Theoretical performance of 30 vs 20 series cards.

Talk about Hardware used for Deep Learning
Post Reply
User avatar
swapration
Posts: 6
Joined: Thu Sep 10, 2020 1:21 am

Theoretical performance of 30 vs 20 series cards.

Post by swapration »

The 3070 is supposed to have equal FP16 and FP32 performance at 20 TFLops each compared to 2080 ti which has a higher FP16 of 26 but a lower FP32 at 13. Seems like a 'jack of all trades' vs 'master of one' scenario.

Which theoretically should have a higher impact on performance in faceswap - 20 and 20 vs 26 and 13?

User avatar
bryanlyon
Site Admin
Posts: 467
Joined: Fri Jul 12, 2019 12:49 am
Answers: 39
Location: San Francisco
Has thanked: 3 times
Been thanked: 112 times
Contact:

Re: Theoretical performance of 30 vs 20 series cards.

Post by bryanlyon »

Hypotheticals like this aren't really feasible to discuss until the card comes out and we test it. That said, if you want to talk about this sort of thing, feel free to drop by our Discord.

User avatar
swapration
Posts: 6
Joined: Thu Sep 10, 2020 1:21 am

Re: Theoretical performance of 30 vs 20 series cards.

Post by swapration »

There's some tensorflow benchmarks out for the 3080. Which can be found here:
https://www.pugetsystems.com/labs/hpc/R ... tiveCharts

Looks like I failed to read the note below the chart. The 3080 was only tested against an RTX Titan using tensorflow 1.15 and the 20 series cards where all run using tensorflow 1.13.

That said since the 3080 placed just under a RTX Titan, comparing the RTX Titan's performance in v1.15 to v1.13, my guess is that the 3070 performs between the 2080ti and the 2080.

Post Reply