just outa curiosity I compared these two cards running firestrike. 970ssc and 1080sc. the 970 scored 9,636. 1080 was 16,684. granted the 970 is in my spare machine. specs for both rigs are below
 
NZXT switch 810
Dell 2560x1440p 144hz gsync
i5 6600k 4.3ghz
Gskill 16gb 3200mhz
gigabyte Z170X G6
eVga 1080SC
corsair 1000wt psu
noctua cpu cooler
corsair force gt 180gb
WD 500gb storage
WD 300gb velociraptor
windows 7 pro 64bt
 

NZXT 810 FULL TOWER
EVGA 970SC GPU
ASUS Z77 SABERTOOTH MOTHERBOARD
I5 3570K 4.4GHZ CPU
G SKILL SNIPER 16GB 1866GHZ RAM
CORSAIR H80 LIQUID CPU COOLER
CORSAIR 1000 WT PSU
CORSAIR FORCE 120GB SSD
LG DVD/BLUERAY BURNER
FAN CONTROLLER
WINDOWS 7 HOME 64BT

AzN-SoLjA
Interesting, throwing out some other numbers here.
Mine 780Ti 11,200+ & 680FTW 6,800+
I would figure the 970 do better than my card in the past

ooops. I ran extreme. just did the basic run.[1.1] score was 9636. dang. still lower than the 780ti

http://gpu.userbenchmark.com/Compare/Nvidia-GTX-970-vs-Nvidia-GTX-1080/2577vs3603
 
There you go.

Common knowledge no doubt but the over all score if testing ONLY GPU isn't representative for comparison unless the CPU has the same threads. I have several PCs and consistently the overall score (i7 cpu) on lesser cards beats the overall score on a machine with an i5 but better card. (Within reason of course)
 
If I am interested in JUST measuring the GPU it's more accurate to use the total graphics score only. Then the tests are close to what would have been expected. Bottom line. Firestrike scores are heavily influenced by how many cores, or virtual cores (i5xxx vs i7xxx for instance) one has. I am not saying it is a bad thing, but that's just how these tests are designed.
 
I know it is not very relevant so far in this thread, and I know just about everybody who loves PCs, building, OC, or just testing knows this already. But thought I'd post anyhow.

stalinx20

Gawg36
Common knowledge no doubt but the over all score if testing ONLY GPU isn't representative for comparison unless the CPU has the same threads. I have several PCs and consistently the overall score (i7 cpu) on lesser cards beats the overall score on a machine with an i5 but better card. (Within reason of course)
 
If I am interested in JUST measuring the GPU it's more accurate to use the total graphics score only. Then the tests are close to what would have been expected. Bottom line. Firestrike scores are heavily influenced by how many cores, or virtual cores (i5xxx vs i7xxx for instance) one has. I am not saying it is a bad thing, but that's just how these tests are designed.
 
I know it is not very relevant so far in this thread, and I know just about everybody who loves PCs, building, OC, or just testing knows this already. But thought I'd post anyhow.

Well yeah... Most people already know that. 

Ah....., yeah......, Ahem. I guess it was a rather pointless post, but you made it better!  Thank you for that wonderful bold enlarged part. Now no one can be in any doubt!!