Three Kingdoms Benchmark Thread

[La♔De♔Da♔Brigadier Graham]

Benchmarks and all that.

Post your Benchmarks here with you PC specs it may give other players some idea of what to expect!
brig

Nvidia Battlebox Utah STRIX


CPU
Intel® Core™ i7 7700K CPU (Quad Core, 4.2GHz - 4.5GHz, 8MB Cache)

Liquid CPU Cooler
Corsair Hydro Series™ H100i v2 Extreme Performance Liquid CPU Cooler

RAM
1Corsair Vengeance® LPX 32GB (2x16GB) DDR4 DRAM 3000MHz C15 Memory Kit - Black


Graphics Card
Aorus GeForce GTX 1080 Ti Extreme11GB


Motherboard
Asus ROG STRIX Z270F GAMING (Socket-1151, 4x DDR4, ATX)


Operating System
1 Microsoft Windows 10 Home 64-bit EN (OEM)


Storage HDD / SSHD
Seagate BarraCuda 2TB 3.5" SATA III HDD


SSD
Samsung 500GB 960 EVO M.2 NVMe PCIe SSD

Case
Corsair Carbide Air 540 Black (ATX)


Power Supply Unit
Corsair RMx Series™ Modular RM850x — 850 Watt 80 PLUS® Gold

--------------------------------------------------------------------------------------------
nvidia version 430.64
All settings on ultra, AA turned off, Large Battles,1080p....


Battle Benchmark.....

Spoiler Alert, click show to read: 

Campaign Benchmark...

Spoiler Alert, click show to read: 

[pyj99]

Alienware 17
CPU 4770 HQ
GTX1080
32 GO ram
SSD 500 GO

2560x1440
Everything on Ultra including Unit Size on Extreme

62fps on battle benchmark

[SirAndrean]

WIndow 10 64-bit
CPU i5-4670 @3,40GHz
MSI GTX 970 4GB (Over Clocked with MSI afterburner)
16GB RAM
Standard HDD

I played in 1080p
I'm able to use extreme unit size, with some ultra and mostly high. Its start only to lag if i play 3 unit stacks battle or 2 unit stacks in large city with fire and smokes.
35fps on battle benchmark

[perablenta]

On my PC :
CPU: Intel i7-4770K @3.9GHz
MB: MAXIMUS VII HERO
Memory: Kingston DDR3 16GB 2133MHz HyperX Beast
GPU: Gygabyte GTX 1080 G1
SSD: Kingston SATA III SA400S37/480G A400
PSU: Chieftec 1000W APS-1000C




At 1440p and all ultra settings with TAA, I am getting average of 43.7 with Extreme units, 45.1 ultra units size, 46.0 on large unit size.




Latest nVidia drivers v430.64.

[♔Greek Strategos♔]

Be aware that benchmarks are usually NOT very trustworthy, as ingame performance could be rather different depending on many many aspects.

I'll soon make a dedicated Tech thread, like the ones I have for Rome, Attila, Tob etc with tech tips, suggestions for hardware,advanced settings etc

Cheers!

[LestaT]

Ryzen 5 2600
Sapphore RX 580
16 GB RAM

Done battle benchmarks on :
Recommended (High with Normal unit size).
Recommended (High with Extreme unit size).
Ultra with Extreme unit size.

[Pathstrider]

Be aware that benchmarks are usually NOT very trustworthy, as ingame performance could be rather different depending on many many aspects.

True, though imo they have gotten better recently - the Skaven benchmark in WH2 was much better than the original battle one, and the Three Kingdoms looks similar to that.

I've been fortunate in that I've got a few systems in the house at the moment to test on which should provide an interesting comparison.

All of these are on the battle benchmark and show Average/Min FPS

i5 6500/1600mhz DDR3/GTX970/1080p/High Settings

Unit Size Large: 44/39 Ultra: 44/39 Extreme: 42/33

I also tried with FXAA instead of TAA: Unit Size Large: 54/47 Ultra: 50/39 Extreme: 46/33

This is almost bang on CA's recommended settings for High - I seem to be under the expected 45-55 FPS but I think that could be due to slow memory or the VRAM usage going over 3.5GB (and thus into the slower speed 0.5GB the 970 has)

i7 7700k/ 2133mhz DDR4/GTX 1660TI/1080p/ Ultra Settings

TAA: Large: 52/46 Ultra: 47/42 Extreme: 44/39
FXAA: Large: 54/49 Ultra: 52/45 Extreme: 49/38

My experience in the past is that Hyperthreading made no difference to Total War games, but I tested with Hyperthreading off:

TAA: Large: 48/43 Ultra: 46/39 Extreme: 44/35
FXAA: Large: 54/49 Ultra: 51/43 Extreme: 48/37

I think the graphics card was more of a limit in this case, but there are certainly some improvements to the min FPS with Hyperthreading.

