FINAL FANTASY XIV: Dawntrail Official Benchmark

[DR4G00N]

Media PC, 1920x1080 High.

 

[Dr. Dro]

i7-3770 @ 4.22-4.43GHz and RX 470 4GB

Maximum preset 1920x1080
View attachment 343796

High preset 1920x1080
View attachment 343797

Old girl still got it! Nice.

 

[adilazimdegilx]

neat 10-20-30k points

 

[sLowEnd]

12700K + 6650XT

 

[Pasi123]

Xeon E5-2690 and GTX 960 2GB

Maximum preset 1920x1080: 3547

High preset 1920x1080: 5429

Standard preset 1920x1080: 8765

 

[Ikkyusan]

i5-13500 + RX 6600 (Stock clock) @ 1920 x 1080 Maximum

i5-13500 + RX 6600 (Stock clock) @ 1920 x 1080 High (Desktop)

 

[Keullo-e]

Ran on my 2nd PC, 6700K @ 4.5GHz & R9 290X Matrix Platinum (no manual OC)

 

[RootinTootinPootin]

Thanks guys!!

1080p run

This kinda looked choppy on my 48' OLED screen..lol..

Gonna try and run this on my ROG Ally.

 

[DR4G00N]

Gaming PC, 1920x1080 High. GPU hovers around 1900MHz @ ~100W?

(Might grab the WX 9100 and try it in this system and see if it does any better than before). Edit: Got about 12500pts with it in this PC.

5120x2160 DSR. Better than expected, ~ 35-50FPS.

 

[Marcus L]

Strange GPU usage in some of the scenes with usage hovering around 50% and lower boost clocks bringing FPS down to 80-90 maybe CPU bound as at 1080p? though here is mine on RX 6800 FSR off/1080p

Score *19284*

 

[Dr. Dro]

Strange GPU usage in some of the scenes with usage hovering around 50% and lower boost clocks bringing FPS down to 80-90 maybe CPU bound as at 1080p? though here is mine on RX 6800 FSR off/1080p

Score *19284*View attachment 343899

FFXIV is known for being rather single threaded, having an Intel chip will give an outright edge. 5700X is your bottleneck here guaranteed

 

[Vayra86]

Yeah... I'm not even gonna bother with this one, I'll be posting the absolute bottom 7900XT score on an 8700K...

Strange GPU usage in some of the scenes with usage hovering around 50% and lower boost clocks bringing FPS down to 80-90 maybe CPU bound as at 1080p? though here is mine on RX 6800 FSR off/1080p

Score *19284*View attachment 343899

Recognizable behaviour from the earlier bench they had. Scores can also vary wildly between runs. 1k point gaps happen.

 

[Ware]

 

[QuietBob]

Maximum quality @ native 1080p, CPU and GPU oc'd
Avg fps 113, min fps 60
Loaded in 9.7 s

 

[Knight47]

I dont think this benchmark is accurate at all. The minimum fps is the drop that happens when it loads the new scene and it looks like 3d cache doesnt help much when your gpu bottlenecked.

 

[Det0x]

7800X3D + 3090 at my daily driver settings

Was pretty sure the 3090 was the bottleneck, so i lowered the resolution scale to 50% but keept the same maximum preset

Then i sat option to all low at same 1080p + 50% scale

Decent benchmark, but seem to be pretty limited by GFX even at 1080p (?)

 

[Dr. Dro]

I dont think this benchmark is accurate at all. The minimum fps is the drop that happens when it loads the new scene and it looks like 3d cache doesnt help much when your gpu bottlenecked.

AFAIK this is one of the games where Raptor Lake has a wide performance gap over Zen 4 3D, i.e. the cache is of little help at all. It's also sensitive to GPU VRAM bandwidth, more almost always translating to faster.

The benchmark is "inaccurate" because if frame rate drops below 60 it will activate FSR or DLSS, and it seems that to obtain the 15000 points for Extremely High you will need your minimums to exceed 66 (60 fps +10%), with averages in the 110s, my RTX 4080 barely manages it.

 

[SuperMumrik]

Decent benchmark, but seem to be pretty limited by GFX even at 1080p (?)

Correct assumption, I was pretty gpu bound. I will test a lower res scale tomorrow

 

[Super Firm Tofu]

Laptop (7735HS / RX7700S)

1080p Maximum:

Score: 14879
Average Frame Rate: 107.1491
Minimum Frame Rate: 59

----------------
1080p Standard (Laptop):

Score: 19965
Average Frame Rate: 148.3951
Minimum Frame Rate: 66

 

[Pasi123]

Xeon X5670 @ 4.2GHz and GTX 760. The scores are only around 5-10% lower than on the GTX 960 which isn't bad considering how badly Kepler has aged in some other games.

Maximum preset 1920x1080: 3219

High preset 1920x1080: 4880

Standard preset 1920x1080: 8390

 

[Paramount]

6900XT

 

[RootinTootinPootin]

tweaking things a bit more.

 

[Knight47]

AFAIK this is one of the games where Raptor Lake has a wide performance gap over Zen 4 3D, i.e. the cache is of little help at all. It's also sensitive to GPU VRAM bandwidth, more almost always translating to faster.

The benchmark is "inaccurate" because if frame rate drops below 60 it will activate FSR or DLSS, and it seems that to obtain the 15000 points for Extremely High you will need your minimums to exceed 66 (60 fps +10%), with averages in the 110s, my RTX 4080 barely manages it.

In Limsa I get 160-170 fps(31% cpu usage) and Alliance raid 300 fps(27% cpu). I imported my game settings(1080p) for Benchmark, it gives 200-110-85-75-110 with 17% cpu usage(mostly comes from a single core)

 

[Itz_Vexx]

In Limsa I get 160-170 fps(31% cpu usage) and Alliance raid 300 fps(27% cpu). I imported my game settings(1080p) for Benchmark, it gives 200-110-85-75-110 with 17% cpu usage(mostly comes from a single core)

Yeah the original quote you mentioned is incorrect because what I've noticed in the REAL game is better performance with X3D chips. The benchmark tells one story but whenever I compare to a friend in game who has a 13900K/4080 setup at 1080p (hes upgrading his monitor in a month or two), Im always a decent amount ahead in fps during peak time Limsa. In the same exact spot, directly next to each other, and facing the same thing. The benchmarks for XIV have always shown Intel's 12th, 13th, and 14th gen ahead of X3D but the X3D helps MASSIVELY even in the benchmark.

In game, its even better due to cache being weighed so heavily with tons of characters around. I get around 110-120fps in a heavily crowded Limsa next to the Aetheryte. Like 1-3% of that could be GPU bottleneck since my Reshade preset is very heavy and my GPU usage JUST hits 97-99% at 120fps ish. Only place the EW benchark saw any kind of CPU work was in the Old Sharlayan scene which even then had a mild amount of NPC's. Even in the EW benchmark I went from 20.5K on my 3950x, to 25.5k on my 5900x, to 31k on my 5800X3D all with the same exact setup and settings using my 3080 TI. All at native res for me which is 1440p.

 

[SuperMumrik]

The benchmark tells one story but whenever I compare to a friend in game who has a 13900K/4080 setup at 1080p (hes upgrading his monitor in a month or two), Im always a decent amount ahead in fps during peak time Limsa. In the same exact spot, directly next to each other, and facing the same thing.

Yeah, this makes NO SENSE at all vs a 5800x3d

Then i sat option to all low at same 1080p + 50% scale

The 7800x3d is impressive. What memory settings were you running here?
My 14900kf is pretty much tapped out on this board (Apex running 8800 is gone).