GTX 590 reviews

[cowie]

12 posts? i will look into that one..

ss of the"GAME OVER" guy the day before,he only had 1.04 ss say same.
840!!!!
but is that not what wizzard was going before his car died @1.21v?


sry about that i hate that too.

[RSC]

'Mate', if you were familiar with statistics you would know something about weighting.

Again, he is entitled to that. No problem there. As I said before, my only issue is with the discrepancy between test scores and final score of both cards. Nothing else.

So then why are you ripping on IPS? Do you just not understand there is more at play than just pixel response times, or is that why you bother to take into consideration numbers below 1920x1080 which are pretty much pointless for anyone who's gonna blow that much cash on something like this?

1- Cause IPS are not considered the best monitors for gaming...?

2- I'm not taking anything into consideration. He is, by testing the cards at those resolutions. If he is not taking scores below 1920x1200 into account why test them at those resolutions? What's the point?

I'd let him hold on to that 4% which may very well be within the margin of error, without even considering anything below 1080p is pointless. You can't really take anyone using that argument seriously

Have you seen the advantage of the 6990 @ 2560 on the tPU! review? The huge lead at high res that made it get a 9.1 score, do you know the value? 3%...

on testing at lower resolutions:



http://www.techpowerup.com/forums/sh...3&postcount=38

If you want to ask him about something, ask him about the EPIC FAIL anagram and why he forgot/decided not to overvolt the 6990 on his review.

[Postmodum]

Cowie... now, there's what i would like to see on a GTX590 ...
Unfortunately, as it seems, those are suicidal benchmarking sessions! eheheheheh

Jeezzzz... this is the weirdest launch by nvidia i ever saw...

[Final8ty]

The latest drivers nvidia has released for the 590 have the card downclocking below 600 Mhz. The card is dropping as low as 550mhz when the OCP kicks in while gaming.

So nvidia is now having to reduce performance from stock to keep the card stable. Hopefully we will start to see review sites revisiting their reviews soon with the new drivers to show the loss not improvement in performance drivers are bringing to these cards.

This is unacceptable on nvidia's part, to now reduce the consumer's performance to keep the cards on the market.

It's time for a recall.

http://hardforum.com/showpost.php?p=...&postcount=141

[Sherman Tank]

Did you read the review? I mean actually READ it? I thought I was quite transparent about the card's shortcomings.

As for the "puff of smoke" issue, I am taking a wait and see approach rather than running around like my hair is on fire.

Hmm SKYMTL I have a few suggestions if you don't mind?
I think we have seen enough for you to issue a WARNING expressing caution on your review at a minimum.
I also trust you will revisit your review and update it with Nvidia's latest drivers to actually give users an accurate idea of performance they will get when their card now that it down-clocks itself to 550mhz depending on load?

You may also like to update the overclocking section to note the card can't effectively be overclocked as power management just down-clocks the card to less than stock clocks with is particularly unproductive when you need the extra speed.

I would have thought the above would have severe effect on minimums, No?

Also the "DAM INNOVATIVE AWARD" rings a little hollow now, maybe something like "DAM EPIC-FAIL AWARD" would work a little better as we've already seen?

Keep up the good work

[cowie]

Cowie... now, there's what i would like to see on a GTX590 ...
Unfortunately, as it seems, those are suicidal benchmarking sessions! eheheheheh

Jeezzzz... this is the weirdest launch by nvidia i ever saw...

yes i agree but is it software messed up on earlier drivers or api volt control or what??how does he get 840 on 1.04? others are not even close bs?good chips that have low volt sweet spot??
since even weak componant cards will not die within reason if proper safeguards are in place.

I have been bed ridden for a month and have no life so i play mister dectective the last week...like ufo hunters lol
i want to try one on old drivers to really see if api/software/bios have something to do with it...none in stock in the states thuo.

[cowie]

http://hardforum.com/showpost.php?p=...&postcount=141

i read that too
can anyone with a card back that up?
2day poster there idk?

just the facts ma'am

[Sherman Tank]

yes i agree but is it software messed up on earlier drivers or api volt control or what??how does he get 840 on 1.04? others are not even close bs?good chips that have low volt sweet spot??
since even weak componant cards will not die within reason if proper safeguards are in place.

