Victoria 3 Performance Benchmark

[Notme1]

Game pauses a tiny bit every time it renders a frame as to ensure stability and correct display of data from I heard.
So low FPS cap results in game running faster at speed 5.
This effect seems to be small, was bigger in 1.6.

 

[Ruck]

Game pauses a tiny bit every time it renders a frame as to ensure stability and correct display of data from I heard.
So low FPS cap results in game running faster at speed 5.
This effect seems to be small, was bigger in 1.6.

But that doesnt explain the 30 fps cap 1840 test.

 

[Notme1]

But that doesnt explain the 30 fps cap 1840 test.

Early game so its all is within margin of error

 

[Louella]

Your save probably has much more going on than the test saves here which were created by just letting the game run till 1900.

But maybe your system runs out of memory and slows down, because it has to read and write to the pagefile constantly.

I recently upgraded the memory on my system, to 32GB. It did not make a difference, only deleting the trade routes did.

Sounds like trade or military units might be slowing down game excessively, make bug report.

Was the trade routes. It's a thing that's already been reported though, so I didn't report it. Should be simple to try out though.

 

[esha26]

GPU: GTX 1660
CPU: 14th Gen Intel(R) Core(TM) i5-14600k
RAM: 32GB (2x16) 5600 MHz DDR5
SSD: 970 Evo Plus 2TB
Game Version: Rolled it back to 1.7.6

1840 Time: 1:04.23
1840 Slow Tick operations: I'm sorry, I forgot!

1900 Time: 2:07.03
1840 Slow Tick Operations:

With the GPU I have, the 'real' performance may be a little worse than what these times suggest. My FPS was also capped at 144, not sure if that impacts performance at all. (edit: it seems to have been off anyway)

 

[Ajaxwindowcleaner]

Apparently AMD's new 9800x3D does extremely well with Clausewitz games.

 

[bball617]

Guess I'll be the first 9800x3D post and will add more tests when I get a chance. This is without PBO enabled and no curve optimizer. Curious how much extra that will gain whenever I get a chance

GPU: 3080TI
CPU: AMD 9800x3D
RAM: 32GB (2x16) 5200 MHz DDR5
SSD: 970 Evo Plus 2TB
Game Version: 1.7.6
Windows 10

1840 Time: 1:03
1840 Slow Tick operations:

1900 Time: 2:14
1900 Slow Tick Operations:

EDIT: A couple more benchmarks with PBO and EXPO enabled. CPU was a +200 and RAM now running at 6000 MHz

1840 Time: 59sec
1840 Slow Tick operations:

1900 Time: 1:59
1900Slow Tick operations:

Also did a run with 60fps counter set and man was it flying.

1840 Time: 45sec
1900 Time: 1:44

Everyone should have the FPS counter set to 30 or 60fps. +10% improvement with just FPS Counter

 

[Caracus]

I would have expected a higher difference between the i5 14600k and the 9800X3D without PBO. Especially given their price. V3 using multiple threads and also needing high boost clocks makes me wanna wait for the 9950X3D. If you got more latest gen X3D results please keep em coming

 

[bball617]

I would have expected a higher difference between the i5 14600k and the 9800X3D without PBO. Especially given their price. V3 using multiple threads and also needing high boost clocks makes me wanna wait for the 9950X3D. If you got more latest gen X3D results please keep em coming

I was too until I saw the 14600k above my post was ran with an FPS cap. That's why I added and extra test with the fps cap on my post.

Comparing apples to apples, with an FPS cap on, the 9800x3D was almost 25% faster in the 1900 run and a whopping 30% faster in the 1840 run vs a 14600k

 

[esha26]

I was too until I saw the 14600k above my post was ran with an FPS cap. That's why I added and extra test with the fps cap on my post.

Comparing apples to apples, with an FPS cap on, the 9800x3D was almost 25% faster in the 1900 run and a whopping 30% faster in the 1840 run vs a 14600k

Actually, I'm the 14600k user. Turns out I didn't have an FPS cap turned on, as when I booted Vicky 3 again, there was no FPS cap. This checks out because I ran through the benchmark without the cap again and the resulting time was a second shorter. It's early on in the 9800x3D's life though, and maybe the L3 cache has a bigger impact in more active gameplay.

 

[bball617]

Actually, I'm the 14600k user. Turns out I didn't have an FPS cap turned on, as when I booted Vicky 3 again, there was no FPS cap. This checks out because I ran through the benchmark without the cap again and the resulting time was a second shorter. It's early on in the 9800x3D's life though, and maybe the L3 cache has a bigger impact in more active gameplay.

Oh nice, 14600k is really kicking butt!

I am really curious if you get the same type of improvements with an fps limiter. If you get a chance, I'd love to see the results with the FPS locked to 60fps to compare the times.

 

[esha26]

Oh nice, 14600k is really kicking butt!

