Buy one GTX 580. Where you can get 2x 295GTX for 100euros? It would be Quad-SLI and that doesnt work on like 99% games properly . Also you would need monster PSU, and single GTX 580 is faster than one 295GTX.
Buy one GTX 580. Where you can get 2x 295GTX for 100euros? It would be Quad-SLI and that doesnt work on like 99% games properly . Also you would need monster PSU, and single GTX 580 is faster than one 295GTX.
I have a 1200 Watts PSU, that should do the trick.
Also, is there that much a difference between the performance of a single GTX 580 to a GTX 295?
From what I've seen in several benchmarks, it appears that the GTX 580 has merely some 10% better performance than the GTX 295, is this false?
Well the 295 is 2x 280s in single PCB, and they are nearly as powerfull than single GTX 580. Tho 295 doesnt support DX11 etc.. 1200watts psu doesnt say anything, it could be some chinese bomb. Just get the 580.
I have a single 295GTX and it's a good card and still capable of running most things maxed (except some special cases like Witcher 2 ubersampling). But it's an old card with no DX11 support, and I plan on upgrading whenever nVidia gets the new kepler cards released.
Is there a reason that can explain why my ASUS GTX 560 DCU II sounds like its going to fly out of my case when I play games ? I'm not talking about Metro 2033 or GTA 4, I was just playing Dead Island, and the coolers were @ 40% and sounded like a hair dryer .
I don't usually care about the sound coming from the case, but this is exaggerated .
Bumping this back up, you'll see some stuff popping up in the usual places soon enough.
TSMC's 28nm process is doing well; they can achieve 45% higher clocks on 40nm. The 28nm process does use different materials, so noone knew what to expect, as this can't be compared to 40nm the same way 40nm could be compared to 55nm. AMD are aiming a bit lower than that 45%, but stock clock speeds over 1 GHz are quite likely. There's an event for their partners in 2 days, hopefully I'll get some more details following that
Various sites still report late December as a target. Not entirely right; we'll see some stuff around that time, but CES is where it's at (2nd week of January). That's also when availability is planned, although it's not quite certain how good it will be. Gotta wait and see if TSMC can do high volume reliably
Meanwhile, NVIDIA are preparing GF110 shrinks, no Kepler yet. That'll be late
I hope Bulldozer team doesnt have nothing to do with these new Radeons There has been rumours that 7970 single gpU will be as fast as 6990 DUAL GPU. Personally i dont believe this, cards are gonna be fast but not that fast for sure.
I hope Bulldozer team doesnt have nothing to do with these new Radeons There has been rumours that 7970 single gpU will be as fast as 6990 DUAL GPU. Personally i dont believe this, cards are gonna be fast but not that fast for sure.
Any stuff like that is something someone pulled out of his or her ass, AMD aren't ready yet.
Now whether it's possible or not? Definitely is. You gotta keep in mind that with the current generation of cards (GF110 and Cayman) an x% overclock yields almost an x% performance increase as well (it's about 10:9 roughly). I'm talking about shader bound games and benchmarks of course, if it's not shader bound an overclock has little effect. Combine this with the fact that TSMC can get to 45% higher clockspeeds, it's safe to assume AMD will be shooting for 30-35% or so; that puts us in the 1.1-1.2 GHz range.
So that's the higher clockspeed; we can expect memory bandwidth to be way higher again as it always does, but memory bandwidth is hardly an issue in most games.
Aside from that though, there's a brand new architecture created from scratch and we simply have no idea how good or bad that will be. Since it's a smaller node, in theory they'd also be able to put even more shaders into the same small die (because I have no doubt they will stick with the small die strategy) - then again, we have no idea what GCN looks like so I don't know how valid this comparison will be. Even if it's just 20% faster than the VLIW4 architecture on Cayman though, you're already looking at a 50% performance increase. This is also its downfall though, new architecture means a lot of driver optimisation has to be done
If they had done so, the new CPU line would've been productive and quite more successful than it is now. I think the Radeon products pretty much keep AMD afloat for now...
If they had done so, the new CPU line would've been productive and quite more successful than it is now. I think the Radeon products pretty much keep AMD afloat for now...
Even better, the GPU division is what made AMD *profitable* - marginal profit IIRC (i.e. in the single millions ) but profit nonetheless, after a hundreds of millions loss streak for 5 years or so
Perhaps some of the GPU engineers of ATI should crossover to the CPU department.
You forget Intel have some of the finest fabs in the world at there disposal. AMD don't have access to such facilitys. so they need more than just engineering help
Corsair 750D :: 750W DPS-G:: Asus x370 PRO :: R7 1800X ::16gb DDR4 :: GTX 1070::525gb SSD::Coolermaster 240MM AIO::
Slizza makes a good point. Intel probably has not "some of the finest" but just "the finest". They're always the first ones to take things to a smaller node, which shows just how good they are.
That said, GlobalFoundries are ramping things up lately, as they can now obviously focus on just that one goal. TSMC seem to be getting overwhelmed the last 2 years, they're having problems keeping up.
Tahiti is the proper GCN next-gen Radeon GPU's, not the low-mid-mobile gunk that was shown earlier.
AMD's long awaited next-gen 28nm GPU is finally set to release in January 2012, according to a report from Fudzilla. AMD announced during its latest financial conference call that the first next-gen GPUs were to ship by the end of 2011. Whether this shipment was for a Dec 2011 launch or in anticipation of an early 2012 remained a matter of speculation. The first next-gen GPUs to release is reported to be based on the high-performance Tahiti chip.
Will be interesting to see what a full-node shrink can do (straight from 40nm to 28nm since TSMC blew 32nm, the node meant for Cayman series).
Signature/Avatar nuking: none (can be changed in your profile)
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
It would be Quad-SLI and that doesnt work on like 99% games properly 
There has been rumours that 7970 single gpU will be as fast as 6990 DUAL GPU. Personally i dont believe this, cards are gonna be fast but not that fast for sure.
