Fallout 4 CPU benchmark thread(need some Zen3 and Zen4 results!)

CAT-THE-FIFTH · 23 Apr 2022 at 12:51

Robert896r1 said:
I got a x3d coming so will test. Maybe I can beat my own scores and lock out the podium lol

It will be great to see your results!

CAT-THE-FIFTH · 23 Apr 2022 at 13:54

@KompuKare Fallout 4 loves memory bandwidth - I have not tried tweaking the RAM any further with the Ryzen 7 5700X,so its quite possible I could get better timings.

CAT-THE-FIFTH · 23 Apr 2022 at 15:06

KompuKare said:
Wonder if the two scenes can show this somehow, say IF one was more bandwidth demanding, and IF the other was more latency.
Guess not, but I added a S1/S2 ration column:

(conditional formatting ignores Tired9's results as they the drawcalls in their results were nowhere near 11,700 and 8,000).

Unsure whether that tells us anything though?

Corvega is very drawcall heavy because you are high up in a large built-up area so is more a rendering thread test IMHO. Diamond City has a much higher NPC density.

Having said that test 1 shows the least improvement from Zen2 to Zen3.

CAT-THE-FIFTH · 25 Apr 2022 at 11:10

zero989 said:
Was bored so I tried it on my laptop, 11900H (24MB L3 Cache).

~4800Mhz CPU 1.23v 4.4ghz cache
16GBx2 3200MHZ C16-17-17-33-1T 1.2v Gear 2
RTX 3080 16GB Laptop GPU

S1 11711 73 FPS
S2 7945 88.1 FPS

AVG 80.5 FPS

Windows 10 21H2. Can probably get slightly higher with a fresh install of win10 as that maxes out memory performance.

AMD users should be using BGS (Bank Group Swap) Alt Mode and should also disable GDM (Gear Down Mode) in addition to optimizing subtimings.

kindle12015 said:
12600KF (5.125GHz, 100.5BLCK x 51, x40 ECores, x43 Ring, 2x8GB DDR4 3920MHz 15-15-15-32 Gear 1 CR1 1.55V Samsung B-Die), RTX 3080, Win11
Save1 (Corvega): 11693 - 96.8 FPS
Save2 (City): 8029 - 123.4 FPS
AVG: 110.1 FPS

https://imgur.com/aX0c027
https://imgur.com/whkNc9o

I wonder what OCed 12700K and 12900K results will be like. They should be a lot faster cause of bigger cache.

Thank you very much for the results!

@KompuKare BTW,if you look at the Ryzen 7 5800X3D results with 3200C16 results the AT numbers look higher - I wonder whether that is because the 64GB kit has weaker subtimings or whether the GTX1080TI seems to have less of a driver overhead??

The Corvega results are much lower with the RTX3080 but have higher drawcalls - the Diamond City results have less drawcalls but probably a higher immediately NPC AI load.

Yaayuh! said:
The Creation engine truly is utter ****. I really hope Creation 2 turns out some big improvements.

It's an ancient engine which has had bandaids put over it's Gamebryo engine roots.

CAT-THE-FIFTH · 30 Apr 2022 at 14:43

Jonj1 said:
Just to copy my post over from Hexus :

Ryzen 7 5700G - no graphics card - 32GB DDR4 3200Mhz

Corvega=43.7 FPS (Drawcalls=11667)
Diamond City= 57.3FPS (Drawcalls=7975)

4GB ram allocated to the CPU for graphics.

Surprisingly close to the Ryzen 5 4650G with a GTX1080TI and faster than a Core i5 3570K with an RX470 4GB! It shows you how CPU limited Fallout 4 is!

CAT-THE-FIFTH · 30 Apr 2022 at 16:19

Jonj1 said:
I was surprised by the result

It's really CPU limited!!

CAT-THE-FIFTH · 27 May 2022 at 16:44

KompuKare said:
Well, that's definitely the top score so far:

Still, I'd rather suspect your 12600K rig was a bit cheaper!

I suspect the reason why the Ryzen 7 5800X3D is underperforming with an RX5700XT is down to the game prefering Nvidia dGPUs and increased AMD driver overhead. Also look at the Ryzen 7 5800X and RX6800XT results compared to the Ryzen 7 5800X3D results and the RX5700XT numbers. There is about a 21% to 22% increase in performance using a Ryzen 7 5800X3D with a much slower dGPU. I will add some results with an RX5600XT with my Ryzen 7 5700X and see how it performs. Then hopefully I can do the same with a Core i5 10400.

