Slashdot Mirror
NVIDIA Unveils GeForce GTX 1080, GTX 1070, Faster Than Titan X For a Lot Less (hothardware.com)
MojoKid writes (edited and condensed): NVIDIA has unveiled its next-generation Pascal-based GeForce graphics cards -- known as the GeForce GTX 1080 and GeForce GTX 1070. NVIDIA's Pascal architecture is based on 16nm FinFET technology, similar to that of NVIDIA's high-end data center Tesla P100 processing engine though the GeForce cards are targeted at the consumer gaming market. NVIDIA's GP104 GPU at the heart of the new GeForce cards is comprised of some 8 billion transistors and features a 256-bit memory interface with 8GB of Micron GDDR5X graphics memory on the GeForce GTX 1080. The GTX 1070, however, employs standard GDDR5. The core clock speed of the GeForce GTX 1080 hit 2.1GHz at one point during the demonstration, though GTX 1070 clocks were not disclosed. NVIDIA CEO Jen-Hsun Huang claimed the new GeForce GTX 1080 is faster than a pair of GeForce GTX 980 cards in SLI and faster than the company's very expensive Titan X graphics card but at half the price. The new GeForce GTX 1080 will be offered in two versions, a standard card with an MSRP of $599 or a highly-overclockable Founders Edition for $699. The standard GTX 1070 will arrive at $379, while a Founders Edition will be priced at $449. Availability for the GTX 1080 is slated for May 27th and the GTX 1070 for June 10. Anand Tech has more information.
Think VR
90fps minimum, stereo, and high fov
Assuming their "A NEW KING" graph isn't a lie, it's $600-$700 for 1.7x the performance of a single 980. This also jives with their claim about it being faster than 2 980s in SLI.
The graph is missing the 980 Ti though, which is what people paying $600-$700 for a GPU will be comparing it to.
I can understand not including the 980 Ti since we're talking about the 1080 and not a 1080 Ti. But they went ahead and included the Titan X. That's some major bullshit - the Titan X is an expensive piece of shit compared to the 980 Ti. A reference 980 Ti is nearly identical in performance to a Titan X at $400-$500 cheaper. A non-reference 980 Ti will easily beat a Titan X and still save you $200-$300. (The Titan X is only available with the reference cooler.) The non-reference 980 Ti is also at the same $600-$700 price of the GTX 1080 (MSRP).
My guess based on their chart (using the Titan X as a baseline) is that the 1080 is about 30% faster than a reference 980 Ti and 15-20% faster than any of the dozens of non-reference designs out there now priced around $600-$700. (There are dozens more non-reference 980 Ti models priced higher and higher than that, if you've got money to burn.)
Anyone on a current generation card really should wait for the presumed 1080 Ti and AMD's Vega.
> Do video card upgrades even matter anymore?
Yes.
* VR requires 90 Hz minimum (Thank god!)
* 4K Gaming at 120 Hz requires beefy hardware.
* ENB mods
If you can't even tell the difference between 24 Hz and 60 Hz ....
OWE my eyes @ 24 fps !
Silky smooth @ 60 fps !
> That's some major bullshit - the Titan X is an expensive piece of shit compared to the 980 Ti.
Whoa, hold on. Context is extremely important.
For gaming yup, that's some serious shenanigans(*) ! BUT for rendering the Titan X is faster then the 980 Ti.
(*) Obviously, many people don't feel the Price/Performance of the Titan X vs 980 Ti is worth it, myself included. Words along Over-priced, Greedy bastards come to mind, but if performance is king and money is no object then for rendering + scientific computing, the Titan X was the previous crown holder.
It all depends on context.
AMD gets paid every time Intel sells an x64 chip. Tired of hearing this crap. AMD isn't going anywhere including broke. Go download a 64 bit iso for Linux. I don't care what distro it is. If it is 64 bit then it is labeled AMD64. Wonder why that is. Maybe you should Google that and learn something.
X-Plane 10, with AIPilotX's HD Mesh v3 and the Massachusetts Pro VFR Scenery will eat up 32 GB of RAM easily.
Flight simulation is one of those areas where I doubt there will ever be such a thing as enough memory. There's always something to model in greater detail.