Skip to content
View in the app

A better way to browse. Learn more.

The AVSIM Community

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

GPUMark, an FSX GPU benchmark

Featured Replies

I've been wanting to do this for a long time.

This is a benchmark flight, very much like FSMark11 & FSMark07 but focused on the GPU. The idea is to figure out which GPU models & architectures run better at high IQ levels in FSX.

It doesn't require a clean FSX install or anything like that, the only requirements are:

 

- Max out your Cloud Coverage Density (8 or more)

- Nvidia users, load the .nip Inspector profile provided (right click on the GPUMark.nip file, select "open with..." and browse for the nVidia Inspector executable)

- Run the benchmark like you would FSMark11

- Set your resolution to 1920x1080

 

Let's see what's the best GPU to crank up those AA settings!

 

Download GPUMark

  • Replies 86
  • Views 14.4k
  • Created
  • Last Reply

Top Posters In This Topic

  • Author

This is my score:

 

GTX580 @ 900MHz / 2200MHz (memory)

Frames, Time (ms), Min, Max, Avg

5284, 300000, 15, 20, 17.613

 

GTX580 Stock clocks: 772MHz / 2004MHz (Memory)

Frames, Time (ms), Min, Max, Avg

4569, 300000, 13, 18, 15.230

 

Scaling 1:1 with core clock

 

...so it seemed like the obvious question that arises is, what role does memory speed play here?:

 

GTX580: 900 / 1800MHz (Memory - downclocked)

Frames, Time (ms), Min, Max, Avg

5282, 300000, 15, 20, 17.607

 

Interenstingly, memory speed is a non factor!

Thanks Dazz

 

Shall make a test

 

Hasse

  • Author

Thanks Dazz

 

Shall make a test

 

Hasse

 

Tak Hasse!

This is my score:

 

GTX580 @ 900MHz / 2200MHz (memory)

Frames, Time (ms), Min, Max, Avg

5284, 300000, 15, 20, 17.613

 

GTX580 Stock clocks: 772MHz / 2004MHz (Memory)

Frames, Time (ms), Min, Max, Avg

4569, 300000, 13, 18, 15.230

 

Scaling 1:1 with core clock

 

...so it seemed like the obvious question that arises is, what role does memory speed play here?:

 

GTX580 Stock clocks: 772MHz / 1800MHz (Memory - downclocked)

Frames, Time (ms), Min, Max, Avg

5282, 300000, 15, 20, 17.607

 

Interenstingly, memory speed is a non factor!

 

I think you mean you did the last test (underclocked) with the core at 900MHz.

 

Another problem with the test is you've got people using different cloud textures and resolutions. You could get people to test with the 1024x1024 32bit clouds that come with HDEv2.0.

Luke Harvest

  • Author

I think you mean you did the last test (underclocked) with the core at 900MHz.

 

Correct, thanks a lot Luke ^_^

 

Another problem with the test is you've got people using different cloud textures and resolutions. You could get people to test with the 1024x1024 32bit clouds that come with HDEv2.0.

 

Again, you're right, I edited the first post to include the requirement for 1920x1080

Not sure how much having 3rd party textures can affect the scores, I'll do some tests

Dazz

Shall i do the test with both 670 4gb and 680 4gb ?

 

Hasse

 

Correct, thanks a lot Luke ^_^

 

 

 

Again, you're right, I edited the first post to include the requirement for 1920x1080

Not sure how much having 3rd party textures can affect the scores, I'll do some tests

 

If you've got REX try running 1024x1024 DXT5 and 4096x4096 32bit clouds, with high AA there should be a massive difference.

Luke Harvest

  • Author

Dazz

Shall i do the test with both 670 4gb and 680 4gb ?

 

Hasse

 

Well, if you have to go messing with swapping cards, don't bother man. One of those is more than enough

 

 

If you've got REX try running 1024x1024 DXT5 and 4096x4096 32bit clouds, with high AA there should be a massive difference.

 

testing it right now, but yeah, you're most likely right

Well, if you have to go messing with swapping cards, don't bother man. One of those is more than enough

 

I swappted CPU today and shall swapp back the 2700K , if a make a fast swap with the cards run the tests tomorrow.

 

The 3770K gone retaire until i am ready with the 24/7 SS.

Hi Dazz

Dont know what a doing wrong.

Cloud 8 evry setting as FSMark11 exept 1920x1080 in FSX and NV CP.

Inspector file loaded ok.

KFA2 680 4gb Std 1110mhz/3000mhz

 

3770K @4.7 downclocked

 

Frames, Time (ms), Min, Max, Avg

9560, 300000, 27, 36, 31.867

 

  • Author

Hi Dazz

Dont know what a doing wrong.

Cloud 8 evry setting as FSMark11 exept 1920x1080 in FSX and NV CP.

Inspector file loaded ok.

KFA2 680 4gb Std 1110mhz/3000mhz

 

3770K @4.7 downclocked

 

Frames, Time (ms), Min, Max, Avg

9560, 300000, 27, 36, 31.867

 

Thanks Hasse

 

Do you have REX installed?

Can you please load the flight and check if your GPU load is close to 100%? no need to rerun the test, just a quick check

GPU load in PresX 75%

I have rex shall i put in 4096 and change cfg?

what teme in rex?

 

Hasse

  • Author

GPU load in PresX 75%

I have rex shall i put in 4096 and change cfg?

what teme in rex?

 

Hasse

 

In my tests yesterday it didn't really matter what REX theme I had installed, the scores were unnafected.

That GPU load is too low, something's not right.

Can you please confirm you have 4x Sparse Grid Supersampling and 16xS antialiasing set in Inspector?

Can you see lots of clouds around the plane in the test?

Create an account or sign in to comment

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.