I have been bed ridden for a month and have no life so i play mister dectective the last week...like ufo hunters lol
i want to try one on old drivers to really see if api/software/bios have something to do with it...none in stock in the states thuo.

The problem is hardware based.
All that the new drivers do is downclock the card well under stock clocks to somewhere like 550mhz when it get's stressed and performance goes all over the place and stutters. This effectively makes it imposable to effectively overclock now.

And guess what? cards are still dyeing...

[insurgent]

Have you seen the advantage of the 6990 @ 2560 on the tPU! review? The huge lead at high res that made it get a 9.1 score, do you know the value? 3%...

If you want to ask him about something, ask him about the EPIC FAIL anagram and why he forgot/decided not to overvolt the 6990 on his review.

The thing is, I didn't look at the total score when I read his or any other review, I looked at the graphs and read the conclusion. The score I learned from forums (probably from you), it's obviously subjective same as the other sites, I just trust that the benches they publish are correct since that's the data that's presumably objective.

It seems like you're the one who wants to ask him something, because I just don't care why it's scored like that. I went to TPU because his sample blew up

[RSC]

The thing is, I didn't look at the total score when I read his or any other review, I looked at the graphs and read the conclusion. The score I learned from forums (probably from you), it's obviously subjective same as the other sites, I just trust that the benches they publish are correct since that's the data that's presumably objective.

It seems like you're the one who wants to ask him something, because I just don't care why it's scored like that. I went to TPU because his sample blew up

Well, I did. And that's why I questioned it.

[deathman20]

Rush the product out and this is what you get? Wow.

Man would of figured the GTX 590 would of trumped the AMD 6990, guess its neck and neck and loosing ground. Props to AMD for pushing the limit.

[insurgent]

Well, I did. And that's why I questioned it.

The rating is subjective, if you disagree with him then that's that, but I did wonder why his 6990 review didn't have info on overvolting.

[Lanek]

http://hardforum.com/showpost.php?p=...&postcount=141

Huum im like Cowie on this one, want confirmation by 590's owners, this post look like a free affirmation without any backup.... ( maybe true, i will not be surprised, but a screenshot of gpu-z can be good. )

[Sherman Tank]

Lol at QC sticker!

[[XC] hipno650]

Downclocking the card nvidia???? someboady dropped the ball over there big time.... looks like they took the whole power consumption thing a little to much into effect.... they should stop wasting time and recall the cards. Intel found a bug in their stuff and recalled them. time for Nvidia to do the same.

[XRL8]

A)The "best of the best" for gaming are screens that work @ 2560...? Isn't 2560 a resolution where the majority of screens are IPS? IPS and gaming? Sure...

B)As I said, you have no hard data to back that statement.

C)Again, show me a study or a statistic and then we'll talk. In the meanwhile you're just (wrongly) assuming that everyone who spends money on dual GPUs is a connaisseur.

D)So why did they test it @ four resolutions? If the lower resolutions aren't relevant, why do it at all? And if they did it, why ignore those results? ...

F)Mate, you've got some serious comprehension issues. The "all round" performance result chart gives the GTX590 a 4% lead over the 6990. Stop trying to spin that, it's a fact.

G)S-c-e-n-a-r-i-o. Cause it was one resolution only.

H)It blew up highly overvolted and with voltages way above the limits recommended by Asus.

I)Stop making stuff up...

(http://www.techpowerup.com/reviews/A...TX_590/27.html)

Sure... He can (and he did) toss all the results up in the sky and give it the mark he wishers (and the, very subtle, EPIC FAIL anagram).

J)The card blew up on him during testing @ stock? No. The card blew up on him while OCing? No. The card blew up on him while overvolting the out of it to run... stock clocks. That's a really interesting test.

1- 11% advantage @ 1024x768, 7% advantage @ 1280x1024, 3% advantage @ 1680x1050, and same performance @ 1920x1200. But hey the 6990 got a 3% advantage @ 2560x1600, so it clearly deserves 2.1 more points.

2- If the GTX590 is loud imagine the 6990, but I'm sure that's a plus for the 6990.

3- Overclocks higher is certainly a con, right? No comments.

Good luck picking a red t-shirt among other t-shirts of other colors. It would be quite the feat for you.

