Discussion in 'Graphics Cards' started by damric, Mar 27, 2011.
if you are referring to dual video cards/gpu's, then i would agree, to a degree
its more proof of how 3Dmark is a waste of money, and a terrible way to compare video cards.
This. Gaming is the best way to test cards. 3dmark is only good to test initial stability of GPU overclocks. That's the only use I see for it.
The first card is overclocked to 1GHz but the scores still shouldnt be that close
good advice here
The 590 doesn't need to win some silly little benchmark because it's bi-WINNING - it wins here and it wins there!
The scores are fake, hampered by some problem and/or just in error. The scores are way off. Besides, if you base your GPU configuration off of 3DMark, you have issues.
QFT. One cannot contain the power of one thousand suns.
Not fake, although I did put the CCC settings to optimal performance. I understand it must be quite upsetting to see these scores. You can look them up in the 3dmark11 database if you like. You will find similar Graphics scores for all GTX 590/HD6990 cards. P scores in 8000's range, and X scores in 3000's range, JUST LIKE IN W1ZZARD's REVIEW.
Keep in mind that my HD 6850's are clocked at 1ghz. It's not my fault that they scaled so well for this particular application.
Now considering that I paid $150 per HD 6850 (after rebate), and those flagship cards cost $700...yeah that would make me cry if I was duped into buying one of those cards, especially the GTX 590 that doesn't even overclock without exploding
Not upsetting at all! Nice work.
*Are you using the 11.4 "preview" drivers?
This all came about when I put the 2nd HD 6850 in my rig this week and I saw it underperforming until I disabled the ULPS in Windows registry. I was only scoring like P5400 with the 2nd card. Since I was scoring P4500+ on just one card, I knew something was wrong. Once I got my crossfire/clocking issue resolved, I wanted to compare some scores. I stumbled across the GTX 590 and HD 6990 scores and I lol'd.
Yes, these are 11.4 beta drivers, and the HD 6990 is using the same drivers if you look closely at the scores.
I'll post my single card scores from a few months ago with the old drivers, as it makes sense, since it's about half the score.
While I don't dispute your numbers, one of the reasons that W1zz takes the time in his reviews to show the numbers not only in synthetic benchmarks, but also in gaming specific benchmarks, is that the synthethic benchmarks do not tell the whole story.
It's pretty amazing what AMD did with these drivers. Looking forward to see how the WHQL's turn out.
Um, i prefer to look at it truthfully, as see how BAD thier drivers were before, and now, they are just now getting to where they were back over a year++ ago.
with 69xxx, new architechture, with a very fundamental shift in design, is gonna take soem time to optimize...funny, March is lamsot over, and there is no 11.3 driver...I dunno if that's good....or bad!
Sure. Also take into consideration the different shader config with the 6 series. But yes, I know all about ATi driver issues in the past.
This is my single GPU score from the older drivers. It has a slightly higher clock, 1050/1225 if I recall.
Ha, edit fail.
You left AMD cards when it was seemingly at it's worst...I'm pretty happy with HD6950, at this point. Still got the Eyefinity issues, but multi-card scaling is much better, for sure.
Heck, I'm playing most often just on 4870...@ 1920x1080...
That post was written at one time, no editing.
Yes, u mad I left AMD bro? Lol. I love wasting my money on different cards, I can afford it. I recommend it to everyone.
Nice clocks.....One problem though, ...... 95% of graphics card owners don't overclock..... therefore they HAVE to pay if they want the performance
No, my edit. It's all about you though, isn't it?
And no, point is that you swapped when it was best to. Can't knock that!
I'm happy somebody understands.
not suprising, my stock score is over 9k for performance mode in 3dmark 11.
overclocked score is just shy of 11k. (12 pts shy. lol)
both of these duals are less impressive than they should have been. the 590 should have been at 700 core at least. Then it would be at 9k as well.
dual gpu cars do offer crossfire/sli for those who don't have mobos that support it though. Combine that with the quasd sli/quadfire crowd and there's enough reason to make them. They just won't be able to compete with 6970 crossfire much less 580 sli.
I thought that was fairly obvious to everyone...? If you want the best bang for your buck, you're purchasing either an HD 6950 or a Geforce 560TI. If you aren't concerned about budget and want performance, you are buying 2x HD 6970 or 2x Geforce 580. The argument over the single slot is worthless, as do you really think someone looking to spend their money on such a solution really has only one PCIe 2.0 x16 slot? And for those that really want to set world records, they are looking at Core i7 XL-ATX motherboards with 4xPCIe slots for four of these cards, not for two dual-gpu cards. While the HD 6990 isn't quite as crippled as the Geforce 590, both are still watered down to operate at reasonable conditions.
These super noisy dual-gpu behemoths are really only for bragging rights by the manufacturers. IMHO, they shouldn't even waste their R&D to create them, and should leave it simply up to the AIB partners to make them if there is demand.
Separate names with a comma.