Fair enough, basically it's not like Kepler cards stopped playing games or something. I get your point now, I apologize again for the confusion on my part.
The only thing I said about the GTX 780 was that while it aged horribly when compared to the 290/290X, it is still relatively "good enough" for my GF who games at 2560x1080.
470 will compete with cut down GP106. 460 will be extremely low end by desktop standards (sub pitcarin level), useful primarily for mobile. As polaris showed, AMD is not significantly ahead of maxwell power efficiency either (AMD will go from being flat out completely noncompetitive in mobile to around level - AMD will not have the expected generational improvement over maxwell in mobile).
Nvidia has the market covered and will have something for the lowend by the end of the year. AMD will have nothing until 2017.
Any way you want to slice it Nvidia is releasing a full lineup, AMD is releasing lowend only.
AMD simply doesn't have the resources to do a full top to bottom refresh; they have said as much.
By looking at the GTX 1060 DX-12/Vulkan performance, RX 470 will be equal if not faster at those games at $100 less price (if you could find any $250 GTX 1060 cards) and perhaps at lower power consumption (110W TDP).
This here is right on.
Going forward in DX12 & Vulkan the 1060 will have to be compared the 470. The 480 is basically a full tier above.
Im truly surprised how behind Nvidia is in DX12, they have the market share and the money for research and development, but it looks like DX12 has blindsided them and now they are playing catch up in the DX12 department. I just hope its a one generation turn around and by next year Nvidia is rolling out strong DX12 cards.
In 6 months, Nvidia will release another card that does better in DX12, change even more and make money. People will praise Nvidia for responding so quickly.
In 6 months, Nvidia will release another card that does better in DX12, change even more and make money. People will praise Nvidia for responding so quickly.
You cannot redesign an architecture in 6 months. Nvidia's next architecture Volta is due in 2018. AMD will continue to have the lead in DX12 / Vulkan atleast till Volta arrives.
I really don't think Nvidia anticipated AMD's Mantle was going to have an impact.Im truly surprised how behind Nvidia is in DX12, they have the market share and the money for research and development, but it looks like DX12 has blindsided them and now they are playing catch up in the DX12 department. I just hope its a one generation turn around and by next year Nvidia is rolling out strong DX12 cards.
Don't be so sure that Volta is a 2018 product.
Fermi - 2010
Kepler - 2012
Maxwell - 2014
Pascal - 2016
Volta - 2018.
Nvidia has been on a 2 year cadence for 4 generations. I don't see it changing especially given that a grounds up new architecture is no easy task. Volta is expected to be a the real successor to the legendary Fermi which started out with issues but when they got fixed it was a DX11 powerhouse.
I didn't say margins don't matter. I said that isn't the primary concern right now. Making money isn't simply about looking at margins and going to sleep happy.
They need market share. That is long-term thinking. See the part of your statement that I bolded? That is exactly what I said. I never said they don't matter. Go back and read.
margins is narrow, short term thinking, which nVidia both has the luxury to rest on right now, but is also sitting on the precipice of failing because of such short-sided thinking. They have already lost the technology generation for the next 3 years with their current architecture because AMD wins with DX12 and remains competitive in the part of the market that matters for DX11 (mainstream, 1080p).
Ahh yes the Nvidia is on the verge of failing argument we have heard since the AMD acquisition a decade ago.
You cannot redesign an architecture in 6 months. Nvidia's next architecture Volta is due in 2018. AMD will continue to have the lead in DX12 / Vulkan atleast till Volta arrives.
In 6 months, Nvidia will release another card that does better in DX12, change even more and make money. People will praise Nvidia for responding so quickly.
And Vulkan will not be in most games until about 2018. The argument will be that most games are DX11 and so Nvidia wins. By the time DX12 is popular, Nvidia will release their new card. While Nvidia has been on a 2 year schedule, it does not prevent them from going to market early. If DX12 is adopted faster than Nvidia expects, they will speed up their next arch.
Never said that. I'm only saying they have lost this current generation when it comes to performance and design. I'm sure they will be fine as a company, of course. Far more than fine.
Performance and design? Seems to me that NVIDIA has designed a much more power and area efficient architecture with a better board cost structure.
Yeah now i definitely can see the Genius of AMD Going Lowend First.
1. Launch RX480 with little to no stock available anywhere.
2. Price it more than the current price of 970.
3. Weeks later still no custom boards available.
4. Competition comes out with faster card at lower price.
5. Competition already has custom cards available for purchase.
6. Profit??? (for the competition)
Yeah Now I See the Genius of AMD Going Lowend First
Not in DX-12/Vulkan, only in DX-11
Yeah you're right.In UK its cheaper but in mine its RX480~$400 and 1060~$340.With Cards in stock TODAY, RX 480 is cheaper with 2GB more Ram and Faster in DX-12/Vulkan.
If you want you can go to ocuk and buy them
Cheapest RX 480 4GB RAM in stock at 189 UK pounds
Cheapest RX 480 8GB RAM in stock at 225 UK pounds
Cheapest GTX 1060 6GB RAM in stock at 270 UK pounds
Even another reason not to buy a current NVIDIA card, because it will be obsolete even faster this time.