Directx 10 test fails at 1680 x 1050

[khayman]

Hi, as the title says, I'm having a problem with the DirectX 10 test which is "unable to run at the desired resolution" - that's what the software says - and penalizes the outcome of 3% when I'm able to make it run at a different resolution (1680 x 990). Basically. when the test screen comes up it's all black and frozen, along with everything else (meaning I cannot use my pc) until I call up the task manager via hot keys (ctrl+alt+del), then drag the test window, even very slightly, and only then does the actual test begin. That's when I can read the different resolution warning; this happens both when I run the test singularly and when I run all tests. I've tried running my GPU at both stock and OC clocks but I get the same outcome; same goes for my CPU: I've run it @ 4.2, 4.4, 4.5, 4.6, 4.7 and 4.8 Ghz and I still get the same error. Any help would be appreciated.
Here's my system:

Asrock Z97 Extreme6
Intel I7-4790K @ 4.2 (underclocked) + Cooler Master Nepton 280
G.Skill Tridentx 2400 Mhz DDR3 2x8GB
Asus GTX 960 Strix Direct CU II OC Edition
Samsung SM951 256GB
OCZ ZX1000W
Windows 8.1 64-bit English

Please note that both OS and drivers are up-to-date!

While I'm here, I may as well report that Passmark's hard disk speed readings are extremely inconsistent; meaning I get very different results independently of the test settings but my SSD, I assure you, always performs the same way or unnoticeably differently.
Thanks, Daniel

[David (PassMark)]

What resolution are you running your monitor at (while on the Windows desktop)?

Samsung was shipping RAM caching software with their hard drives, do you have that installed? As that certainly messes with the results. Even when we ask the O/S for no caching, it gets cached anyway and you end up measuring the RAM speed.

[khayman]

Resolution is the recommended at 1920 x 1080.
As far as the hard disk goes, the SM951 comes from a different branch of Samsung's and they do not provide the Magician software; so no caching.

What resolution are you running your monitor at (while on the Windows desktop)?

Samsung was shipping RAM caching software with their hard drives, do you have that installed? As that certainly messes with the results. Even when we ask the O/S for no caching, it gets cached anyway and you end up measuring the RAM speed.

[khayman]

One other thing: my monitor is actually a tv and it's connected via hdmi cable....

[khayman]

One other thing: my monitor is actually a tv and it's connected via hdmi cable....

I've tried switching monitors but it doesn't make any difference.
And there's another problem:the direct 2d test is also very inconsistent; I've tried various different gpu settings but the results are ambiguous and I'm unable to pinpoint the exact reason why that occurs. I think this software simply doesn't like my hardware.

[David (PassMark)]

We are aware of a issue with the Direct2D test in Win8. It looks like a operating system bug as far as we can tell.

If you go into the Windows control panel and change the monitor's resolution, does the video card / monitor combination that you have support 1680 x 1050?

[khayman]

We are aware of a issue with the Direct2D test in Win8. It looks like a operating system bug as far as we can tell.

If you go into the Windows control panel and change the monitor's resolution, does the video card / monitor combination that you have support 1680 x 1050?

Yes, it does.

[khayman]

We are aware of a issue with the Direct2D test in Win8. It looks like a operating system bug as far as we can tell.

If you go into the Windows control panel and change the monitor's resolution, does the video card / monitor combination that you have support 1680 x 1050?

Ok, I can confirm that Windows 10 has solved the Directx 10 problem; however, the Direct 2d one still persists. To be honest, there is a slight improvement but the result is still nowhere near where it's supposed to be.

[David (PassMark)]

We don't really know the cause of the issue. But thanks for the follow up.

[khayman]

Ok, I can confirm that Windows 10 has solved the Directx 10 problem; however, the Direct 2d one still persists. To be honest, there is a slight improvement but the result is still nowhere near where it's supposed to be.

Also, the SSD speed readings are still off, way off.
Edit: I went into Edit/Preferences...drive under test I shifted from "user appdata" to C local drive then "test duration" anything but short will do the trick and, lastly, "number of iterations" I did 2 instead of 1; this seems to have solved the hdd problem.

[David (PassMark)]

Yes, long tests will give better results in general (less margin for error).

The folder used shouldn't matter much (unless you have 3rd party software, like Anti-virus programs, watching and scanning files in specific folders).

[khayman]

Yes, long tests will give better results in general (less margin for error).

The folder used shouldn't matter much (unless you have 3rd party software, like Anti-virus programs, watching and scanning files in specific folders).

Still nothing on the Direct 2D test; I can't tell what the problem is. As I've already said, there's been a slight improvement since I installed Windows 10 but it's still very far from where it used to be: former results were over 36, now I get 19.9 (16.2 with W8.1). I really hope somebody comes up with a solution here... it's very frustrating.

[khayman]

Nevermind; I've decided to uninstall Performance Test: I'll wait for the next edition!

[Tim (PassMark)]

Microsoft have confirmed the Direct2D slow down is a windows bug.

[khayman]

I understand. A bug not yet fixed, even with Windows 10 (btw, I've reverted to W8.1 in the meantime), and if we're gonna wait for them to rectify this problem, we'll likely grow a white beard before they do, lol.
Yesterday, just for kicks, I took out my "old" Samsung 850 pro and installed W8.1 on it, minus 250 ish updates (including Office's though), and ran PT: Direct 2D scored 41!!! with frame rates from 40 to 50.... I was amazed; the overall score was just under 6250 which is very close to my SM951's (6412, ran today: D2D = 15.3) which is 3 times as fast as the 850 pro.
They could at least tell us which update is causing the bug...