With our review of the GeForce GTX 1070 Ti going up moments ago (we got to test the ASUS ROG Strix version) we thought we'd share the before and after benchmark results. And that's because when our new GPU arrived all we did was replace the existing Nvidia GeForce GTX 1070 Founders Edition with the fancy new Ti version.

Which means all other hardware stayed the same.

Here are the 1070 Founder Edition results.

http://www.ausgamers.com/gameres/5549/images/source/1070founders1.jpg

http://www.ausgamers.com/gameres/5549/images/source/1070founders2.jpg

And here are mostly the same tests with the ASUS ROG Strix GeForce GTX 1070 Ti

http://www.ausgamers.com/gameres/5549/images/source/asusrog1070tibench1.jpg

http://www.ausgamers.com/gameres/5549/images/source/asusrog1070tibench2.jpg

As you can see, with no overclocking we saw 10-25% improvement across the board, which is impressive. Especially when the 1070 Ti has been designed to run much higher than the out of the box clock speeds.

As you can see, this is a s*** way to display comparison graphis.

*scrollscrollscroll*

For the same price you can get a standard 1080 lol...

NOT WORTH IT AT ALL

http://staticice.com.au/cgi-bin/search.cgi?q=geforce+gtx+1080+8G&spos=20

https://i.imgur.com/Gh10XXv.png

f***, they did prey very high/high for GTXi but for GTX they only did very high.

https://i.imgur.com/s5TdReu.png

The average increase in FPS is 12.31!

Next time you guys decide to graph anything, shoot me a PM and I can help with the formatting. Soz but those graphs are terrible.

lol yeh man if you have that in excel format then send it to me

lol yeh man if you have that in excel format then send it to me

I used R and ggplot2 to graph that. Overkill I know but it lets you modify every single aspect. Here is the code:

library(ggplot2)
library(reshape2)
game_gtx <- c("ACO(U)", "ACO(VH)", "DOW3(MS)", "Doom(U)", "Prey(V)", "Prey(VH)", "GRW(VH)", "GRW(U)", "TD(H)", "TD(U)")
GTXi <- c(70.0, 75.0, 90.74, 133.7, 0, 129.8, 83.71, 54.95, 119.3, 94.2)
GTX <- c(65.0, 70.0, 66.5, 116.2, 145.0, 114.8, 74.08, 48.73, 103.5, 81.8)
to_plot <- data.frame(x=game_gtx, GTX, GTXi)
melted<-melt(to_plot, id="x")
ggplot(melted,aes(x=x,y=value,fill=variable)) + geom_bar(stat="identity", position = "dodge") + theme(axis.text.x=element_text(angle=90, hjust=+0.8)) + xlab("Games") + ylab("FPS")

Like I said, overkill.

Sorry, but your review comments about GPU temperature are not believable. Serious reviews elsewhere put the 1070 and 1080 Tis at around 80 degrees under BF1 max load (your 60 is a dream). Sorry.
I live in a hot dusty country - Australia.
I have to buy 3 new boxes this Christmas. Budget $15,000 with new network. And maybe booster air-con. Choices of large highly filtered towers are hard enough.
I have TRUSTed you guys forever.
PLEASE, include some BF1 max load temps next time.
I really need to make the right choices. THANKS!

Okay well I didn't test BF1. An hour of Wolf 2 on max settings and it was 60-65 degrees. Also this is specific to the ASUS ROG STRIX in default OC mode. This version of the 1070 Ti has exceptional cooling - hence it's 2.5 slot size.

last edited by KostaAndreadis at 01:58:19 06/Nov/17

Oh and Yeti, another review that found the same temps: https://www.techpowerup.com/reviews/ASUS/GTX_1070_Ti_STRIX/35.html

Depends on the fan curve too, default fan curves on pascal cards is woefully inadequate. My founders edition 1080ti sits between 60 - 65 under heavy load because I have made a custom fan curve that ramps the fan up a lot faster, so it doesn't at all suprise me that an AIB card with a custom cooling solution can sit around 60 - 65 degrees under load. if its hitting 80 degrees you'll be starting to thermal throttle the card so you really, really, REALLY don't want that.

@Yeti - the strix cards run much cooler than founders edition.

My ASUS Strix 1080 sits very happily at 69 degrees full load while overclocked.