By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - PC - Carzy Zarx’s PC Gaming Emporium - Catch Up on All the Latest PC Gaming Related News

Captain_Yuri said:

Here is Anandtech's take on FSR:

https://www.anandtech.com/show/16723/amd-formally-unveils-fidelityfx-super-resolution-open-source-game-upscaling

After reading this I now expect best case for FSR is going to be as good as DLSS 1.0 which was not very good.  Well hopefully that mean by FSR 2.0 they finally have a useful alterative to DLSS 2.x but look like we going to have a long wait for a competitive alternative.

I am excited for what AMD could be releasing in 2022 with zen 4 + 3D V-Cache Stack.

On APU side hopefully this is a sign of things to come because am not interested in 5xxxG but a future 6xxxG with RDNA2 (and hopefully that mean native HDMI 2.1) I would defiantly be interested in a super small form factor computer to hook directly up to my TV for streaming uncompressed ultra Blu-ray files etc for maintaining max visual and sound.



Around the Network



                  

PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850

RDR2 getting DLSS
Doom Ethernal getting Ray Tracing and DLSS

Lots of other games getting Ray Tracing and DLSS and Nvidia Reflex

Not much other news.

I'd skip the 3070 Ti and get a 3080 instead. 8GB vram for $600 is way too low imo. I guess $1200 is the new MSRP for the Ti series going forward which is urghh!



                  

PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850

$1200 MSRP for 3080TI wonder if that was always the plan or they increased it because of everything going on. I would of expected a $1000 MSRP because that what 2080TI was and for the 3080 they kept msrp same as launch 2080 price and the jump from 2080 to 2080TI is probably bigger then jump from 3080 to 3080ti.

Not that it really matter since I doubt anyone getting one for any where close to the MSRP.



Captain_Yuri said:

I'd skip the 3070 Ti and get a 3080 instead. 8GB vram for $600 is way too low imo.

Back when the 3070ti was rumored to have 16GB of vram I was seriously considering it, but not with 8. The regular 3080 still looks like the best bang for the buck if you can get it for a reasonable price. At this point I may end up skipping the gen or buying once it becomes last gen, though.



Around the Network
TallSilhouette said:
Captain_Yuri said:

I'd skip the 3070 Ti and get a 3080 instead. 8GB vram for $600 is way too low imo.

Back when the 3070ti was rumored to have 16GB of vram I was seriously considering it, but not with 8. The regular 3080 still looks like the best bang for the buck if you can get it for a reasonable price.

Yea pretty much. This is the one thing that annoys me is that the laptop 3080 has the same cuda cores and etc as the 3070 Ti but there is an option for 16GB of Vram. Like it's much slower sure but still, quite urghh worthy!

This gen is basically like this imo:

Best bang for buck:

3060 Ti
3080
6800 XT

Meh:

3060
3070
6700XT
3080 Ti
6900XT
3090

Skip:

3070 Ti
6800



                  

PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850

Cyran said:
Captain_Yuri said:

Here is Anandtech's take on FSR:

https://www.anandtech.com/show/16723/amd-formally-unveils-fidelityfx-super-resolution-open-source-game-upscaling

After reading this I now expect best case for FSR is going to be as good as DLSS 1.0 which was not very good.  Well hopefully that mean by FSR 2.0 they finally have a useful alterative to DLSS 2.x but look like we going to have a long wait for a competitive alternative.

I think that'll depend on how it looks in motion. There has to be a reason for AMD to be confident in their spatial scaling implementation if they already knew the kind of temporal issues Nvidia was getting with DLSS 1.0.



 

 

 

 

 

haxxiy said:
Cyran said:

After reading this I now expect best case for FSR is going to be as good as DLSS 1.0 which was not very good.  Well hopefully that mean by FSR 2.0 they finally have a useful alterative to DLSS 2.x but look like we going to have a long wait for a competitive alternative.

I think that'll depend on how it looks in motion. There has to be a reason for AMD to be confident in their spatial scaling implementation if they already knew the kind of temporal issues Nvidia was getting with DLSS 1.0.

The reason could be it is the best solution they can do at the moment without delaying it another year and they made the calculation getting something out now even if it not great is better then having nothing at all.



Cyran said:
haxxiy said:

I think that'll depend on how it looks in motion. There has to be a reason for AMD to be confident in their spatial scaling implementation if they already knew the kind of temporal issues Nvidia was getting with DLSS 1.0.

The reason could be it is the best solution they can do at the moment without delaying it another year and they made the calculation getting something out now even if it not great is better then having nothing at all.

They could have went for checkerboard rendering and temporal upsampling if that was the case. These are already fairly comparable to DLSS 1.0 IMO.



 

 

 

 

 

Honestly think its pretty great. A more open solution for more platforms is what we need. It might not look like a upgrade from standard TAA or DLSS 1.0 rn though the benefit of much better performance and not hardware-agnostic is a big plus. That alone will give it a huge boost on uptake compared to proprietary technologies. Instead of companies doing their own solutions we have a universal one that will be easier for smaller teams to implement and something the industry should strive for.

Tbh with the way GPU's sitcu is still going people (like me) who have Pascal cards, this is quite a boon to have.

Last edited by hinch - on 01 June 2021