Jump to content

VR Hardware Benchmark late 2020 - Ryzen 2000/5000 - 1080ti vs 3080 vs 6800XT


JayRoc

Recommended Posts

Hey folks,

 

EDIT: All the performance issues discussed in here with the AMD 6000 series are solved with setting your PCIe slot to GEN3 in your mainboard-BIOS with AMD CPU´s!

 

my brother (med0ra) and me spend again a lot of hours testing and benching hardware.

I'd like to thank Cassiop aswell. He made some interesting testing on different RAM speed and compared a Ryzen 5000 to an older Intel.

I will put his findings in here aswell.

 

I'd you guys to participate and help to make this kind of complete hardware guide. I will discripe how we made our benchmarks and if you have the same VR Headset and either a matching graphic card or a matching CPU you could send me your results. I will put them in here aswell.

 

How we testet: We used Gryz's DCS benchmark files which you can find here. Benchmark file 1 was used.

For settings we used the standard DCS VR settings which you can find in the options. The fps were counted and monitored with fpsVR (here - no one of us is related to them). I can absolutely recomend fpsVR to fine your settings in VR, since you can easily figure out if your gpu or cpu are bottlenecking with the settings you are using. 

 

For testing you have to open fpsVR which gives you the follwing windows.

Open the Settings and turn the history on.

 

image.pngimage.png

 

Now go to Basic where you can fin the Reset-Button. Press this button on the start and on the end of the Benchmark als the Benchmark tells you to.

 

image.png

 

The default path for the History-Viewer is: C:\Users\*username*\Documents\fpsVR - as statet in the fpsVR settings above.

The tracks are listet there and should be slightly over 3minutes.

 

Following hardware was used:

 

Ryzen 5900X System:

- Gigabyte B550 Aorus Pro

- G.Skill Sniper 32GB 4x single ranked  @3600 CL19 (OC)

- different SSD for game and operating system

 

Ryzen 5600X System:

- Gigabyte B550M Aorus Pro

- Crucual Ballistix 32GB 4x single ranked @3840 CL16 (OC)

- different SSD for game and operating system

 

Ryzen 2700 System:

- MSI X370 Gaming Carbon

- Kingston HyperX 32GB 2x dual ranked @3200 CL16

 

AMD Radeon RX 6800XT

- Stock

- OC 115% Power Target - 2100Mhz for Ram

 

Gigabyte RTX 3080 Aorus Master 10G

- Stock 1845Mhz 19002Mhz (actual boost up to 1995Mhz)

- OC 1940Mhz 20002Mhz (actual boost up to 2085Mhz)

 

Zotac GTX 1080ti - with Arctic Watercooling 120

- boosts a bit higher than Air cooled ones due to lower temperatures

 

Cassiops findings:

 

 

Quote

 

Chapter One : The CPU

First question : Does the CPU Clock matter : MEH

 

Second question : Does the CPU Matter at all ? Well...

 

Does DCS care about SMT ? (AMD's equivalent of hyperthreading). NO

 

Chapter 2 : the Ram !

 

First : Does the speed of the ram have an impact ? YES. well, a bit

 

Do i need to run ram in Coupled mode (fclock = ram speed/2) or not : YES, BY A LOT

 

Does DCS Care about Ram's timing and subtimings. running at 3733 Coupled ? No.

 

Does DCS Care about running very high frequency of ram ? No.

 

Can DCS run on 16gb of ram in Multi Player ? eeeeehhhh...²

 

 

Here is the Link to his Reddit where he explained everything in detail: Cassiops Reddit
 

Our results:

There are still results missing. But we benched on two computers for over 12h and at some point you just have enough.

I hope we will fill the missing bits, but its really time consuming putting the graphic cards back and forth.

 

 

image.png

 

Pimax can force the Nvidia driver to use fixed foveated rendering. This means that the rendering quality outside of the sweet spot, which is blurry anyways, will be rendered with less quality. As you will notice there is quite an improvement in performance. We had just a short look at it and it seemed to be fine.

There is a discussion about it somewhere in the Forum.

 

image.png

 

Below the OC reffers to the graphics card. DCS is set to SS 1.0 except when written Both.

 

image.png

And some Benchmarks for the Reverb G2 benched on Ryzen 5600X

Standard VR settings aswell.

 

image.png

 

I think the results already available can give you guys a good estimate if and what to upgrade if you´re on a budget and cant get a complete new system. And it can give you a good idea which upgrate is worth it for you.

 

As i meantioned earlier, if you have some hardware that matches in, feel free to benchmark and send me the results.

I´d like to see some Intel system in here aswell.

 

Regarding Frametimes: The Radeon is absolute King and therefore way better for VR!

