Anonymous at Tue, 15 Feb 2022 19:37:48 UTC No. 882618

>>882616
96

Anonymous at Tue, 15 Feb 2022 20:42:36 UTC No. 882647

>>882618
sounds like bs

Anonymous at Tue, 15 Feb 2022 23:00:04 UTC No. 882689

>>882647
Educate yourself kid

Anonymous at Tue, 15 Feb 2022 23:16:41 UTC No. 882702

>>882647
blendlet spotted.

Anonymous at Tue, 15 Feb 2022 23:18:52 UTC No. 882704

>>882616
A 3090 is the sweet spot for rendering currently, supports nv-link for vram pooling too.
>>882689
A6000 is good but only if you absolutely need 48 per card. The 3090 provides the same horsepower (cuda cores, RT cores, tensor cores) for half the price.

Anonymous at Tue, 15 Feb 2022 23:25:29 UTC No. 882706

>>882704
I have two in my workstation (at work). Privately I use two 3090. I also do not use blender.

Anonymous at Tue, 15 Feb 2022 23:48:43 UTC No. 882711

>>882706
Do you think the 3090 is getting a little old at this point? Its been out for 16 months and they say that nvidia is launching new cards at the end of this year. Should I be concerned about that?

Anonymous at Wed, 16 Feb 2022 03:28:04 UTC No. 882754

>>882616
Rendering what?
And in what timeframe?
You can render whatever you want with proper management and enough time, using some old ass shit.

Anonymous at Wed, 16 Feb 2022 03:37:01 UTC No. 882758

>>882689

Anonymous at Wed, 16 Feb 2022 04:04:29 UTC No. 882762

>>882754
heavily tiled udim 4k textures on displaced objects with enough speed in the render preview to make quick judgments and not have to wait all day

Anonymous at Wed, 16 Feb 2022 04:12:06 UTC No. 882763

>>882762
why would you ever need that?

Anonymous at Wed, 16 Feb 2022 04:23:16 UTC No. 882765

>>882763
what ? SP has udims and film quality expects tons of tiles along with displacement. Are you trying to stir the pot?

Anonymous at Wed, 16 Feb 2022 04:30:56 UTC No. 882766

>>882765
>he still uses displacement

Anonymous at Wed, 16 Feb 2022 04:32:21 UTC No. 882767

>>882766
yeah, i use displacement because i animate in maya and cant have 50 million tris on a mesh im trying to animate

Anonymous at Wed, 16 Feb 2022 04:33:43 UTC No. 882768

>>882767
>he thinks it's a matter of geo vs rendertime displacement

Anonymous at Wed, 16 Feb 2022 04:35:02 UTC No. 882769

>>882768
why are you on this board

Anonymous at Wed, 16 Feb 2022 04:35:57 UTC No. 882770

>>882769
>he's butthurt that he's not ahead of the curve

Anonymous at Wed, 16 Feb 2022 04:39:37 UTC No. 882771

>>882770
you are obviously a lot smarter than everyone in the world combined so please enlighten us with your knowledge senpai so we can all learn from you

t. 155 iq brainlet

Anonymous at Wed, 16 Feb 2022 04:41:48 UTC No. 882772

>>882771
https://perso.telecom-paristech.fr/boubek/papers/TFDM/
There, now you can start researching on your own, lazybones.

Anonymous at Wed, 16 Feb 2022 05:00:55 UTC No. 882775

>>882772
can you provide me with the vulkan source anon

Anonymous at Wed, 16 Feb 2022 13:00:34 UTC No. 882854

>>882775
Where is the source?

Anonymous at Wed, 16 Feb 2022 13:43:50 UTC No. 882863

>>882775
the recipe is right there
just write it urself bro

Anonymous at Wed, 16 Feb 2022 13:49:20 UTC No. 882865

>>882863
I will implement it into my engine and I look forward to seeing your version as well. What's stunning is how much detail they could get so much out of only 35mb of memory versus over 2gb of memory for displacement on a single object. I guess this goes to show how powerful Adobe's team is rn

Anonymous at Wed, 16 Feb 2022 14:08:46 UTC No. 882871

>>882711
Two 3090 are more than enough for day to day work, unless you do crazy large simulations or load textures in the hundreds of GBs

Anonymous at Wed, 16 Feb 2022 20:29:43 UTC No. 882979

What do you think of the 3070 8gb for $999?

