v3.55 - Nvidia GTX280 fan speed half the RPM's

[chinobino]

Hi, I have noticed that the fan speed on my GTX280's has been reported as half speed (after version 3.36), it seems that you may have used a different divider?

I have included a report in HTML format and the debug dump if that helps .

If you need any more info just let me know.

v3.55 Report & Debug

And here is the report & debug from v3.36 with the correct fan speed;

v3.36 Report & Debug

[Martin]

Thanks for the data. I'll analyse it and let you know as soon as I have a solution.

[chinobino]

Thanks a lot!

I appreciate all your hard work.

[Martin]

Please try this build: http://www.hwinfo.com/beta/hw32_355_817.zip and let me know if the GPU fan speed is reported correctly now.

[chinobino]

The new build (v3.55-817) is now reporting the correct fan speed - Excellent, thankyou!

There is possibly one other small 'bug' that I noticed.

When watching the CPU 'Clock' on the System Summary page, my QX9650's four CPU cores are showing as changing frequency independently, which is not physically possible on a Core 2 Duo according to the Intel white papers.

My E8700 also does this (although not as often, having only 2 cores) as the load on them varies.

In previous versions I noticed that you had locked all four cores so that they moved together, I was wondering if the change was intentional or a C2D bug perhaps from the new i7's?

It's not a major issue - I was just curious as to why it suddenly changed.

[Martin]

I made changes to the clock calculation in v3.53-3.55 mostly to reduce DPC latency issues caused by previous versions and changed the scan method a bit (instead of parallel on all cores, it's now serial), so the clock monitoring might appear different.
However, it's hard to say what is correct without really seeing how it behaves. I also don't think it's possible to catch the true situation in all cores realtime when the P-States change very frequently.

chinobino Wrote:The new build (v3.55-817) is now reporting the correct fan speed - Excellent, thankyou!

There is possibly one other small 'bug' that I noticed.

When watching the CPU 'Clock' on the System Summary page, my QX9650's four CPU cores are showing as changing frequency independently, which is not physically possible on a Core 2 Duo according to the Intel white papers.

My E8700 also does this (although not as often, having only 2 cores) as the load on them varies.

In previous versions I noticed that you had locked all four cores so that they moved together, I was wondering if the change was intentional or a C2D bug perhaps from the new i7's?

It's not a major issue - I was just curious as to why it suddenly changed.

[chinobino]

I see - that makes sense.

Those P-states change are very fast, like you say, so it's really a non-issue.

Now that the sensors are all spot on I'm off to re-configure my LCD display, thanks Martin.