 

Have a good one, med & jay 

 


Edited by JayRoc
FVR chart was wrong, added G2 results, added fpsVR specifics
  • Like 3
  • Thanks 2

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Link to comment
Share on other sites

Placeholder Frametimes

| F-16C | FC3 | TM Hotas Cougar | CH Products Pro Pedals |

Pc1: AMD 5900x, RX6800xt, 32Gb 3600 CL19, Gigabyte B550 Aorus Pro | Pc2: AMD 2700, GTX 1080ti, 32Gb 3200 CL16, MSI x370 Gaming Carbon

Samsung Odyssey, Pimax 8kx, HP G2 (incoming), Pimax 5k+ (ex), Pimax 4k (ex), Vive Pro (ex), Oculus CV1 (ex), Oculus CV1 (ex)

Link to comment
Share on other sites

  • JayRoc changed the title to VR Hardware Benchmark late 2020 - Ryzen 2000/5000 - 1080ti vs 3080 vs 6800XT

If I understand correctly now, would it do me more to keep the GTX 1080TI and replace my AMD Ryzen 2600X with an AMD Ryzen 5900X or 5800X?
So far a new GPU would be less useful ...

Hardware: Windows 11 64Bit, AMD Ryzen 7 5800X3D, ASUS TUF Gaming X570-Pro Wifi II, 32 GB Ram 3600 MHz DDR4, TUF RTX 4080 OC, M.2 SSD ADATA SX8200 2TB, Meta Quest 2, ASUS TUF VG279QM Monitor, TM HOTAS Warthog , VIRPIL VPC WarBRD Base mit TM Hornet Stick und Saitek Pro Flight Rudder Pedalen.

 

Deutscher Guide zu: Mirage 2000C, MiG-21bis, F5 Tiger II, Mi-8MTV2, F-14B Tomcat, AJS-37 Viggen und Fulgabwehrsysteme

Link to comment
Share on other sites

If you're using a "lower res VR HMD" absolutely yes! The frametimes will also be more consistent which is really nice to have in VR.

If you´re planning on upgrading to a "high res HMD" like G2 or Pimax 8K it depends. A lot of multiplayer? Single player? Frames before quality?

 

In your case i would start with tuning your RAM. And if you keep your Rift - the CPU right after it. That would be the most cost efficient upgrade.

 

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Link to comment
Share on other sites

OK thanks I already thought about replacing the RAM with 3200. I will keep the rift for the time being. I'm mainly in SP. I already thought about getting a Ryzen 3800Xt because of the better single core 4.7 MHz. Or would the 5900X be better for Zen-3 in general? That would still be cheap enough in relation to the GPU, although I wouldn't need that many cores. Your tests help a lot, I've missed something like that until now. Really great job.

Hardware: Windows 11 64Bit, AMD Ryzen 7 5800X3D, ASUS TUF Gaming X570-Pro Wifi II, 32 GB Ram 3600 MHz DDR4, TUF RTX 4080 OC, M.2 SSD ADATA SX8200 2TB, Meta Quest 2, ASUS TUF VG279QM Monitor, TM HOTAS Warthog , VIRPIL VPC WarBRD Base mit TM Hornet Stick und Saitek Pro Flight Rudder Pedalen.

 

Deutscher Guide zu: Mirage 2000C, MiG-21bis, F5 Tiger II, Mi-8MTV2, F-14B Tomcat, AJS-37 Viggen und Fulgabwehrsysteme

Link to comment
Share on other sites

You will get a lot more for your money with 5600x instead of 5900x especially in SP. The 3800x is somewhat bad, because of the 105W TDP (3700x has 65W) it is a hot CPU which needs a very good cooler and is not really faster than the 3700x. I think with 3800xt it will be the same problem. You should check that. So i would go 3600x, 3600xt, 3700x, 5600x. 5900x is only worth it, when you dont care about the price. Maybe 5900x is helpfull in MP. I think we will do a short comparission. But first we have to fly a little bit 😄

  • Like 1

| F-16C | FC3 | TM Hotas Cougar | CH Products Pro Pedals |

Pc1: AMD 5900x, RX6800xt, 32Gb 3600 CL19, Gigabyte B550 Aorus Pro | Pc2: AMD 2700, GTX 1080ti, 32Gb 3200 CL16, MSI x370 Gaming Carbon

Samsung Odyssey, Pimax 8kx, HP G2 (incoming), Pimax 5k+ (ex), Pimax 4k (ex), Vive Pro (ex), Oculus CV1 (ex), Oculus CV1 (ex)

Link to comment
Share on other sites

If someone had told me at the start of the year that I would be seriously considering a 100% AMD upgrade I would have laughed in their face!

I am looking to get a 5600X, hoping to see a Aorus or ASUS EK 6800XT appearing.