Anonymous at Thu, 17 Feb 2022 07:12:48 UTC No. 883114

>>882616
Just get a 90, literally 0 issues if rendering is your thing. Plus the speed, my god. But I did scalp a 60ti for 1500 so it defrayed the cost of the 90 considerably.

Anonymous at Thu, 17 Feb 2022 11:11:42 UTC No. 883137

>>883114
I pulled the trigger on a 3060 for $780, the 2 fan variety. It turns out that 3 fan gpus wont fit into my case so if I were to get a larger card I would have to upgrade my case which would be a nightmare. Also I got it because of the relatively high amount of memory and from all the benchmarks ive seen it has 50% the perf of a 3090 ($3090) for only 30% of the cost. I've seen videos of people doubling up 3060s to get 3090 like perf at only 60% of the cost and i can do this because i have 2 pcie slots and have done this in the past with much older cards. Also, getting a 3060 means im not that commited to the 30 series and still have lots of money left over to upgrade to something nicer if/when the 40 series comes out. I'm coming from a 970, with 3.5gb vram.

Anonymous at Thu, 17 Feb 2022 11:21:34 UTC No. 883141

>>883114
I pulled the trigger on a 3060 for $780, the 2 fan variety. It turns out that the length of the 3090 wont fit into my case so I would have to upgrade my case which would be a nightmare. Also I got it because of the relatively high amount of memory and from all the benchmarks ive seen it has 50% the perf of a 3090 ($2800) for only ~27% of the cost. I've seen videos of people doubling up 3060s to get 3090 like perf at only 60% of the cost and i can do this because i have 2 pcie slots and have done this in the past with much older cards (2x 970 although i only run 1 970 now). Also, getting a 3060 means I'm not that committed to the 30 series and still have lots of money left over to upgrade to something nicer if/when the 40 series comes out. I would love to have 24gb vram but again, a 3090 from my measurements simply will not fit in my case so i would have to rip everything out and get a new case and have to install everything again or build a brand new pc with a new mobo which would be a real pain, especially this far into the life of the 30 series.

Anonymous at Thu, 17 Feb 2022 11:43:45 UTC No. 883143

>>882979
wow, the GPU market is really fucked...

Anonymous at Thu, 17 Feb 2022 22:26:13 UTC No. 883239

>>883141
Vram will be limited to whatever size that is even with two since it can’t nvlink. I’m guessing 12gb but you can double it up for speed, but if you spent 780 for 2, might as well see if you can find a 90 if you’re by a Microcenter at that price and actually get 24gb, but if it doesn’t fit your case you do you. Cheap nzxt cases will work just fine honestly. Though I had a 70 that was mostly limited by 8gb vram, maybe 12 is just enough for most hobbyists.

Anonymous at Fri, 18 Feb 2022 03:31:36 UTC No. 883289

>>882772
>almost no increase in memory usage
well that's nice, now we can ign-
>dynamic bvh increases render time by 5870%
ah shit nigga what are you doing

Anonymous at Fri, 18 Feb 2022 12:26:57 UTC No. 883327

>>883289
they say it right in the paper that as compute continues to improve the hit will be less over time

Anonymous at Fri, 18 Feb 2022 13:50:49 UTC No. 883341

>>883289
Relative to its already substantially decreased render times.

Anonymous at Sat, 19 Feb 2022 15:43:44 UTC No. 883524

I might have fucked up. I ordered a new cpu - a 12900k but now i dont know if renderman is compatible. Any help?

Anonymous at Sun, 20 Feb 2022 01:37:28 UTC No. 883610

Is maya compatible with windows 11? So far, the only program i see with official suppor for windows 11 is Blender 3.0. How can free software be so far ahead? I look on autodesk forums and they are angry at no official support and that is paid software.

Anonymous at Mon, 21 Feb 2022 13:55:49 UTC No. 883896

>>883524
>>883610
https://start.duckduckgo.com/