CAT-THE-FIFTH · 27 May 2022 at 18:43

Ok,benchmarked game with an RX5600XT. I used the latest supported driver.

I used an average of three results for each scene.

Corvega=61.33FPS(11709 Drawcalls)
Diamond City=75.67FPS(8022.67 Drawcalls)

At some point might try the Core i5 10400 with the GTX960 4GB and RX5600XT to get some more data for the table.

CAT-THE-FIFTH · 1 Jun 2022 at 10:22

If you can run the RX6900XT also,even if the score is lower it doesn't matter because if it improves over the RX5700XT score,you could have the highest AMD score. Also thanks for going to the effort - must be a PITA to take out all these cards!

CAT-THE-FIFTH · 1 Jun 2022 at 10:25

Robert896r1 said:
No offense @LtMatt but your scores don't line up with any of the x3d results. Nothing you're doing tuning wise is wildly beyond the capability of the chip limits.

He is running 4400CL14 DDR4 which is much faster than the other Ryzen 7 5800X3D results?

CAT-THE-FIFTH · 1 Jun 2022 at 10:34

Robert896r1 said:
He's not. That's the kit spec. He's then tuned it down keep in sync. Pay attention to his timings screenshot.

But interestingly enough the top score for the Ryzen 7 5800X3D was at stock,but with the best memory settings. So I would expect with betting memory settings and a slight overclock it should beat the JZ- has. The 1st test is only a few FPS higher(115FPS vs 120FPS),its the second test which is where the biggest improvement seems to be in.

CAT-THE-FIFTH · 1 Jun 2022 at 10:36

LtMatt said:
I'm now about to reinstall Windows, but once I've done that I'll maybe add some 6900 XT scores.

The settings are definitely right this time, so not sure what else I can do and my screenshots look identical to all the others bar slightly different camera positions, but that's just trying to get close to the target draw calls as specified in the OP. FPS can change depending on where you look, I question this benchmark methodology.

It's running at almost 4000Mhz CL14 with infinity fabric at close to 2000Mhz.

Feel free to leave my scores off the leader board guys, don't want to create an argument over a benchmark score. Not that bothered as it's a pain in the ass to run anyway so I can focus on other game benches.

I knows it a PITA,so me and KompuKare appreciate the efforts of anyone who can run this benchmark! It simply does not have much benchmarking done on newer CPUs and dGPUs.

CAT-THE-FIFTH · 2 Jun 2022 at 23:12

What it looked like on 2015 era CPUs:

https://i.imgur.com/vWuOBwK.png

Core i7 6700K(2016 article) memory scaling:

https://i.imgur.com/GNG5hq8.png

Anyway,I am going to test this on a Core i5 10400 with RAM running at 2666MHZ.

CAT-THE-FIFTH · 5 Jun 2022 at 18:09

Got some new results on my second PC with an RX5700XT borrowed off a mate.Will try and test my RX5600XT,my RTX3060TI and the old GTX960 4GB I have at some point.

Specifications
Core i5 10400 set at 80W power(short and long duration power).
Alpenfohn Black Ridge low profile CPU cooler
Asus PRIME B460I-Plus mini-ITX motherboard
16GB 3200MHZ RAM(run at 2666CL15-17-17-36 2T)
500GB WD SN700 NVME SSD

The CPU seems to run at 4006MHZ consistently.

https://imgur.com/a/MywXnXc

Diamond City=64.4FPS(Drawcalls=8016)
Corvega=54.93FPS(Drawcalls=11730)

CAT-THE-FIFTH · 6 Jun 2022 at 20:46

I have run this on two different systems with three different CPUs, and 4 graphics cards so far, so not sure what you are doing wrong(I have a whole tranche of results to add). The ini files have to be changed per the instructions copied over from AT forums and Vsync has to be disabled in Nvidia drivers.

My Ryzen 7 5700X was pretty much close to Humbug's score, so not sure why you are not beating both our systems!! There is a Ryzen 5 5600x in the table with faster RAM which topped our scores. Plus the Ryzen 7 5800X3D is the leading CPU.

CAT-THE-FIFTH · 6 Jun 2022 at 22:02

Some more updates on the Core i5 10400 system.

