Fiery

It is still the case I'm afraid.

Do you use Windows 10 64-bit, with latest patches applied? Thanks, Fiery

It seems the issue is the lack of administrative privileges. Make sure to check that the user who's running AIDA64 has got administrator rights, and that you've got KERNELD.X64 file present in the AIDA64 installation folder.

Both of you: please apply the Windows 7 AES security patch (if it's not installed yet) named KB3033929, then update to the latest AIDA64 beta available at: http://www.aida64.com/downloads/latesta64xebeta After upgrading to this new version, make sure to restart Windows to finalize the upgrade. Let us know if it helps. Thanks, Fiery

Please note that 950 Pro is extremely slow when you use the built-in (default) Windows 10 NVMe driver called StorNvme. It provides over 1800 MB/sec sequential read but less than 100 MB/sec sequential write performance using that driver. However, after installing Samsung's own NVMe driver for 950 Pro, the same drive accelerates up to 800 MB/sec write speed. So it's really not a great idea to stick to the default NVMe driver of Windows

I understand your issue, but I'm not sure if there're any ready-to-use solutions that would fit your requirements LCDInfo sells a nice colour panel that may fit the bill (called USBD480), but it is just the panel, not a full device that could be slotted straight into a double 5.25" bay. As crazy as it may sound, your best bet might just be a cheap Android based phone/phablet that you could hack into the drive bay using a 3D printed (or otherwise sculpted) plastic frame. Then you could use AIDA64 RemoteSensor or Odospace: https://forums.aida64.com/topic/2776-display-pc-sensor-values-on-android-device-connected-via-usb-odospace-remote-panel/

It means the issue caused by a bug in the OpenCL compiler which is part of the OpenCL driver which is part of the video driver for your GPU. Try to update your video driver, it should help.

1) Yes, but it's still not the utilization of the GPU. GPU is a hardware component, and the listed items are software components. AFAIK it's not possible to measure GPU load on Intel GPUs. 2) Are you sure it's only one 2 GB chip soldered to the motherboard of your system? Anyway, it's best to simply run the AIDA64 Cache & Memory Benchmark, and see how the memory can perform. If it can push more than 10667 MB/sec (while running at DDR3-1333 speed grade), then it's definitely using a 2-channel configuration.

Previous AIDA64 versions didn't have the compatibility / supportedOS tags in the application manifest. We've had to add them to properly support Windows 10. However, by adding them to the manifest, it seems Windows drivers behave differently. Apparently Intel's legacy video drivers cannot handle such situations. We've checked, and it's not just AIDA64. The latest version of OpenGL Extension Viewer simply crashes when you start it under 64-bit Windows 10 with a Clarkdale IGP

1) What GPU-Z shows is not the GPU load, but the Direct3D load. 2) N2830 does support up to 2 memory channels. Do you mean your system utilizes only one memory channel out of the two?

Does it hang at the GPU test or CPU test? In other words: left column or right column on the GPGPU Benchmark Panel?

Thank you. As it turns out, this is a bug of Intel's Windows 10 video drivers. We've run into the same issue than this gentleman: http://stackoverflow.com/questions/34676697/embedding-manifest-into-the-application-binary-breaks-opengl-2

Thank you for the data. The issue will be fixed in the next AIDA64 beta update due in a few days from now. I'll post a message into this topic once the new beta is available for download.

Do you know approx. at what RPM your fans are spinning while your computer is not under heavy load (e.g. when you're using AIDA64 or browsing the web), and how many fans are spinning?

We've checked all those, but without a luck so far The registers you've mentioned are all zero

That is due to a bug in the latest Saitek X52Pro drivers (Version 7.0.47.1). When you install those drivers, the driver installer fails to copy the 32-bit DirectOutput.dll file into the folder under \Program Files (x86)\Saitek. We've checked it, and it seems both the latest driver (7.0.47.1) and also the previous driver (7.0.40.9) have this bug. The last driver that works fine is 7.0.39.0. You can either revert back to the last working v7.0.39.0 driver, or I can send you the 32-bit DirectOutput.dll file that we extracted from 7.0.39.0 driver. But I cannot guarantee the DLL from the old driver would work properly with the newer driver as well.

Thank you for letting us and other users know about the resolution.

I'm afraid we don't have any contact at HP

Thank you. Are you using a 32-bit Windows or 64-bit?

We've never noticed that as an issue on our test systems so far. Please let me know what video card do you have, which Windows version do you have installed, and the video driver version you've got installed.

I'm afraid we don't have any idea about what could prevent your system from providing drive activity statistics to AIDA64 We've done test runs on various Win8.1 systems, but on every system we could monitor drive activity without any quirks. There must be a special configuration setting or a background service/process that causes the issues on your system, but I have no idea what could it be.

AIDA64 doesn't support setting the number of decimal digits per sensor item, and we don't have plans to enable that. Regards, Fiery

Thank you for your feedback. The mapping of misc temperatures on your board are: Temperature #1 = T_Sensor1 Temperature #2 = EXT_Sensor1 Temperature #3 = EXT_Sensor2 Temperature #4 = EXT_Sensor3 I'm not sure why they show bogus values sometimes. The only issue I can think of that may cause that is a background process or service that concurrently monitors sensors. The most infamous of those is Asus AI Suite that is still not synchronized with 3rd party monitoring software like AIDA64, HWMonitor, SpeedFan, CoreTemp, HWiNFO, SIV, etc. While those listed software are all synchronized with each other to avoid such collisions due to concurrent sensor access. As for the extension fans, I'm afraid those are only available via EC banked access. And switching EC banks is dangerous, since when a collision (with e.g. AI Suite) occurs during an EC bank switch operation, the system could lock up or the CPU might get overvolted

Yes, that's to prevent you from losing a column If you set one to zero width, it will revert back to the default width the next time you open the System Stability Test.

It is not an option, but a small new feature. AIDA64 will now remember the column widths of the statistics list on the Statistics page of the System Stability Test.