Discussion in 'Hardware' started by Guest-44432, 20 Jan 2012.
cant wait till i get my new card. i only have 1gb vram but play at 2560x1440
I did sell my 580's to early, as I was hoping the HD7970 CF drivers were fine. But as it turned out, it wasn't, So I did what I thought was best and got a couple 3GB versions to tie me over.
If what Pete J is saying is correct, about the GTX 680 being released to knock out the HD7970, and then AMD to bring out either the HD7990 or HD8970 to knock that out. Then Nvidia releases the GTX780 again this year to knock out the HD7990/HD8970.
Then I will be holding off for the GTX 780.
My 3GB GTX580's will keep me gaming till September, when they do release the GTX780...Enough time, to save up for Ivybridge-E and 3 GTX780's, and Windows 8 of course.
Ha, they will turn up at your door, once you have paid for them
Indeed I think thats what I might do, although I am very tempted to get some HC Cheesecake 580's
thanks for the reply.
If you are referring to this, then I forgot to reply. Sorry.
I don't know, is my honest answer. I don't think they would be able to clock as high, as the work that's gone into the Classified's are purely based on OCing, The VRM's can give clean power upto 1.7v vcore (although you would be needing LN2 for that).
To compare the Classifieds to my Zotac 580's, when watercooled. I would say the fan on the classified does a awesome job at keeping the GPU cool and quite, and only a couple of degrees above my watercooled 580's. Can't wait to get a waterblock on these.
You're a nutcase mate
Should have just held out like I am, I couldn't even be bothered to pick up another 570 for the time being - which means I now annoyingly have 3 screens staring at me and waiting to be played on... -.-
i see. cheers
Just as an interesting note on the Crysis benchmark. I'm getting almost identicle performance at 1080. My monitor is actually 1920 x 1200 but I just ran the benchmark at 1080 for comparison, you can see previous test in the log, and the difference between 4x AA and 16xQ which strangely gives better performance.
i5 2500k at 4.7 under custom waterloop.
Crossfire Asus 5850 DirectCU @ 810 and 1020
Corsair Vengeance 1.35v 1600mhz 8GB 9-9-9-24
Corsair Force GT 128GB
You haven't clicked the 64bit...
It crashes when I use 64 bit, so I've never bothered to re-install it. Can't see that making any difference.
I just wanted show its more likely the software causing the low fps for you guys cards.
How did I miss this thread?
So you sold 3x 580GTX and bought 2x brand new 3gb versions?!
You sir ether have a very understanding wife/girlfriend, or are hung like a donkey...
But he's also got sh!tl04ds of cash - eh, eh?
lol I don't think it's a matter of being minted or having **** loads of cash to buy these cards, but I think it's rather where priorities lay and what the interest is. He obviously loves gaming so he loves buying GPUs, as where somebody else might love going car's and spend all there money on those
Either way any hobby is costly if you look into it deep enough
Would that make a difference?
How well do the classifieds work?? I'm running evga ref design 3gb 580s and I can (with my own bios) hit 1050core without much issue although it's not a 24/7 thing!
Well, I'm glad I did buy these cards. As after Playing Crysis 2 maxed out, I was running over 2.1GB of vram at 2560x1600. Same as BF3 etc.
Anyhow the missus doesn't have a say in my spare money, as I pay the mortgage + bills. Where she is happy spending her money on pampering her self with £100 hair cuts, Nails, Spa etc.
Yep, as it will be restricted to how much memory the game can use, so you will see more performance in 32bit.
I would keep with what you have got. As 1050mhz is not to be sniffed at. I can run 1000mhz on each card, but not in SLI for some strange reason...Which is outstanding for a air cooled card.
I would like to get some waterblocks for these, as I think I can get to your clocks speeds quite easily.
970mhz is the max I can get stable on air in SLI.
In the not too distance future I will be able to add some results to this thread.
Separate names with a comma.