I'll probably test again on high at some point to really see what the 4 Core/8 Thread limits are in this title.

i5 8400/ 2666mhz DDR4/GTX 1080/1080p/ Ultra Settings

Large: 71/65 Ultra: 67/58 Extreme: 62/50

The CA recommended specs for 60 FPS ultra is a i7 8700k, but the i5 8400 handled it fine - I suspect that CA made a decision to recommend the i7 to make it easier to compare to older I7's (or they got paid by intel) but from what I saw on CPU usage, Three Kingdoms really likes having 6 cores to work with.

I also ran at 1440p:

Large: 47/42 Ultra: 46/40

i7 9700k @4.9ghz/ 3200mhz DDR4/GTX 1080TI/1440p/ Ultra Settings

Large: 64/57 Ultra: 61/54 Extreme: 59/50

I'm fairly happy with this paired with a G-sync monitor, though the Extreme unit size lows are a bit much. From what I could tell the game used all 8 cores, but favored 6 - which backs up my theory it's be targeted at 6 cores.

[Pathstrider]

As an addendum, I’ve done a little bit more testing on the i7 7700k/GTX 1660 TI build. It looks like moving from TAA => FXAA removes the GPU bottleneck. From there the lows seem to be more to do with CPU - I had to drop shadows to medium to maintain a min of 60 FPS. This sort of suggests to me that a 4core/8thread processor isn’t enough for Ultra unit size + ultra shadows, although it might be solved by overclocking the 7700k and using faster memory.
I also tried dropping down to FXAA on the 9700k/GTX1080ti build - this had quite a substantial impact, gave me on average +15 FPS or so. That a much bigger impact than I was expecting, and is in the realms of what it costs me to go from FXAA => 4xAA in Warhammer 2 (c20 FPS).

[A Barbarian]

System: 6850K@4.4Ghz HT on, 1080Ti FE@1950Mhz, DDR4-3000@CL13-12-12-32-1T

Battle 1080p:
Med: 180.9 (125-245) fps
High:130.2 (96-158) fps
Ultra: 94.4 (81-116) fps

Spoiler Alert, click show to read: 

Ultra + Extreme Shadows + Extreme Unit Size: 76 (56-108) fps

Spoiler Alert, click show to read: 

Battle 1440p:
High: 84.5 (75-98) fps
Ultra: 64.2 (57-75) fps

Spoiler Alert, click show to read: 

Update: Patch from 28th May seems to have essentially identical performance

Tip: If somebody struggles with CPU limited performance, don't forget to optimize DRAM speed (besides OC CPU). TW games seem to be pretty DRAM speed and also latency sensitive. In Attila I could measure up to 40% more frames from just doubling the RAM speed (see my post here: https://www.twcenter.net/TWC-Total-War-Attila-benchmark-thread). And TW3K seems to like fast RAM too.

[Odinarius]

i7 4790K (4.4Ghz)
16gb DDR3 RAM
Nvidia 1070Ti, stock settings

Getting around 55-58fps in benchmark on 1080p on Ultra settings.

Had problems with screen tearing, so I turned on Adaptive sync in Nvidia drivers (Vsync when above 60fps, turned off when below). Worked wonders, no screen tearing at all, buttery smooth motion since I've done that. Big recommendation!

[Mr.Freud]

5930k @ 4.4
32gb ram @ 3200
2080ti @ 2100/7500

max. settings with extreme unit size and extreme shadows @ 1080p

Battle bench. average: 79.5 lowest: 50

campaign average: 93.5 lowest: 64

Now I wonder if they implement Dx12 support and if they do - by how much will it improve CPU bottleneck situations
P.S. Anyone knows what "Battle cinematic smoothing" does?

[Pathstrider]

Tip: If somebody struggles with CPU limited performance, don't forget to optimize DRAM speed (besides OC CPU). TW games seem to be pretty DRAM speed and also latency sensitive. In Attila I could measure up to 40% more frames from just doubling the RAM speed (see my post here: https://www.twcenter.net/TWC-Total-War-Attila-benchmark-thread). And TW3K seems to like fast RAM too.

I found the same thing for Warhammer 2 - going from 1600mhz to 3200mhz with my 7700k added about 10 FPS to my minimums.