http://www.tweaktown.com/news/52996/nvidia-culls-geforce-gtx-1060-3gb-arrive-1050-december/index.html
>1060 3gb rebranded into $170 GTX 1050
>1060 6gb now cost $250
AMD status?
>>55562867
I am not a shill so I will say nothing
>>55562867
now? as in, the founders edition is 250 now? bc the msrp is 250 for the 6gb. if true, pretty interesting. tired of waiting tho
>>55562951
>>55562867
But don't nvidia cards get gimped over time? Why would I want something that could last a while potentially but won't because of drivers.
so its GTX970 with 0.5GB less ram for $170?
>>55562991
dunno. im new to this whole gpu circus because i wanna upgrade. depends on what all the shills on /g/ wanna sell you
AMD BANKRUPT AND FINISHED
>>55562991
Nvidia cards don't get gimped, only AMD gimps cards
>>55563051
>>55563060
Well I just found this in >>>55560947
take a look at that 780 and the ti version, a fucking 960 is tiers above it. Remember when nvidia told maxwell user to wait for async compute drivers in the beginning of the year? Nothing yet. That's the problem, nvidia offers great performance, they're just too jewy to let you keep it for too long. It's ironic that a 290x now outperforms a 780ti just because of Nvidia's gimping.
>>55562991
It's already gimped. No dx12, 192bit bus, no sli.
>>55563110
Notice something?
AMDummies must be crying themselves to sleep these days
>>55563127
I definitely noticed how Nvidia shitcanned unified memory for Maxwell, and its still not present in Pascal either.
Why the fuck are they releasing fucking 1050 before the 1080ti?
>>55563152
The New Titan and 1080ti make the 1080 look like a fucking 980 in comparison. They won't release 1080ti even though it's just sitting there, ready...until they milked all the money from inferior...cut down gtx 1080 and new Titan. Cunts.
>>55562991
Nah, nVidia doesn't gimp their cards. Their hardware is TOO perfect. So perfect that they reach their peak performance just before the next-gen hardware is released to retail. After that, don't expect anymore from deprecated EOL hardware.
>>55563127
FP64 was shitcanned, Unified Memory and Stacked DRAM are still undelivered
Literally how, did they really decided to scrape GPGPU R&D for muh gaems?
>>55563147
Pascal is a rushed product that was totally unplanned. Literally a die shrink and some software changes. Nvidia is really having trouble with the next gen, it seems.
>still can't into proper DX12/Vulkan specs
Enjoy your outdated crap.
>>55563233
Looks like fp64 is only for top tier products now (GP100) and not even halo game cards GP102 are expected to have it, suspect 102 will be same fp32core count as the 100 though. Now if that meant cheaper prices for us it'd be great, but it won't. Going to be funny when another titan comes out with no fp64 again.
AMD on the other hand might not be able to afford two top tier products and have their top end gaming chip be the same as their compute card.
>>55562991
No, I'm pretty sure whatever gimping you have you get straight from the factory.
>>55563219
>Nah, nVidia doesn't gimp their cards.
3.5gb ring any bells?
>>55563060
>>55563219
Come on faggots, how much do you get paid?
>>55563390
and before that on kepler cards wasn't there a software side gimp they later fixed?
>>55562991
>>55563051
This is now Nvidias cards get gimped. Its the way they build their architecture. When they go from Kepler to Maxwell they create an entirely new chip set. This means new new drivers will be geared to Maxwell...because it's new so you wont see much support for older cards you will most likely see a decrease in performance because of the drivers. Why does Nvidia do it this way? Nvidia just wants a new **efficient chip** set from scratch.
When AMD creates a new architecture they add in new components. This is why you will see GCN1, GCN2, GCN3 and 4, it's still GCN. When a new driver is made the some of the same components you see in GCN1 is still there in GCN4. So you will still see support and improvements on older AMD cards. Why does AMD do it this way? **Better long term support** and they don't break any thing. This is also why AMD chips draw more power. The have components that have been brought along from the previous GCN generation plus new components.
Which is better? Long term AMD short term Nvidia.
>>55563711
>better long term support
No, they do it because it costs less and they love rebranding shit to sell to you
>>55563110
What kind of fucking ass-backwards scaling are the bars in this graph anyway? It's Nvidia marketing department bullshit tier.
The top 1080 bar that represents an increase of 23fps (144->167) is bigger than any other bar on the graph by magnitudes for no fucking reason at all. The bottom card 760 has a 27fps bar that is microscopic. The whole chart is fucked
>>55562867
AMD in Zen+ (or whatever) should introduce double-precision decimal FPU then they would totally shit on intel in scientific applications.
because then you don't loose precision that much.
