Hey,
I wanted to post this on the video cards sub forum but seeing that forum is dead posting here instead. feel free to move the topic if necessary.
I got a 3080 a month ago and really impressed with it's performance. Currently im running 3080/980ti/1070ti setup and want to replace 980ti as well.
Im eyeing at 3060ti as it's better price/performance where I am and better thermals. But the memory difference is making me ponder about getting another 3080 instead.
3080 has a ddr6x memory with 320 Bit bus width and 760.3 GB/s bandwidth while the 3060 is 256 Bit and 448.0 GB/s respectively. Also there is a 2gb difference in mem capacity.
How would this affect the performance? I figure it will affect the scene loading time onto the GPU at first but in animation would it make a big difference? A 2gb upgrade is fine but I dont see myself creating bigger scenes than I normally do.
Thanks.
3080 vs 3060ti
Forum rules
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
I think the best way to gauge performance is to look at the actual benchmark numbers. Loading something into an 8GB pool will in my opinion not be any faster because of faster VRAM - PCIe 4.0 compared to 3.0 sure but even then I'd be a serious doubter you could ever even prove it as even x8 PCI-E 3.0 delivers 8GB of bandwidth. I can imagine the situation being different on a 48GB card.
Personally, I just stick to the benchmarks when dealing with "unknown" hardware I've never owned before. And do check out all of the benchmarks, CUDA and RTX and other renderers. If a certain card is 2x faster than the other then no matter the bus width or core clocks it will be about ~2x faster.
Do note though that whenever possible check out benchmarks that are the most relevant to you - if you are doing logo mograph type of work then checking an RTX benchmark scene won't be a proper representation for your workflow.
Personally, I just stick to the benchmarks when dealing with "unknown" hardware I've never owned before. And do check out all of the benchmarks, CUDA and RTX and other renderers. If a certain card is 2x faster than the other then no matter the bus width or core clocks it will be about ~2x faster.
Do note though that whenever possible check out benchmarks that are the most relevant to you - if you are doing logo mograph type of work then checking an RTX benchmark scene won't be a proper representation for your workflow.
Thanks for the reply. I decided to go with 3080 again just to keep the 10gb vram advantage. I didn't factor in the win 10 vram grab when I was considering the 3060ti.
I do a wide variety of animation work which includes open terrain landscape work with hdri lighting and the rtx boost on those scenes are crazy.
I do a wide variety of animation work which includes open terrain landscape work with hdri lighting and the rtx boost on those scenes are crazy.
Honestly, with the rise of photoscanned assets and just larger texture sizes in general (which is good imho) having a larger VRAM pool is really useful if you ask me. Unless you really know you never get OOC / or get close to utilizing the VRAM pool on a 6GB / 8GB card because your projects are that small in terms of assets, I think opting for the higher VRAM option makes sense.
That said, if you need to get projects out the door asap and you can't really justify spending 500$ on a card, well, the 3060ti is an awesome option.
Guess it depends
Anyway, hope you'll have fun with it!
That said, if you need to get projects out the door asap and you can't really justify spending 500$ on a card, well, the 3060ti is an awesome option.
Guess it depends
