Battlefield 6: Multiplayer CPU Test, 33 CPU Benchmark

To be fair, 2042 had 128 players and gigantic maps which did drag down performance a lot. I don't think there's an engine that can handle that particularly well.

[-]

Midland3640@reddit

at what resoltion are you playing? just curious

[-]

Firefox72@reddit

Locked 80fps in smaller modes like Rush.

Locked 70fps in Conquest/Escalation

1080p.

[-]

norhor@reddit

Yeah, it is important context. Also what settings is used.

[-]

trololololo2137@reddit

CPU bottleneck is crazy in BF6, denser areas easily drop to 70-80 FPS on 5950X lol

[-]

Risley@reddit

What kind of potato are you playing on. I play with a 13700 and a 4090 and at no point in the game have I seen a drop. And my graphics are in overdrive.

[-]

Raphaeluss@reddit

If someone still plays in 1080p, it might be useful to them

[-]

DataLore19@reddit

1080p is the internal render resolution of your GPU if you're using 4k resolution with performance upscaling (FSR or DLSS).

[-]

BlackPet3r@reddit

Or well you know, everyone using DLSS or FSR while playing in 1440p. Quality preset at that resolution renders at 1080p, which increases CPU load.

[-]

Cireme@reddit

1440p DLSS Quality is even lower, 960p.

4K DLSS Performance is 1080p.

[-]

Turkish_primadona@reddit

Some of the comments here confuse me. I'm running a R5 7600 and a 7700xt

1440p with a mix of high medium, I get a consistent and pleasent 75-85 fps.

[-]

DataLore19@reddit

Reddit always good for a little "works fine on my PC". 😂

[-]

Regarding the VRAM-recap section in the video. Did he account for the VRAM-leak issue/bug with the overkill texture setting? Currently there is a bug where the game continuously eats more VRAM the longer you play if you have the texture setting at overkill (this also affect the multiplayer, where your FPS will decrease map after map).

For example, my 5070 ti will play at 150+ fps during the first map, but if I play long sessions it drops down significantly - down to the 90s. Changing the overkill texture setting down, then back up again fixes the issue ( or by restarting the game).

[-]

RandyMuscle@reddit

So that’s what happened. When I had textures at overkill, my VRAM just got destroyed during my second game when it seemed fine at first. FPS took a crap and game got super choppy. I’ve been totally fine since turning textures back to ultra.

[-]

_OccamsChainsaw@reddit

I haven't noticed this with my 5090. Granted I might not have played for a long enough session to reveal the problem but I'd assume several hours should do it.

Conversely CoD (specifically warzone) would pretty routinely crash for me for the same reason.

[-]

Exajoules@reddit

I haven't noticed this with my 5090. Granted I might not have played for a long enough session to reveal the problem but I'd assume several hours should do it.

I guess it took around 5-6 games in a row before I started to notice performance drops. Since the 5090 has much more vram, it likely takes much longer for it to become a problem (if ever).

[-]

RandyMuscle@reddit

5800X3D and 5070 Ti here. Playing on high with textures on ultra and filtering on overkill at 4K with DLSS set to balanced and my FPS almost never goes below 110. They just need to fix the mouse stuttering. No matter what FPS I’m getting, the mouse movement looks awful. I play with controller mostly and it doesn’t impact controller for whatever reason. Hope it’s fixed soon for the mouse people.

[-]

XavandSo@reddit

The inevitable 5800X3D marches on forwards.

[-]

Geddagod@reddit

Interesting to see the 12900k fare a bit better though. On launch IIRC, on average, the 5800x3d was pretty much on par. Do newer games like the 12900k more than the 5800x3d?

[-]

Gippy_@reddit

At launch the 12900K was tested with trash DDR5 because it was new: GN with 5200 CL38, HUB with 6000 CL36. At the time, their conclusion was that going to DDR5 wasn't worth it because it was barely faster than DDR4.

When the 5800X3D launched, HUB tested it against a 12900KS running 6400 CL32 and they traded blows against each other.

However, in this video, the 12900K was tested with 7200 CL34 which really extracts the last bit of performance out of it, while the 5800X3D is still stuck with 3600 CL14 DDR4.