Windows 10 Pro | ASUS RANGER VIII | i5 6600K @ 4.6GHz| MSI RTX 2060 SUPER | 32GB RAM | Corsair H100i | Corsair Carbide 540 | HP Reverb G2 | MFG crosswind Pedals | Custom Spitfire Cockpit

Project IX Cockpit

Link to comment
Share on other sites

Good that you buy what makes sense for you and not only what you think is best. I am happy to be back on full AMD. The stock 6800xt is a really good card, performance and noise wise. I would buy any 6800xt if you can get them for a decent price.

  • Like 1

| F-16C | FC3 | TM Hotas Cougar | CH Products Pro Pedals |

Pc1: AMD 5900x, RX6800xt, 32Gb 3600 CL19, Gigabyte B550 Aorus Pro | Pc2: AMD 2700, GTX 1080ti, 32Gb 3200 CL16, MSI x370 Gaming Carbon

Samsung Odyssey, Pimax 8kx, HP G2 (incoming), Pimax 5k+ (ex), Pimax 4k (ex), Vive Pro (ex), Oculus CV1 (ex), Oculus CV1 (ex)

Link to comment
Share on other sites

Last time I owned ran with AMD was with a HD 5950 back in 2009, absolute beast of a card in it's day. It will be my first AMD CPU though.

Looking forward to AMD and Ray Tracing performance in Cyberpunk will look like, can't be any worse than my current 2060 Super. 

Windows 10 Pro | ASUS RANGER VIII | i5 6600K @ 4.6GHz| MSI RTX 2060 SUPER | 32GB RAM | Corsair H100i | Corsair Carbide 540 | HP Reverb G2 | MFG crosswind Pedals | Custom Spitfire Cockpit

Project IX Cockpit

Link to comment
Share on other sites

17 hours ago, JayRoc said:

 

How about the GPU memory usage? Do you think the 10G of 3080 is enough for high res HMD, like G2?

Hope you can add 3090 in the benchmark.

Many thanks!

13700k @5.6GHz | 32GB DDR4 4000MHz CL14 | ASUS Z690 TUF  | GIGABYTE RTX4090 Gaming OC | Samsung 970PRO 512G + 980PRO 2T

GIGABYTE M32U | HP Reverb G2

Warthog HOTAS | TPR | JetSeat | TrackIR5

Windows 11

Link to comment
Share on other sites

Would be nice if someone would bench that und submit me the results.

 

In the Benchmark 6.9Gb or so was used. In Multiplayer it can a issue when switching airframes a lot the game will load news stuuf and not unload the old stuff and that results

in really bad performance. If thats not a Problem or you dont switch airframes on servers you wont have a problem.

There is a whole Thread on that toppic.

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Link to comment
Share on other sites

1 hour ago, JayRoc said:

Would be nice if someone would bench that und submit me the results.

 

In the Benchmark 6.9Gb or so was used. In Multiplayer it can a issue when switching airframes a lot the game will load news stuuf and not unload the old stuff and that results

in really bad performance. If thats not a Problem or you dont switch airframes on servers you wont have a problem.

There is a whole Thread on that toppic.

 

How do you activate the test once you're in autopilot mode?

Win 11Pro. Corsair RM1000X PSU. ASUS TUF Gaming X570-PLUS [WI-FI], AMD Ryzen 7 5800X 3D, Sapphire Radeon RX 7900 XTX Nitro+ Vapor-X 24GB GDDR6. 32 GB G.SKILL TridentZ RGB Series (4 x 8GB) RAM Cl14 DDR4 3600. Thrustmaster HOTAS WARTHOG Thrustmaster. TWCS Throttle. PICO 4 256GB.

WARNING: Message from AMD: Windows Automatic Update may have replaced their driver by one of their own. Check your drivers.

M-2000C. Mirage F1. F/A-18C Hornet. F-15C. F-5E Tiger II. MiG-29 "Fulcrum".  Avatar: Escadron de Chasse 3/3 Ardennes. Fly like a Maineyak.

 

Link to comment
Share on other sites

@K1nsal
We will bench the G2 on the other systems aswell. So if there is any unknown problem, we will see it.

| F-16C | FC3 | TM Hotas Cougar | CH Products Pro Pedals |

Pc1: AMD 5900x, RX6800xt, 32Gb 3600 CL19, Gigabyte B550 Aorus Pro | Pc2: AMD 2700, GTX 1080ti, 32Gb 3200 CL16, MSI x370 Gaming Carbon

Samsung Odyssey, Pimax 8kx, HP G2 (incoming), Pimax 5k+ (ex), Pimax 4k (ex), Vive Pro (ex), Oculus CV1 (ex), Oculus CV1 (ex)

Link to comment
Share on other sites

14 minutes ago, med0ra said:

@K1nsal
We will bench the G2 on the other systems aswell. So if there is any unknown problem, we will see it.

