers, maybe you will get it.

[Johnny87au]

It wins, but are you buying it?

I sure am, cant wait till the blocks are out for it and gonna throw 2 of these babys under water

[STEvil]

1- Is it that hard to understand that I was only referring to the tPU! review? On tPU! he doesn't test system temp, nor mosfets or VRMs temp. He tests core temp. And on that matter, as I said and as it is written on both reviews, the GTX590 got... better temps.

2- The card comes stock @ 613MHz core clock, he got it to 775MHz core clock. That's 26% OC. Don't try to spin it...



My point is not what I should be comparing. My point is what the tPU! review showed in their tests and how, independently of that, the GTX590 got a 7.0 score when the 6990, that had worse all round results, got 9.1.




1- Mate, I was talking, as you understood by now, about the "Relative performance (All resolutions)" chart. So don't try to change the subject. Cause on that chart, which is the global performance chart for the tPU! tests, the GTX590 is FASTER than the 6990.

2- Do you have any stats on the resolution at which dual gpu cards are used? Any study, etc? No, you don't. So you can't say that they will be used at this or at that resolution. It is like saying all Asus Rampage boards were bought by people who overclock. This is computer hardware and lots of people buy stuff just cause it's expensive.



You missed the idle temps also, didn't you? And the OC temps as well...




That is irrelevant to me. Again, my point is just the difference in score on the tPU! reviews. GTX590 got better all round results and lower score. And that, in my book, is shady. And I won't even start on the EPIC FAIL anagram...





Take N: I was talking about tPU! review only. And they measure only core temp.




Price is the same on their review. $699 each.




Have you read why? Go search this thread about SweClockers and drivers, maybe you will get it.

This is why you should read multiple reviews.

[bamtan2]

what sticks out to me now is just how broken SLI is in stalker and metro. something about those darn eastern europeans. it is clearly not the 1.5gb of memory because 580 and 6970 are close just like they're supposed to be. but the SLI pairs suck donkey dong compared to the crossfire pairs.




if a reviewer compares their 6990 vs 580 results to their 580 vs 580 sli and 6970 vs 6970 crossfire results, I expect they will find that all the games where 590 sucks compared to 6990 are the games where the 580 sli scaling sucks. meaning some software guys somewhere need to fix their crap.

[Lanek]

If it was only that ... we will be all happy ....

[JohnZS]

I will make my decision in May.
At the moment I am leaning towards the 590, however the VRM issue concerns me, if only those who over volt the cards are effected by this then it is not an issue for me, however I am concerned it could hit people on stock volts.
Drivers are already boosting performance of the 590.
I am happy enough to wait with my 295 Single PCB as it still plays my games at high and even maximum details @ 1920*1200.
I know to some people it may seem silly upgrading from this card, but I would like to have more than 896VRAM and better PhysX performance and also the bonus of DX11
John

[Mungri]

The 600 Mhz clock is holding it too far back, each GPU is severely bottlenecked by the clock speed. And they can hardly be overclocked without blowing up

You need to get the GPU clock up to around 850 Mhz on the Geforce 500 cards to have them running at their best potential.

[Postmodum]

True True ...
my 580 just loved the 960Mhz overclock ... it was getting 30500marks in vantage ... not bad considering the 36000 marks of 590 in the same bench ...

[sniper_sung]

We have calculated the average score of top 10 Performance results from 3dmark.com for 590 vs 6990, removing redundant submissions. Apparently the PWM of 590 is too fragile to OC, making 590 losing slightly against 6990.

GTX 590:


HD 6990


Mean score for 590 is P9797.9, while mean score for 6990 is P10540.4. Waiting for more 590s to exceed 700MHz and upload results.

[XRL8]

1- Mate, I was talking, as you understood by now, about the "Relative performance (All resolutions)" chart. So don't try to change the subject. Cause on that chart, which is the global performance chart for the tPU! tests, the GTX590 is FASTER than the 6990.

2- Do you have any stats on the resolution at which dual gpu cards are used? Any study, etc? No, you don't. So you can't say that they will be used at this or at that resolution. It is like saying all Asus Rampage boards were bought by people who overclock. This is computer hardware and lots of people buy stuff just cause it's expensive.

Well, if you did talk about average of all results than ok.You just didnt put it that way before.You said it was faster at all resolutions.
As for the second part, no i dont have a study.But as you said yourself, people buy stuff thats expensive (if theyre uneducated) ,so they wont pair a 590 with a 17" crt from 1995 to play it at 1024x .Theyre gonna get best of the best, thats 2560x or multimonitor setup.And even if they choose just the most common denominator, its gonna be Full HD aka 1920x .
If you really believe that such a card will be paired with a low res monitor be it an joe enthusiast or joe gold teeth.I guess youre just out of touch with reality.
But getting back to the review score, they almost definitely took the highest res results only for consideration of the performance, because thats the target audience of this kind of cards.Ultra high end.

You missed the idle temps also, didn't you? And the OC temps as well...

