Originally Posted by Shinobi Jedi
I'm not communicating what I'm trying to say very well. Because I see what you mean about those statements being contradictory, and yet I'm trying to say essentially the same thing you are in your response.
See, those results are impressive, that's what I'm talking about being down with. Where I see that I am not clarifying really well, is I was comparing the 590 and the 6990/Mars in my head and not the 590 against itself.
Meaning, in the benchmark reviews, even with the overclock, the 6990 was almost always less then a 10fps difference between the cards where the 6990 does come out ahead, even when it's overclocked.
Whereas the 590, brought me a 10fps+ in the Metro 2033 bench just going from the 270-275 series of drivers. But because a 6990 edges out the 590 by such a small difference in FPS, even though your eye most likely won't see the difference (unless it's at the border of being choppy FPS) - It's the "big winner"
But the reviewers that are Overclockers first and gamers second, don't really factor that it's leaf blower loud and there's been a history of bad driver support at AMD, though that seems to be changing. Of course this still applies if the results were flip-flopped. If someone wanted to game with 6 monitors and headphones, then they shouldn't care if the 590 is now out performing the 6990 in many benches.
And now we have the Mars, and the hypocrisy in scores between 7.0 and 9.4 because one overclocks better by %10, is exactly what I mean. There's too much emphasis on overclocking performance and not on the other attributes.
If I was reviewing the Mars card as a gamer, then the Mars score would also have to be graded on real world, gamer applicable qualities. The two biggest one being size and price.
First, this being a 3 slot card, is there even a mobo available that has the room to run two of these in SLI if some Richie-Rich wanted to? If not, then the fact that you're going to be stuck with this one card in your machine should be a factor in the final review score.
The second is price. Being that the suggested MRP is equal to getting two 590's for Quad SLI, should completely be a factor in the card's grading.
But no, because the card comes with true 580 level stock clocks, it gets a 9.4 - Because too much importance is placed on overclockability, rather than all the criteria that someone who is into PC Gaming, only for the games, can make an informed decision on.
That's what I'm trying to say. The fact that people like you are getting such awesome performance increases and this isn't making news in Hardware enthusiast sites, is hypocritical.
I have a feeling that reviewers like W1zzard are too worried that if they go back and take a second look at the 590 and admit they got a little too eager the first time, that they'll lose some kind of credibility, when truthfully, they would actually gain credibility by displaying true objectivity.
So, what I meant to say was, there's was mostly only a less than 10fps difference between the 6990 and 590 with overclocks applied (I'm going off memory, so I may very well be wrong) - And what I was trying to say was, show me an overclock that gives 20-30fps bump of one card over the other 2 out of the top 3 on the market and then you can color me impressed. (You made me see that 10fps is too generous and easy to obtain)
So, If I haven't confused things any further, while I think you're misunderstanding what I was trying to say because I didn't communicate it well - I do concede that when looking at it from your perspective it does seem contradictory. But I forgot to clarify that I meant the 590 vs. 6990/Mars and not against itself - in regards to which one makes the best purchase for the high end gamer.
While I'm pulling internet faux pa's and actually admitting online when I'm wrong, let me really dig into some crow by saying in regards to Deus EX:HR -
Rush and Masked were totally right and I was wrong about the "Micro-Tearing".
My buddy is getting the same thing on his new HP laptop that uses AMD's crossfire (In the same way Nvidia applies Optimus in their notebooks) to switch between his notebook's 6750m and the intergrated GPU.
Only enabling Vsync gets rid of this.
Also now that I'm deep into the game, my GPU usage and FPS are now all over the place too. Anywhere from 50fps for a milli-second, then all the way back up 120fps before u know it.
So, I'm pretty confident this is the game's optimization/coding rather than any of us having issues with our 590's and rigs.
If I was getting these FPS in a competitive MP game, I'd be pissed. But since I'm playing a SP game for the story and everything is fluid for the most part, I don't mind.
I still fracking love these cards. I love them so much, my wife is starting to get jealous.
Although, I'm sure it's probably not the case, I'm justifying to myself to buy a couple of SSD's to see if that helps with any of the Microstutter. One for my OS and one to run games off of. Of course, I know it probably won't help much, but I just really want one.
Anyone know if there's a certain brand or model that's really good for games? Similar to how WD Velociraptor HD is designed with games in mind?
Sorry for the long post. I wasn't planning on writing a book.
What it comes down to, all the "590 hate" stems from TWO reviewers. Sweclockers and W1zzard who does the good TPU reviews benching everything an analyzing numbers.
Both of these guys killed their 590 samples. W1zzard I believed killed two. While I applaud him pointing out the issue to nVidia, he utterly and unequivocally destroyed the reputation of this card.
Our max voltage I will modify any of our custom BIOSes is 1.063v. Thats really high in my opinion. So high in fact, I only recommend people to use water cooling and that voltage. If anyone has seen my oc thread, its very cautious.
The voltage W1z killed the GPUs at was 1.213v. The maximum voltage for a 580 using a modified BIOS
. It wasn't even incremental, like testing every 10mV, he went straight from 1.1 to 1.2v I believe. I will find the initial article later.
He then did childish things like list the Pros-Cons in a way such as every letter spelt Epic Fail:
It has since been "edited". Despite every benchmark he ran at stock, being better than 6990, equal to 6990, and slightly behind the 6990, the card was scored a 7.4, where as a 6990 was scored as a 9.0 and Editor's Choice.
Every dead GTX 590 image stems from 3 or 4 incidents of cards dying, yet somehow it turned into EVERY 590 was destined to explode. Rotated images/shots were claimed as new "users" killing their cards. Raziel (an OCN member) claimed his died at stock clocks. I have said it again and before, if all were true, you can count the number of "exploding" 590s on two hands. At least the public ones.
What is annoying isn't really the reviewers doing their thing, its the encouraging a million little parrots on the internet to repeat the same junk either out of context or out of a realistic operating parameter such as "all 590s are destined to explode" instead of "all 590s will probably blow up at 1.213v."
Furthermore, if it wasn't for this total douche baggery with how they handle the extremely public popular reviews - our PDL would barely even exist/be much lessened. Instead we have two hands a foot handcuffed when it comes to OC and in some cases, just general performance.
Oh well, its very true what has been said before, the overclocking emphasis has been taken to the extreme on this card. I can get a 30% OC on my GPU, which is fantastic, yet its still considered a "Fail" card. Even on OCN, all they care about are these initial review, not what people are actually achieving with it. Its this misinformation which fired me up to start the benchmarking thread.
But with all things, you can only go so far to wipe out ignorance, you have to have a willing mind.