A) Uhm, IPS gaming screens are in many areas better than TN.TN has only one advantage, thats response time.And yes, people who do like to watch quality screen go IPS or VA.Games arent only unreal tournament ones you know.Good IPS screen has good response time with much better picture quality.So ,yes, IPS is for gaming too.And most importantly people who can afford these cards really like quality.And even if you go for TN, its 1920x1080.

B)Well, i have logic.As it does not apply to you, i can just fireback, you dont have any data to back up your statement too.So thats a draw mate.

C)You cant say i wrongly assume people who spend 700$ on GFX cards have a taste for the best of the best if you cant "back up" your statement with data too, now dont you ?

D)Because to get all data points needed for comparisons for every card, you have to have unified testing specification for everything from slowest cards to the fastest ones.Moreover having more data you can draw conclusions.Such as, maybe GTX590 is vram starved ?.More data points is ALWAYS better.As to the second part, they didnt ignore the lower resolution scores.They showed you them to make up your mind, didnt they ?However in THEIR assesement they arent relevant.

F) You said "GTX is all round better than 6990" .It isnt.All round means in every scenario.But its faster "on average" not "always".And has its drawbacks.You wont argue that overvolt protection works better on 6990 than 590 ?Or will you :-)

G) Yes it got lower score in one resolution, but its kinda implied that in multi monitor setting it would get worse score also...

H) According to nvidia these cards feature protection from such things.And they doesnt work properly.So thats not a plus.

I)Maybe you should show whole quote from him
From tpus review:

Price-wise both HD 6990 and GTX 590 are tied around $700 which is a lot of money to spend on a graphics card. My recommendation would be to go with a single GPU GTX 580 and wait what the future brings in terms of games - most games are console ports, Crysis 2 is DX9. Developers! The PC needs more love from you.

Yes, clearly AMD paid him off.Thats why he doesnt recommend neither card.And recommends 580.

Im gonna stop here cause its pointless.Score isnt based on median of the results, but what reviewer gets after digesting them all.Get over it.

[Buckeye]

12 posts? i will look into that one..

ss of the"GAME OVER" guy the day before,he only had 1.04 ss say same.
840!!!!
but is that not what wizzard was going before his car died @1.21v?

Cowie if you wouldnt mind, can you please resize that picture. There is really no need to post screens that are so huge, it makes us have to scoll way to the right and back to read every post.

If you use Imageshack you can also get a thumbnail link and use that instead of posting large pictures.

Thanks for understanding

[W1zzard]

Im gonna stop here cause its pointless.Score isnt based on median of the results, but what reviewer gets after digesting them all.Get over it.

i have said it several time in other reviews. our numeric score is just a rough indicator for people who skip to the last page, and then are too lazy to read the conclusion. it's a general "feel" score based on what the reviewer thinks. read the conclusions for more info and to get a summary of the key points. then read the rest of the review. then ignore those things that dont matter for you and come to your own conclusion

at tpu there is no correlation between numeric score and any awards. for editors choice i ask myself "would i buy it?"

[Vinas]

So if you had to choose a card based on planning for future use as opposed to current games, the 6990 is probably the better choice from a performance perspective.

Agree.

[RSC]

A) Uhm, IPS gaming screens are in many areas better than TN.TN has only one advantage, thats response time.And yes, people who do like to watch quality screen go IPS or VA.Games arent only unreal tournament ones you know.Good IPS screen has good response time with much better picture quality.So ,yes, IPS is for gaming too.And most importantly people who can afford these cards really like quality.And even if you go for TN, its 1920x1080.

Is that so? How many 120Hz IPS screens do you know?

B)Well, i have logic.As it does not apply to you, i can just fireback, you dont have any data to back up your statement too.So thats a draw mate.

See, that's the thing. You're the one saying that everyone who buys these cards have screens of a certain resolution. You're the one who is narrowing the field without proof of that. I'm the one leaving it open, as it is. Because, as I said before, hardware purchases, as many other types of purchases, are not always based on logic.

C)You cant say i wrongly assume people who spend 700$ on GFX cards have a taste for the best of the best if you cant "back up" your statement with data too, now dont you ?

Read above.

