New QUAD Core Laptops Refresh Slower

[tmf171]

Has anyone seen an issue with newer quad core computers (that are supposed to be faster) actually run their applications at a slower rate? Refresh to the screen is slower and data acquisition appears to be slower. This is also maybe a comparison of Windows 10 to Windows 7. Specifically we can use a brand new gaming laptop vs a 5 year old workstation. Any insight on this. They both use NVIDIA discrete graphics cards.

[billko]

---

@tmf171 wrote:

Has anyone seen an issue with newer quad core computers (that are supposed to be faster) actually run their applications at a slower rate? Refresh to the screen is slower and data acquisition appears to be slower. This is also maybe a comparison of Windows 10 to Windows 7. Specifically we can use a brand new gaming laptop vs a 5 year old workstation. Any insight on this. They both use NVIDIA discrete graphics cards.

---

You've rambled all over the place and haven't even mentioned what it does to LabVIEW, yet.  So no, no insights yet.

[tmf171]

LabVIEW 2016 Data Acquisition and Display on multiple waveform graphs and intensity graphs. (1 waveform graphs and 3 intensity graphs). We use a queue at acquisition and process the waveforms in the analysis loop. 5 years ago this worked fast on Lenovo Works slower on latest gen computers.

Lenovo W530: https://www.amazon.com/Lenovo-ThinkPad-W530-243852U-Notebook/dp/B00BFFB08A

MSI Gaming Laptop MSI GV62 8RD-034:https://www.amazon.com/MSI-GV62-8RD-200-Performance-Steelseries/dp/B07BWF4H3W/ref=sr_1_1_sspa?s=elec...

Its just strange that the display updates happen less often (we check how many items are in the queue and only refresh the screen if the last item in the queue is being dequeued.) I'm not sure if anyone else has seen slower behavior and there is some setting that will be a eureka moment.

[RavensFan]

Why would a slower screen refresh, assuming that it is slower, cause the data acquisition process to be slower?

[billko]

I wonder how much shovelware is running in the background?

[RTSLVU]

Well LabVIEW is single threaded (I believe) so that  i7-8750H 2.2 - 4.7GHz is not really faster than the i7-3740QM 2.7 - 3.7GHz when running a single threaded application

[altenbach]

LabVIEW is multithreaded and has been for a long time (even before multicore became a thing). I have no problem saturating dozens of cores with a single LabVIEW application.

Yes, there is only a single UI thread, so if you have dozens of graphs with millions of points each, you are saturating the UI the thread.

Still, this does not explain the difference.

New computers are often slow due to bloatware (as has been mentioned), as well as significant background activity (indexing, updates, etc.), so give it a day or two to get comfortable.

[bseguin]

All that depend on your application.

If you are refreshing your data on the GUI on the same thread that you acquire them, yes it can be slower depending of the 2D acceleration capability of your computer. a gaming computer will be good at 3D but not necessarily at 2D. LabVIEW use 2D for his display refresh.

I suggest you to revise your application architecture.

Create multiple loop/sub VI that run at the desired priority.

The display refresh for user is pointless at a speed more than 5 times per second except when it is for video purpose. Your brain will not retain values that changes more than 5 time per second I can guaranty you that.

So 1 sub VI subroutine called dynamically. data exchanged trough get set or any other way that will prevent race condition. (this is your core data acquisition). Other thread that run at lower priority like save data to file (larger chunk is faster for mechanical disk *I wont go in details here*) And the lowest priority is your refresh rate of the data showed to the user.

You better show a graph with statistical analysis than showing all the point real time and let your brain evaluate approximately what it means.

And next time you chose hardware... Gaming PC does not mean faster.

Benoit

[billko]

You did a great job with the explanation.  🙂

[bseguin]

Kudo it.

Benoit