[-]

Doubleyoupee@reddit

I wonder why it's so much slower than the other X3D parts though?

The 7600X3D is much faster than high clocked CPUs like the 14900K so the X3D cache is definitely strong in this game. Yet the 5800X3D is being beaten by a 7600F. I expected the 5800X3D to be higher

[-]

XavandSo@reddit

I guess there's only so much 3D V-Cache can push an aging Zen 3 architecture.

[-]

Firefox72@reddit

One of my biggest regrets was not getting the 5700X3D when it was still in stock to upgrade from my 5600.

[-]

Copponex@reddit

Exact same situation. I didn’t really pay attention toothed cpu market after I bought my pc, and now when I’m looking to upgrade I can see that I have completely missed a golden opportunity.

[-]

xspacemansplifff@reddit

Hell yeah

[-]

EnglishBrekkie_1604@reddit

As is destined.

[-]

Klaritee@reddit

200s boost is cover by warranty so you have no reason not to use it. 2100 d2d frequency is criminal.

[-]

IlTossico@reddit

I can't understand why this man can't do a functional benchmark, like trying different resolution and maybe trying older CPU that people still running, to see if someone need or not an upgrade.

Same for the GPU benchmark, totally useless.

Anyway, i'm pretty sure the finished game run differently than the Beta, the last Beta i tried was way force in performance than the previous one and then alpha tests too.

But if someone have an i9 9900k, and it's curious to know performance, no issue, both on 1080p and 1440p the CPU is chilling, never got above 40% usage.

[-]

Cireme@reddit

But if someone have an i9 9900k, and it's curious to know performance, no issue, both on 1080p and 1440p the CPU is chilling, never got above 40% usage.

That doesn't mean you're not CPU limited. It means that the game uses 6.4 of your 16 threads, but you could still be limited by your single thread performance.

Generally GPU demanding, my 2080 was struggling a lot in 1440p all low, to maintain 60 fps. DLSS was making 0 difference.

Yeah you are defintely CPU limited.

[-]

TopSchnitzel@reddit

Cpu benchmarking is always done at 1080 to prevent GPU bottlenecking so you obviously don't know what you're talking about lmao

[-]

SirMaster@reddit

My 5900x is bottlenecking my 3080 :(

[-]

Jayram2000@reddit

Same here with my 9070xt, it's getting slammed to 100%.

[-]

DynamicStatic@reddit

I can play 1080p on low with my 7950x3d and 3080 Ti and still only hit 120-140 fps. Hmmm.

[-]

AK-Brian@reddit

If it's like the beta, ensure that you've enabled it as a game within the Game Bar profile (if using CPPC Driver preference), as it wasn't picked up automatically and likely still won't be this soon after launch. Disabling SMT also improved performance for me during that test.

[-]

DynamicStatic@reddit

I assume it would be the same if I lock it to certain cores with processor lasso? Either way my performance is pretty bad.

[-]

AK-Brian@reddit

Assuming it doesn't trip up anticheat, it should see a similar result, yeah. Forcing processes to the cache chiplet via CPPC Prefer Cache would also work as a quick and dirty solution.

I don't have the retail version and can't give you any numbers, unfortunately, but even jumping into the firing range and jotting down some quick numbers should give you a ballpark idea of whether or not you're seeing an uptick.

[-]

godfrey1@reddit

can't listen to the video now, any reason for 7800X3D absence?

[-]

TheFinalMetroid@reddit

HUB lent it to another reviewer

[-]

VincePuc9@reddit

So happy i pulled the trigger on a 9950x3d :D

[-]

TheBestestGinger@reddit

I wonder if they optimized the game in between the beta and release.

I was playing the on 1080p with a R7 3800x and a 3080 and was really struggling on the lowest settings (if I remember correctly averaging roughly 60 fps, but as a match went on I averaged maybe 45-50 fps)

I upgraded to a 5700x3D and the game is running smoothly on a solid 120 fps on medium - high graphics.

Looking at the benchmarks in the video it looks like the R5 3600 is getting some decent frames of about 92 on average at low.

[-]

trololololo2137@reddit

feels the same to me between bf labs/beta/release