No i didnt, read my previous post, just this time do it when youre calm ;-).

That is irrelevant to me. Again, my point is just the difference in score on the tPU! reviews. GTX590 got better all round results and lower score. And that, in my book, is shady. And I won't even start on the EPIC FAIL anagram...

Again, back from fairy land.It didnt get "all round" better results.
It got worse results in high res scenarios (which are the most important ones for the reviewer i guess).
It had higher power consumption.
It blew up ,and wasnt alone in it.
As for the price, they listed MSRP, because thats what they got from Nv.Real price is different.And they take into consideration real price not the imaginary one.

And most importantly the score, isnt just a mathematical calculation of all the single results.Its a reviewers opinion based on what HE thinks is important for a buyer of such card.The card blew up on the guy.He wasnt particurarly impressed by that it seems.As for the drivers and blewing up, for one, its nvidias fault and second ,asus gtx590 blew up in pclab.pl with the non-blewing drivers.

But yes, if you take only points you like, then the card is awesome.
-Massive 11% performance advantage (in 1024x768)
-Quieter (from very loud to loud)
-Most of them dont blow up!
-Overclocks higher! (with a big chance of not blowing up using semi correct drivers!)

I strongly believe all TPU`s results are fair.As to the conclusion, reviewer has different opinion then yours, but given all the info this review gives out.Its up to you to really decide.
If he would be on AMDs payroll as you faintly suggest, they would choose specific amd only benchmarks and do them most favorably to AMD.Theres nothing fishy about the results.

[cowie]

That bench 3d11 right? it favors the amd cards i thuoght? just asking...plus just the 3d parts would be better and avg. cpu clock
how does say that heaven bench work out since thats the only other dx11 to compare?

oh and this card can we agree 590 has about 9% overclock with stock volts...i would not know but is that about the same as the gx2 and 295 cards?

What do i care anyways bring on the 28nm already

[Micutzu]

GeForce GTX 590 overclocking | lab501.ro - translation

Serious cooling, 50% overclock on air, the works.

[bamtan2]

GeForce GTX 590 overclocking | lab501.ro - translation

Serious cooling, 50% overclock on air, the works.

this is why god made birds sing. to celebrate this.

[Olivon]

GeForce GTX 590 overclocking | lab501.ro - translation

Serious cooling, 50% overclock on air, the works.

Great work Lab501 team

[Stukov]

How long did they run before they died?

My first one died at about 1.5-1.75 years and the second died 6 months after the it arrived.

[Lanek]

what sticks out to me now is just how broken SLI is in stalker and metro. something about those darn eastern europeans. it is clearly not the 1.5gb of memory because 580 and 6970 are close just like they're supposed to be. but the SLI pairs suck donkey dong compared to the crossfire pairs.




if a reviewer compares their 6990 vs 580 results to their 580 vs 580 sli and 6970 vs 6970 crossfire results, I expect they will find that all the games where 590 sucks compared to 6990 are the games where the 580 sli scaling sucks. meaning some software guys somewhere need to fix their crap.

you compare a 580 clocked at 772mhz to a dual 580 but who are clocked at 607mhz .. you can't use the 580 score as comparaison... I don't say you are wrong, but it's really hard to compare with thoses numbers.

[E.R]

what sticks out to me now is just how broken SLI is in stalker and metro. something about those darn eastern europeans. it is clearly not the 1.5gb of memory because 580 and 6970 are close just like they're supposed to be. but the SLI pairs suck donkey dong compared to the crossfire pairs.

if a reviewer compares their 6990 vs 580 results to their 580 vs 580 sli and 6970 vs 6970 crossfire results, I expect they will find that all the games where 590 sucks compared to 6990 are the games where the 580 sli scaling sucks. meaning some software guys somewhere need to fix their crap.

40,1 vs. 60,1 for Stalker in SLI and 28,74 vs. 52,31 for Metro2033 in SLI dont look that bad but not perfect either.

[bamtan2]

you compare a 580 clocked at 772mhz to a dual 580 but who are clocked at 607mhz .. you can't use the 580 score as comparaison... I don't say you are wrong, but it's really hard to compare with thoses numbers.

40,1 vs. 60,1 for Stalker in SLI and 28,74 vs. 52,31 for Metro2033 in SLI dont look that bad but not perfect either.

Let me help you understand what I mean.

SLI scaling in crysis is 85%. But in metro it is 64%.
CF scaling in crysis is 77%. But in metro it is 80%

By scaling I mean how much extra performance is gained by adding 100% extra (one more) graphics cards. SLI is unusually bad in metro. This explains why reviewers are complaining about GTX 590 performing well in some games, but bad in others like metro. The SLI scaling is bad.

Look at the anandtech numbers yourself. GTX 580 beats 6970 in metro. But when you SLI it, it loses.

crysis numbers: http://images.anandtech.com/graphs/graph4239/36042.png
metro numbers: http://images.anandtech.com/graphs/graph4239/36049.png

CONCLUSION: SLI SCALING = BAD IN METRO AND STALKER