AMD Driver: Adrenalin 22.5.2
Nvidia Driver:512.95

RX5600XT Pulse

Corvega=54.67FPS(11710 Drawcalls)
Diamond City=64.33FPS(8033.33 Drawcalls)

RTX3060TI FE

Corvega=61.67FPS(11717 Drawcalls)
Diamond City=72.53FPS(8028.67 Drawcalls)

GTX960 4GB Windforce 2X

Corvega=58.5FPS(11707.33 Drawcalls)
Diamond City=65.8FPS(8027.33 Drawcalls)

imgur.com

Discover the magic of the internet at Imgur, a community powered entertainment destination. Lift your spirits with funny jokes, trending memes, entertaining gifs, inspiring stories, viral videos, and so much more from users.

imgur.com

Edit!!

It appears whilst fiddling around with the rig the CPU cooler got knocked a tad loose for the GTX960 run!

Corvega=61.2FPS(11715.67 Drawcalls)
Diamond City=72.67FPS(8024.67 Drawcalls)

@humbug Good spot on the high temperatures!

CAT-THE-FIFTH · 6 Jun 2022 at 22:25

humbug said:
Your Core i5 10400 is a warm, next time someone says Zen 3 runs hot i'm pointing them to this thread

It's in 6 litre mini-ITX case with one of these 47MM low profile coolers:

Black Ridge - low-profile cooler in cooperation with DAN Cases -

Premium-Markenhersteller von CPU Kühlern, Lüftern, Festplattenkühlern und Zubehör. Die Produkte von Alpenföhn gehören zu den meist gefragtesten Kühlungskomponenten auf dem Markt.

www.alpenfoehn.de

So it does run very toasty. Having said that thanks for the comment - noticed the CPU was running a bit too warm for the last run,so might re-run it.

The Ryzen 7 5700X is in a 12.7 litre NCase M1 with the following cooler:

NH-L12S

Successor to the award-winning NH-L12, the NH-L12S is a compact low-profile CPU cooler that combines excellent compatibility, surprising efficiency and remarkable flexibility. Thanks to its NF-A12x15 PWM, slim 120mm fan, the NH-L12S provides even better quiet cooling performance than the...

noctua.at

It's still low profile at 70MM,and the case has more cooling but I am surprised how well behaved the B2 stepping Zen3 CPUs are. Still get boosts upto 4.65GHZ!!

CAT-THE-FIFTH · 6 Jun 2022 at 22:30

humbug said:
Nice little cooler

Does make one heck of a racket when it spins up!!

CAT-THE-FIFTH · 7 Jun 2022 at 00:41

Maybe the resolution is not being set properly? Graphics driver re-install ?But you seem to be having the worst Zen3 scores for some reason? I have run this a few times and get mostly similar performance.

Humbug's Ryzen 7 5800X/RTX2070 Super,my Ryzen 7 5700X/RTX3060TI and the AT forums person's Ryzen 5 5600X/RTX3080TI are all within a few FPS of each other which sort of makes sense. Someone also mentioned their Ryzen 7 5800X/RTX3080 result on here and it matched up too. The Ryzen 7 5800X3D results are all clustered within a similar grouping at the top. The Zen2 results with an Nvidia GPU also seem to reasonably close together(the Ryzen 9 has higher clockspeeds and was run with faster RAM).

Will have a play around with DXVK and see if that helps boost performance - good spot.

CAT-THE-FIFTH · 7 Jun 2022 at 00:53

Tired9 said:
I have tried all sorts other than a driver reinstall, it would not make much sense that a fresh Windows install (From when I moved the 5800X to the Gigabyte x570 Aorus Elite from the previous MSI B550i MPG Gaming Max WiFi...) Would be worse than the 5600x in the previous MSI board.

The difference I can see though is that the Gigabyte board has no ability to lock in maximum boost clocks like you could with the 5600x with PBO enabled, but that still does not count the fact that this chip is clocked higher but also running a superior RAM configuration which is 2x64-Bit = dual rank.

I have messed with PCI-E Gen revisions to no avail, in fact I lost like 15 FPS moving to Gen 2.0 LOL.

I have turned on and off a 3DMark performance enhancement setting in the bios which is a legacy option and that did nothing either.

I am on the latest bios for the motherboard so latest AGESA too.

The RTX 3070 is having no troubles, the OC applied right now is the most efficient my CPU has ever been only reaching 74c max as opposed to 83c on auto with max boost hitting 4875mhz.

I can reinstall the driver for the GPU but I am at the point where I just know nothing is going to change, especially wnena wrapper makes such a massive difference and the game don't crash.

I am just trying to think why you are getting those results. I am running the Ryzen 7 5700X under a low profile Noctua L12S in an NCase M1 with an Asus B450I Strix - so you should be certainly above me(and Humbug) in performance!