 

1 hour ago, JayRoc said:

@Thinder

Added that Info in the first post.

 

@K1nsal

Good decission.

 

 

 

Thank you both for your hard work I know Bench testing is not the most entertaining thing. 

  • Like 1
Link to comment
Share on other sites

What systems are you building? I currently have a 3800X 32 GB RAm SSd's radeon 5700 non xt 

asus tuf gaming plus wifi 

 

currently a quest 2 with a g2 on the way

 

possibly looking at a 5600x and a 6800xt

 

gonna start from scratch new ssDs and install the standalone

 

 

I have had OK performance so far gonna start over rebuild windows 

 

I fly the harrier and its a HOG performance wise

Link to comment
Share on other sites

@ JayRoc

I got a little bit of a funny Situation I picked up a used 1080Ti EVGA Hybrid and two days later I got a call from a local Store that he put a 6800XT aside for me (since I bugged him for over two weeks)

I have a 2700 (non X) OC to 3,9 GHZ SMT off with 32GB @ 3333Mhz CL15.  As you can imagine I kept swapping GPUs for and back. And even borrowed a 4K screen to see the differents.
1st of all I awas amazed about the 1080TI upgraded from a 1070TI.
Compared to the 6800XT with really everything on Ultra High (also Shadows) 2x MSAA 2x SSAA and SSLR in the worst possible place I got 3fps less with the 1080TI at 1440p. I think 39 vs 36 fps
 

If you take out SSLR the gap widens, if you take out SSAA you are 77 fps vs 64 lowest.
4K with 2x MSAA made the biggest difference lowest 6800XT was 54 vs 37 on the 1080Ti

(all tests were done Single Player, low to the ground with lots of trees and explosions/ particles around)
(Also I am only interested in the lowest FPS since I cap at 60fps due to Track IR, because I can't reach 120FPS stable without scarifying all the eye candy)

I could imagine the CPU is probably the bottle neck here even though both cards are at 99% when measuring those lowest FPS situations.

Here is my question. on your chart (benchmark) when testing the 6800XT on a 2700x vs 5900x is the GPU utilization in both cases also 99% or is it lower on the 2700x?


Edited by HansPeter1981

My System specs: Cpu 5800x3d liquid cooled GPU 7900XTX Ram 32GB 3600mhz cl16 Motherboard B550M MSI, Windows 10 PRO on NVMe Drive, DCS on its own SSD, Monitor Philips 32" 4k curved adaptive Sync framerate capped at 59fps, Trackir 5, VKB 

Link to comment
Share on other sites

To be honest we didnt keep track of that. But i guess we have different settings in DCS. Some are more CPU stressing and some more GPU stressing.

With the 2700 you will have a bottleneck as you can see in the charts above.

With all the different GPU´s the 2700x doesnt get faster. Its the CPU.

 

In your case i would try to give the 6800XT back and get a faster CPU if you need more frames.

If you have a 450 chipset it could be that your mainboard can handle a Ryzen 5000.

 

 

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Link to comment
Share on other sites

On 12/13/2020 at 9:26 PM, joseridingclub said:

What systems are you building? I currently have a 3800X 32 GB RAm SSd's radeon 5700 non xt 

asus tuf gaming plus wifi 

 

currently a quest 2 with a g2 on the way

 

possibly looking at a 5600x and a 6800xt

 

gonna start from scratch new ssDs and install the standalone

 

 

I have had OK performance so far gonna start over rebuild windows 

 

I fly the harrier and its a HOG performance wise

Processor: AMD Ryzen 7 3800X (8-Core) 4.5 GHz Turbo(waiting for 5600 to restock mainly for the SAM as its on like 5 to 8 percent faster)

570 MOBO

1x Radeon RX 6800 XT 16GB

32 GB of RAM

Link to comment
Share on other sites

Gonna send back some ram giving problems get a new kit had 4x8gb gonna go with 2x16 do you have any recomndations please be specific!

 

Jayrock what modules would you get for a DCS dedicated system>> DDR 4000 2x16gb ram doesn't seam to be as in short a supply as video cards and CPU's can someone link me to a good ram kit?


Edited by joseridingclub
Link to comment
Share on other sites

4 hours ago, helios123 said:

Does anyone have an explanation as to why the 6800XT is performing better than RTX 3080, when in most non-vr games 3080 is on par if not slightly better?

Actually in 4K (lots of pixels..) "all" benchmarks has the 3080 ahead by a margin. But VR benchmarks are almost nowhere to be seen. That what is so great with JayRoc & co's efforts.

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...