The ACC GPU Benchmark Thread

I'll do runs in a couple of day. Working out of town so i wont be home until Thursday.

 

I really wonder if your GPU is that much faster than my Radeon RX6800 or the Ryzen 5600X is that much faster than my Ryzen 3600X. Or just both slightly faster.

 

Ok so I finally did it correctly,

i7 10700k Stock clock
MSI 3070 Suprim X OC (+1oo mhz core, +200 mhz memory)
16Gb Ram 3200Mhz (dual channel)

1080p - Epic

2021-02-24 03:43:04 - AC2-Win64-Shipping
Frames: 11160 - Time: 60000ms - Avg: 186.000 - Min: 148 - Max: 280

2021-02-24 03:49:14 - AC2-Win64-Shipping
Frames: 11301 - Time: 60000ms - Avg: 188.350 - Min: 148 - Max: 280

2021-02-24 03:50:51 - AC2-Win64-Shipping
Frames: 11287 - Time: 60000ms - Avg: 188.117 - Min: 148 - Max: 281

Average across three runs is 187.489 fps

 

I guess it's both + the fact that I overclocked CPU and GPU and that ACC probably runs a bit better on Nvidia cards. I also use probably the best RAM there is and mobo is also a bit better I guess. A lot of guesses though. Maybe some people who do this as a hobby with much more knowledge could give a better answer. I only recently informed myself again on pc builds after my previous build from 7 years ago.

 

I see you run default voltage. My temps are about 72/74. Will try that and up the power limit to max.
Just did a test with core voltage default and power limit 107% and had about the same average as you. My GPU temp goes to 70/72 though. Still pretty safe.

 

Im not able to change my voltage. I checked the boxes but no dice.

Yeah my temps haven't exceeded 70°C while running heaven benchmark for 20min straight so she's comfortable for sure.

I have a Dark rock pro 4 cpu cooler and 6 case fans. 3 intake up front, 1 exhaust in the back and 2 exhaust up top.

I bumper my memory up to +250 mhz and been playing ACC all morning and no issues.

I'll keep tinkering with the settings to see how far i can go.

 

The CPU in his case is a big jump as its much faster single core performance compared to 3600X and its also helped significantly by the VERY quick RAM, 3600 C14 gives 7.77nm latency, this is just about as fast as you could get plus having 4 sticks makes a dual rank memory configuration so bonus there as well. Anything that is CPU limited is affected by RAM speed, and basically all sims are CPU limited with ACC being the worst of them all

 

That's plenty more fans than I use. I could only mount 1 fan on my Noctua cpu cooler since ram sticks are coming too high. Might be able to mount with an offset. Also ran out of fan connectors on mobo. Might look for some 4pin splitters. I have 3 intake fans in front and 1 exhaust at the rear + the 1 cpu fan.

 

Can’t fit a 120mm fan on the front? I have a D15S with a 140mm fan in the middle and a 120mm fan on the front with G Skill RGB RAM

 

I had the same issue with one of the fans on the Noctua cooler rising too high because of RAM. I replaced the fan in question with this one: https://noctua.at/en/nf-s12a-pwm-chromax-black-swap.
You could consider removing the bottom intake fan and using an extra CPU fan instead. I seem to recall Steve over at GamersNexus saying that the airflow is better without the bottom fan because in some cases it will cause warm air to flow up from the GPU. Or something.

 

I use two splitter. 3 to 1. So all the case fans only take up two headers on the board.

Im using a cooler master n600 case.

I just wanted to make sure cooling wasn't ever going to be an issue.

 

Hi guys...I started playing ACC again after a long break. I found the Option "Mirror FPS" and played around with it. I rembered that there was always the problem with the mirrors giving a huge FPS hit. I set it to 1/2 and ran a worst case benchmark at night, in the storm and with 24 opponents. The FPS were significantly higher than I remember from that situation, wothout this mirror FPS option. I could even set shadows from mid to high and the FPS were still much higher than they used to be.

Can somebody confirm?

 

I think the performance in general has improved since you've been away. I also tested the mirror thingy but can't really say I got a lot of boost out of it even when changing mirror settings. Maybe I didn't test thoroughly enough.

 

Keep in mind that you generally want positive pressure in your case to keep the dust out so more air going in then out. That way instead of being a vacuum and sucking in dust that in turn clogs up your heat sinks and fans it pushes it out.

In your case I’d make sure your intake fans are spinning faster then your exhaust fans

 

Sounds good. I'll definitely look into it.

 

Either way... good stuff.

I swapped GPUs again and run an RTX 2070 now. Performance is really nice, maybe I'll rerun the benchmark and look into the mirror FPS thing.

Has someone played around with NVIDIA DSR? Is it working in ACC?

 

Dynamic Super Resolution? Pretty sure that works as well in ACC as in any other game. I haven't tested it specifically in ACC since I run triples and don't want to increase the GPU load further.
It should give the same performance as running a native res monitor in the resolution you choose.

 

DSR works well but you can just use the ingame resolution scaling as well.

 

Got my hands on a RX6700XT gonna try to provide you some results soon guys, anyways has anyone tested the new AMDs with SAM function? Looks like boosting quite significantly in some games, how about ACC? Saw a 13% boost on 6800 test by @Gobrel anyone else tried it?

 

As promised tested on a Gigabyte Gaming OC RX6700XT together with:

• Ryzen 3600X
  
• MSI B450 Tomahawk
• Corsair Venegance RGB Pro DDR4 3200 // CL16

And well not quite pleased being honest, there's a boost over the last generation of Radeon, but well I think I just thought I would get more with SAM. Nowhere near the boost of SAM that @Gobrel achieved with turning on this option, despite the same processor and RAM, and well I'm not sure what could have caused it. Anyways here are my tests:



note the results in this table are based on my custom dash camera (noticed I didnt reset to the default cockpit cam on my 5700xt runs, and didnt want to play with repluggging the gpu)



So SAM generally gave almost nothing here in case of ACC, and it was marginal when it comes to other games. Tested other presets on my custom dash cam, and the results are that most likely I'll set them between med-high. So yeah percentage wise the the upgrade from 5700XT to 6700XT did give a boost, the work culture of the new Radeon is waaaay better staying cool at 60/80 'C temps, instead of 75'C and not shy of hitting 100 on the hotspot on the old gen, so this is way better too. Anyway thought that SAM would give a better boost in performance, but maybe that's some issue with my board and the Radeon driver - 6700XT doesnt have a stable build yet, only a beta driver. All in all Im happy as I bought the 6700xt for 800 euro and sold the 5700xt for 900 which is absurd, but as I heard the 5700XT likes to mine ETH.


Here are the results with the 1st page settings - after I did reset to the default cockpit camera:

2021-03-27 14:37:46 - AC2-Win64-Shipping - 1440P - default cockpit cam
Frames: 5399 - Time: 60000ms - Avg: 89.983 - Min: 64 - Max: 141

2021-03-27 14:39:46 - AC2-Win64-Shipping - 1080P - default cockpit cam
Frames: 7506 - Time: 60000ms - Avg: 125.100 - Min: 88 - Max: 213

So cant compare everything nicely like before, but theres a visable improvement over the dash cam which has to display more of the surroundings, a gain of 10 fps in 1080p, and 8 in 1440p. So yeah if someone wants to compare to other graphic cards then compare this.