Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
you can record from your TV output maybe depending on the TV
but ? still need some kind of input into the pc. mebbe something usb
well from what I read online you can record it on your XBox with voice commands
Not sure. You should look that up to see how others do it as it is possible.
This seems like a wasted of money and overkill. Why would you need such a card just for the monitor/s? Enlighten me.
it can run at least 4, I actually have 3 on mine plus a TV if needed
I have the Gigabyte GTX 970 G1.
moar pr0n
https://www.nvidia.com/en-us/titan/titan-v/?ncid=so-you-tnv-29059
with video
ice cream with pron
wonder what it will be like with rendering?
my nvidia is a 2gb GT730. in my i7.
was thinking of another 730 for my dualcore AMD pc,
My experience is that brute force GPU simply doesn't work. No matter the rig. Of course the better the rig the faster the rendering. But proper scene optimization and some post-processing is the only way to go for fast rendering.
...yikes, just saw the writeup on the Tom's site and 3,000$ for only 12 GB (OK, HBM 2 instead of GDDR5X), apologies not worth it unless your pastime is scientific modelling or AI development (even though unlike the Tesla V100 it has video outputs). 13.8 TFlops of single point performance and 5,120 CUDA cores means nothing if your render file exceeds 12 GB in size.
Now, if you look at the promo shots of the card, there is a nice big NVLink plug on the side... (This could be foreshadowing what will be on the GTX xx70 and GTX xx80 Volta gaming cards.)
From what I can understand the nvidia unified memory is already done with Kepler. And it will improve with Pascal. However, even if memory caching is a big step forward indeed, it may not always be a solution depending on the case. If the memory access is localized, then caching works just fine. If the memory access is not localized, then caching benefits may be overwelmed by the transfer time, or access latency.
So, to achieve good performances, a good scene optimization may be necessary even with the unified memory.
https://devblogs.nvidia.com/parallelforall/beyond-gpu-memory-limits-unified-memory-pascal/
EDIT to be clear what I mean is: if the scene exceeds the GPU memory limit and you render the whole scene, not just a part of it, then I would expect the unified memory model to have a huge drop in performances.
...again you are talking to someone who builds big scenes and renders in large resolution format so not only is preserving detail quality important, but the time it would take to resize a large number of texture files in a 2D programme runs into diminishing returns.
Even if Nvidia has plans to bring NVLink to the Titan V, I still have doubts they would shoot themselves in the foot by competing against their own Quadro line through introducing it into their lower cost consumer cards. Also notice that the Titan V still only has 12 GB (not 16) of HBM2 even though it uses the same graphics processor and setup as the Volta Tesla (which has 16). Linking two Titan Vs would give the one the same memory resoruces as the current P6000 (24 GB) but for an additional 2,200$ (which includes 1,200$ for the set of two NVLink connectors required).
By doing this, I feel Nvida is showing they are being very aware not to overshadow the current or next generation of their more expensive pro grade Quadro line. The x5000 and x6000 Quadro series has also quite possibly reached the end of the line with Pascal. Most likely the Volta architecture will usher in an entirely new series with possibly a 16 GB (GV100?) and 32 GB (GV200?) Quadro respectively as their two flagship Pro grade units (also most liely at a higher price than the Pascal versions as well fiven the jump in cost with the Titan V).
Imagine having a total of 64 GB of VRAM with 2 NVLinked Volta 32 GB Quadros running over 10,000 cores. Even my scenes would be hard pressed to challenge that.
The HBM2 Memory that Nvidia is using right now comes in 4GB chips. Tesla V100 = 4 chips, and Titan V = 3 chips.
I'm guessing that when Nvidia releases the Volta based Quadro cards they will introduce 8GB HBM2 chips for the Quadro V5000 and V6000, and drop the 4GB chips down to the Geforce cards for use in the GTX 1170 and 1180..
On a side note... The Titan V may just be defective Tesla V100 cards that have one bad memory chip on the GPU.
I will have this please
and a windfarm for electricity
Seriously. Looks like dForce hair has finally met its match.