I am really curious if you get the same type of improvements with an fps limiter. If you get a chance, I'd love to see the results with the FPS locked to 60fps to compare the times.

Sure! Here's the additional results:

GPU: GTX 1660
CPU: 14th Gen Intel(R) Core(TM) i5-14600k
RAM: 32GB (2x16) 5600 MHz DDR5
SSD: 970 Evo Plus 2TB
Game Version: Rolled it back to 1.7.6

1840 Time (60 FPS cap): 00:47
1840 Slow Tick Operations (60 FPS cap):

1900 Time (60 FPS cap): 01:43
1900 Slow Tick Operations (60 FPS cap):

Really good times in general, and capping the FPS seems to make a sizeable impact indeed. Now here's to hoping that I don't face any of the degradation issues people mention in regards to these chips...

 

[Whitewizard2750]

I’m thinking about buying this i5 - 14600k CPU but I’m on budget. Is there some good value CPU that would be a little worse but reasonably cheaper? Or is 14600k a great deal on itself? Thank you for advices.

 

[bball617]

I’m thinking about buying this i5 - 14600k CPU but I’m on budget. Is there some good value CPU that would be a little worse but reasonably cheaper? Or is 14600k a great deal on itself? Thank you for advices.

14600k is a pretty dang good deal at $200. But.... I'd be worried about it cooking itself with all the issues that those 2 generations have had. Intel says theyre fixed but can you really believe that?



Thanks @esha26! Pretty even and for $250 cheaper its a steal as long as it holds up.

 

[masteriw]

How come the 14600K outperforms the 9800X3D?

Maybe when the 16-core 9950X3D is out someone will test it...

Also how do you guys set up framerate caps? Through the ingame options menu or via video driver? (NVIDIA Control Panel, for example).

Also, thank you for all the tests!

 

[tottetotte]

How come the 14600K outperforms the 9800X3D?

Maybe when the 16-core 9950X3D is out someone will test it...

Also how do you guys set up framerate caps? Through the ingame options menu or via video driver? (NVIDIA Control Panel, for example).

Also, thank you for all the tests!

The reason I am here is because I am not 100% sold on AMD's X3D hype. I don't understand why it's necessary to put graphic calculations in the CPU, which I thought was the job of the graphics card. But I have not done heavy research on it since I don't play games that rely heavily on graphics; rather, I play Paradox games or other grand strategy games that rely heavily on data calculation. I believe a $250 CPU can outperform a $500 CPU at Paradox games due to the performance in single-threaded tasks. Intel has been roasted by the gaming community for being a 'data-center' CPU, while AMD has been praised for being the gamer's CPU. However, I want to buy a CPU for its performance in calculating data, not for increasing FPS from 800 to 850.

Why is this the case though? Game studios will likely never prioritize full optimization with multi-threading, since they would probably never handle all the bugs that come with it. Sure, they can multi-thread simple, non-conflicting tasks like audio, rendering, and other background processes to off-load the heavy calculations, but the core game logic and all the massive checks will always be the bottleneck on a single thread. Hence why more cores will not help at all for games like Victoria 3

It would be great to see how Ryzen 5 9600X performs over the Core i5-14600K

 

[ThePHD]

How come the 14600K outperforms the 9800X3D?

Maybe when the 16-core 9950X3D is out someone will test it...

Also how do you guys set up framerate caps? Through the ingame options menu or via video driver? (NVIDIA Control Panel, for example).

Also, thank you for all the tests!

I'd like to see more testing before I draw any conclusions. If anyone else gets one then please do share your results. Linus Tech Tips and Gamers Nexus both tested it in Stellaris and the 9800X3D destroyed every other CPU by a large margin. It's odd that there isn't much of a difference in Victoria.

 

[Haresus]

1.8.1 saves uploaded, I'll run the test tomorrow or so. Should be some improvements for late-game performance according to the patch notes.

1840:

Dropbox

www.dropbox.com

1900:

Dropbox

www.dropbox.com

 

[matlion91]

GPU: AMD Radeon RX 6700 XT
CPU: 13th Gen Intel(R) Core(TM) i5-13600k
RAM: 32GB
SSD: Yes, Samsung SSD 970 EVO Plus
1840 time: 1min 28 sec
1840 Slow Tick Operations:

1900 time: 2min 33sec
1900 Slow Tick Operations:

 

[FrontLine1111]

I'd like to see more testing before I draw any conclusions. If anyone else gets one then please do share your results. Linus Tech Tips and Gamers Nexus both tested it in Stellaris and the 9800X3D destroyed every other CPU by a large margin. It's odd that there isn't much of a difference in Victoria.

I don't see how graphics performance should change from early game to late game. What am I missing? My guess would be the exponential growth of buildings, construction, pops, etc., which indicate a game design which does not scale. I haven't spent a lot of time researching it though.