D)Because to get all data points needed for comparisons for every card, you have to have unified testing specification for everything from slowest cards to the fastest ones.Moreover having more data you can draw conclusions.Such as, maybe GTX590 is vram starved ?.More data points is ALWAYS better.As to the second part, they didnt ignore the lower resolution scores.They showed you them to make up your mind, didnt they ?However in THEIR assesement they arent relevant.

What's the point in comparing a dual GPU card with mid/low end cards? As for the vram starved, i'll get back to you later (G).

F) You said "GTX is all round better than 6990" .It isnt.All round means in every scenario.But its faster "on average" not "always".And has its drawbacks.You wont argue that overvolt protection works better on 6990 than 590 ?Or will you :-)

1- In five resolutions tested, it won three, tied one and lost one. That makes it more well rounded.

2- I don't know. Have you seen many 6990 overvolted? The only one I've seen, @ Neoseeker, curiously died shortly after the review. How do you like them apples?

G) Yes it got lower score in one resolution, but its kinda implied that in multi monitor setting it would get worse score also...

Oops...


(http://www.hardwarecanucks.com/forum...review-23.html)

H) According to nvidia these cards feature protection from such things.And they doesnt work properly.So thats not a plus.

Overclock always had risks. And he did go way above stock voltage.

I)Maybe you should show whole quote from him
From tpus review:


Yes, clearly AMD paid him off.Thats why he doesnt recommend neither card.And recommends 580.

There you go trying to spin things... You said "they take into consideration real price not the imaginary one". And they didn't. In their review, both cards cost the same, so price difference wasn't a factor either.

Im gonna stop here cause its pointless.Score isnt based on median of the results, but what reviewer gets after digesting them all.Get over it.

Final score is arbitrary? Makes perfect sense.

[Sherman Tank]

Oops...


(http://www.hardwarecanucks.com/forum...review-23.html)

That review (maybe all reviews) is irrelevant unfortunately, what with all the driver updates and down-clocking with 'all over the place' performance and stuttering.

[RSC]

That review (maybe all reviews) is irrelevant unfortunately, what with all the driver updates and down-clocking with 'all over the place' performance and stuttering.

Have you seen proof of the downclocking? Cause, so far, all I've seen is one user (total count: 3 posts) over at [H] forums claiming that.

[JohnZS]

I know I am probably going to get accused of Fanboyism here, but nVidia are not alone with cards burning up when they get pushed a bit.
ATi HD6990 also has the same issue, just for some reason it is not quite so widely publicised.
My guesses for this (as controversial as it may sound) is either:

1) ATi card can tolerate being pushed a bit longer so the burn out has taken a little longer to develop.
2) nVidia hyped the GTX 590 up so much that it is being scrutinised by a fine tooth comb.

Anyway this may be an interesting addition to the debate...

After our initial testing of the HD 6990 we moved the graphics card over to a backup system that we were using to test new games for our benchmarking suite. We were able to complete testing with the HD 6990 in some of our new benchmarks including H.A.W.X 2, Lost Planet 2 and DiRT 2, however, when we were testing the performance of Dragon Age II the HD 6990 died on us. At the time of it's demise the card was set at the stock 830MHz setting and the BIOS switch was in the default position. The fact that it died could have been that we tested the graphics card at both the 375W and 450W settings, but since the review we have left the settings at default level.

Presently this leaves Neoseeker without a HD 6990 for future testing. AMD will not warranty the card so we are left with no choice but to reach out to their partners to see if we can get a sample. So, from this point on it will only be seen in the benchmarks that were used at the time of the review until we can get a replacement.

Just thought I would post this to try and say that the GTX 590 is not the only Graphics card in the world which suffers from this.

John

[Lanek]

don't worry, it have allready been posted, but i don't know if we have really seen many case yet .. ( well i think it's the only so far i remember, but i don't know maybe some have got same problem ) If some know more cases, will be interessant to know what part have die too ( anyway no luck it's an AMD sample and not an AIB one.. )

[RSC]

The Editor in Chief of Overclockers Club, Bosco, already said that one of their samples also died recently and that he knows of four more dead 6990 and two other that are having problems.

Ya the 6990's are not fairing much better I know of 4 cards that are dead so far with 2 more having issues. One of ours died as well..... shakes head

(http://forums.overclockersclub.com/i...c=183386&st=36)