View Single Post
Old 05-21-12, 04:06 PM   #31
i SPY
007
 
i SPY's Avatar
 
Join Date: Apr 2007
Location: You were sayin'
Posts: 290
Default Re: GK110 to feat a 512 bit bus ?

^
Hopefully, i dont really know, but if i go with GF110 road then yes. Im betting on Q1-Q3 2013, at least that's what my gut is telling me

This tesla version will be apparently Q4 2012.

I got those 2 slides from here,
http://www.nvidia.com/object/nvidia-kepler.html

there is also 23page whitepaper, describing HyperQ and dynamic parallelism among other new features. And imo this only confirms how crippled GK104 really is, they had to use 690GTX dressed as tesla K10 to somehow fill the gap until then..

Quote:
Originally Posted by ViN86 View Post
Dynamic parallelism is a great idea. It will work well for API calls that have well defined layouts (i.e. one function calls the next calls the next, etc.). It will certainly help with slower CPU's, but the feature will be very dependent on driver optimization.
Agree, hopefully it won't be a flop like dynamic vsync (very buggy atm). Im also very interested in HyperQ - it will use higher gpu usage, thus more powah and in geforce variant it could mean more fps ^^

Quote:
Hyper-Q
Slashes CPU idle time by allowing multiple CPU cores to simultaneously utilize a single Kepler GPU, dramatically advancing programmability and efficiency

Now if they skip both then yeah i dont really see a point in GK110 variant, other then 15smx and higher memory bus..
Although it would fill this big gap between 680 GTX and 690 GTX, hopefully they're only gonna slash away vram size and leave GK110 intact.
__________________
intel Q9450 @ 3.656Ghz [1.3875v, LLC off]| GA-X48-DS5 [Memory Enhance: Turbo]|MSi N570GTX TwinFrozrIII OC/PowerEdition|Kingston HyperX 4x2GB PC 8500 @ 1097Mhz [5-5-5-18, 2.25v]| Creative X-FI Pro [SB046A]| Tagan PipeRock 600w [48A]

i SPY is offline   Reply With Quote