RTX 3060 = 2080 Ti for Rendering!
Therewolf
Posts: 3
According to Guru3D the RTX 3060's Vray rending perfermance is on par with the RTX 2080 Ti:
https://www.guru3d.com/articles_pages/asus_geforce_rtx_3060_strix_gaming_oc_review,31.html
For $350 you can get the rendering perfornce of a $1,200 card? With 1 extra gig of video RAM? That's impressive, and hopefully this Vray preformance translates to Iray performance. I think this makes the RTX 3060 the best starter card in quite some time. I know stock is going to rough for a little while, but I'm excited to try two of these in my rig.
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
When do you think supply / demand will normalize? Prices are a bit crazy at the moment!
Well, I'll be able to tell you about Iray in DS specifically sometime today, as I happen to have a shiny new 3060 sitting on my desk right now. (Just waiting on a new power supply to arrive, as my current one is only specced for a 75W card).
3060 = 8gb
2080ti = 11gb
3060 dual comes with 12gb, 3060 ti version has 8gb
Explanation of the difference between the two and benchmark comparisons here
I'm still mad… but buy it anyway - RTX 3060 Review - YouTube
Now that the cheapest of the RTX 30xx lineup has 12GB:s, where are the 3070 and 3080 ones with more VRAM?
And I've just tested my new 3060 with the forum benchmark; at 8.144 iterations/sec, it is indeed on basically the same tier as the top 20 series cards - beats the 2080 Super, not *quite* a match for the 2080 Ti, but staggeringly close for the bottom card (so far) of the 30 series.
Lol, just checked amazon for my country, only one available is priced at 1300 + tax and shipping.
On that note it would be really great if any 2080ti owners, and Turing owners in general could run the Daz benchmark scene again with 4.14 or 4.15 for a proper comparison.
Give me a minute to get the benchmark scene and set it up for reference and I'll post the result for a RTX 2080 TI
Very interesting, so there is a hope for some reasonably priced Nvidia graphics card for iray in Daz Studio.
Of course, it all depends on Nvidia and the graphics card manufacturers,
if they will find enough resources to produce satisfactory number of such cards for everyone interested.
3060 12GB review:
https://www.tomshardware.com/news/nvidia-geforce-rtx-3060-review
Linus going harsh on NVidia:
https://www.youtube.com/watch?v=XfIibTBaoMM
That's great to hear. I'll be replacing two 1070s with the 3060s. I feel that the performce + 12Gb VRAM = best bang for your buck right now. Plus hopefully they'll be easier to buy since NVIDIA nerfed them for mining.
Appearently NVIDIA has delayed all new unannouced cards indefinitly until they get stock issues under control for their current cards. So we might not see a new higher VRAM 3070/3080 until 2022. Nothing's been offically annouced though, so only NVIDIA knows for sure.
We got a new 2080ti time up and it is a few tenths of an iteration faster than the 3060. The 2080ti scored 8.75 iterations per second compared to the 8.1 of another user. Of course this is just one 2080ti compared to just one 3060. Different models with different overclocks and coolers can impact these. And different scenes may also effect rendering. It is possible that the 3060 can handle complex geometry better than the 2080ti, so more complex scenes might push the 3060 on top. The reason I believe that the 3060 might do this is because of the fact that the 3060 is even competing with the 2080ti at all. In any other metric the 2080ti should destroy the 3060, and it wouldn't be close. I'm not sure how we can test it. Unless the 2080ti and 3060 owners would want to try one more test, the dforce strand hair test that was posted a long time ago. I think that would be interesting to see with any Ampere cards. Anybody remember that?
Still, that is a fantastic showing for the 3060. Assuming the MSRP of $329, that compares pretty well to the $1200 for the 2080ti, LOL. And you get 1 extra GB of VRAM to boot.
If you can point me at where the test is, I'm up for that. (I'm not sure how many other 3060 owners we yet have on the forum. I was a lucky sod to get my hands on one at all).
We don't really have a lot of data about how the RTX speed-up varies with the content of the scene; I wouldn't be surprised if the 2nd gen RTX cores mean that in a different test the 3060 might actually edge ahead of the 2080 Ti, but it is a guess.
You can't do that.
https://www.tomshardware.com/news/nvidia-to-support-nvlink-only-for-rtx-3090
Only the 3090 has NVlink for some reason. I discovered this only after I bought a 3080. Great card but it would've been fantastic to be able to add the extra video memory on top.
You can still have two (or more) cards in a rig without NVLink, it is just they will not be able to pool their VRAM. However with a massive 12 GB of VRAM in a 3060 this should be able to handle most scenes that people create, and thus two cards would render them a lot faster than one.
How can two GPU's pool their processing power together but not be able to pool their vram together?
That's always been the case.
Even without SLI/NVlink?
Yes. NVlink is only used for (texture) memory pooling.
I did not know that. I looked into it and apparently SLI/NVLink is designed for gaming specifically, I assumed you had to connect the GPU's between them by default to get them to work together for anything. I guess I might still be able to upgrade in the future.
I found it!
https://www.daz3d.com/forums/discussion/344451/rtx-benchmark-thread-show-me-the-power/p1
This was a cool little bench, the Turing cards were putting up crazy numbers compared to the GTX cards. So I bet Ampere would do numbers that are just absurd.
Oh wait, the download for the scene is gone. I'll try to look at see if I have it and upload a copy.
Ok, I uploaded the scene save file to sharecg. So you can download it there.
I went ahead and ran the bench with my 1080tis, one alone, and then another together. What is interesting is that I did not see any speed increase with 4.15 like we did with the other bench scene.
https://sharecg.com/v/97260/view/21/DAZ-Studio/Iray-RTX-Benchmark-Scene
I think I see why, this scene has no normal maps at all. But the rayDant scene does have normal maps on the outfit. The notes for 4.14 does mention that a change was made to make normal maps more efficient. So I believe this confirms that the speed increase observed in Daz 4.14 is directly tied to the scene using normal maps.
This also means that we can indeed directly compare any new numbers posted to the old numbers posted when this thread was created. That is excellent.
Well, I've just run it, and while 31.993s is a valiant effort (at 18.754 iterations/s), it is still slightly beaten out by the ~28-29s average benchmarks reported for single 2080 Ti cards (assuming those results are still valid) - again at about nine-tenths of the performance; so it seems that the specific geometry of the scene may not actually have much effect on relative Turing/Ampere performance.
It's possible that might change in a few weeks (the 3060 has only been out three days, and the first set of drivers available at launch aren't necessarily as good as the card can get), but for the moment the conclusion would seem to be that the cards aren't exactly equivalent.
Still, considering the 3060's lower MRSP, power consumption and slightly higher VRAM, the only strong argument I can think of for the 2080 Ti as an Iray card is that it supports NVlink. (Of course, when it comes to gaming, the 2080 Ti is leagues ahead, but that's a different kettle of fish).
You can have 2 in your rig - they just can't share RAM
Does anybody got this card and has checked its performance in Daz Studio iray, yet?
I have heard, that the VRAM in RTX 3060 is not as good as VRAM in RTX 3090.
Well the 3060 has 12GB while the 3090 has 24GB, so that is an obvious difference. The 3060 has slower VRAM as well, but that makes absolutely no difference to Iray. It could have DDR3 memory and it wouldn't matter. The memory speed pretty much only effects gamers, and some other applications that are sensitive to memory speed. But Iray is not sensitive to VRAM speed whatsoever. So for Iray this is not even a consideration. The only consideration is how fast the cards renders and how much VRAM it has.
Thanks for the explanation.