The ACC GPU Benchmark Thread

Nah, it was dual channel. Running dual - if you can - is such a basic thing that I don't think it needs further investigation.

 

No, doesn'need investigation, but it should be mentioned how it was tested, don't see it in the sheet.

 

Also how much of that 8gb is actually left with system processes

 

Don't know. I only know that on my system, something between 5 and 6 GB RAM usage was max in ACC. Firefox and some tools in the background were never an issue even with 8 GB.

 

I had 16, just got 16 more and it made no difference in the avg FPS.

It did increase the 0.1% and 1% tough

And windows is booting faster lol

 

Please write down your CPU, GPU, RAM, and ACC version.

2019-09-29 22:57:10 - AC2-Win64-Shipping
Frames: 6430 - Time: 60000ms - Avg: 107.167 - Min: 72 - Max: 115

2019-09-29 23:01:17 - AC2-Win64-Shipping
Frames: 6175 - Time: 60000ms - Avg: 102.917 - Min: 90 - Max: 115

2019-09-29 23:05:03 - AC2-Win64-Shipping
Frames: 6123 - Time: 60000ms - Avg: 102.050 - Min: 82 - Max: 113

cpu i5 4570, 16 ram, rtx 2060 6g, acc v1.0.10

 

An RTX 2060 beating 2080 and 1080Ti? Something isn't right...starting to wonder if the GFX settings preset works the same way for everyone...maybe some of the settings remain untouched when switching from custom to epic. @pankykapus

The recent results are way out pf proportion :-(

 

Where is the 2060 beating the 1080ti?
My 1080ti was min 78, average 112,max 174 compared to 82,102,113!

 

His best run was 107. Compare this to post 411.

Even your 112 vs the 102 would be too small of a gap for those ECUs.
Somebody should check if the Epic preset reliably sets all parameters to a fixed value or of some remain unchanged from the custom preset.

Edit: Here's a GPU performance overview by Gamestar:



https://www.gamestar.de/artikel/nvi...aytracing-fuer-die-massen,3339044,seite2.html

The relations between GPU models are nowhere to what we currently have in here. For example: 2080 Ti should only be 10 % quicker than a 2080 but in our test it's 20 %. RTX 2060 should be 20 % slower than a 1080 Ti but in our test it's equally fast (looking at post 411).

We also have strong deviations between identical models - Ole's 2060 made only 95 FPS etc.

Edit 2: APC900's 1080 Ti even only made 99 FPS AVG.

 

Anyone getting Ram and VRam usage like this ?

Doesn't boad well for 8GB cards or systems with only 8GB !

 

Whoa...that doesn't seem right, at all. Especially the VRAM usage. I see that you run high res but still it seems excessive.

But seeing as APC900 has 11 GB VRAM and 32 GB system RAM and still a pretty poor result for a 1080 Ti (99 AVG) - it doesn't seem like a general thing.

In the past, I was between 4 and 5 GB VRAM usage and about the same RAM usage. But I will test this again. It would also explain the lack of smoothness that got introduced somewhere between 1.0.8 and 1.0.10.

There were several reports about VRAM issues but I don't know if any of them were properly understood and fixed.

 

I reply to myself (in a clear example of egocentrism):

Yes, the CPU is key. Someone offered me a 2700K for £60 and I took it. I changed the thermal paste of course, made the heart transplant and is now giving me 60 fps stable in any game, including the brand new Ghost Recon Breakpoint (Ultra settings) that comes out next friday. I'm also running Assassin's Creed Origins in Ultra. Very happy with this CPU (all this without overclocking card or CPU, blessing silence)

COOLING is another factor. When the side lid of my Alienware is open, with a small silent usb fan aiming to the interiors of the computer, I gain 3-5 fps. So I'm going to leave the lid open.

 

Sorry didn't notice the type of weather until I saw everybody checking Sizes ) Everything is a competition...

 

Ok, thanks for clarifying. Post 1 is quite clear so maybe practice reading more thoroughly We didn't compare our d*cks either. If you re-read the posts after yours, it's mainly about finding the reason for the non-plausible test results. If we hadn't discussed, we would now have a "magic 2060" in our chart.

 

Its a Picture Nothing to read.. Get your mind out of the gutter... My computer is better than yours is that better ) see you on the track

 

Well now we've gone there, i didnt really want to brag but ive been told that mine is rather ......maybe in PM we can share pics






Im joking of coarse

 

Did a test here. Heavy rain vs clear sky gives no different results (nothing out of the usual 3fps avg scatter).

Cooling certainly is playing a huge factor here. I have my gpu fans custom spinning at 100% since 60 deg to guarantee no thermal throttling

Of course cpu is liquid cooled by a kraken

 

Thank you. So the investigation is ongoing...will do some further tests later this week.
My GPU keeps itself at 65 and I haven't seen any throtttling, yet. Could even push the clock to stable 1970 MHz.

 

nice deal but there are plenty of games you’ll never see 60fps with that CPU

 

If I can play Ghost Recon Breakpoint without being optimized (comes out officially this friday) and I can play ACC at 60 fps, without overclocking card or CPU, maybe there are not many future games that I will see at 60 fps, but everything that has come out until now I probably will.

However, I see your point. Is the old socket 1155. I can't play 4k and I won't be able to run anything that comes out next year for sure. And that is true, but for me is all right, I don't need to play new exciting games in PC (I have the PS4 for that ), to be honest I didn't even buy the PC, someone gave me the Alienware for free (lucky bastard). I'm just improving it. I've got a GTX 1070 for £120 and this 2700k for 60. That is my investment. When I sell the remaining parts I will actually recover most of the money.

All I wanted is to run ACC comfortably. Now I just need to make it look good, at the moment it doesn't look that good. The driving is very nice, it has a very nice feeling, but even Project Cars 2 looks better than ACC. Someone said that there is a balance in the graphics to make the game look good, now I have to find that.