Upgrading to LV 2017 from LV 2015 causes a dramatic slow down in application execution speed

[blessedk]

I recently upgraded to LV 2017 17.0f2 (64-bit) from LV 2015 15.0.1f10 (64-bit). Noticed a remarkable slow down in execution speed of an application I originally developed in LV2015. Any Ideas why? A bit disappointed as I expected newer versions to better/faster or atleast as good as the previous ones. To put things in perspective, a loop which used to run every 140 ms now takes 280 ms!!

[Bob_Schor]

---

@blessedk wrote:

To put things in perspective, a loop which used to run every 140 ms now takes 280 ms!!

---

That's astounding!  Can you isolate the loop and let us take a look at it?  A 5-10% difference I could (maybe) understand, a factor of 2?  Not so much.  

Bob Schor

[MichaelBalzer]

I can't say I saw a massive performance degradation with 2017. Can you isolate which function(s) in the loop are slower after the update?

[blessedk]

Agreed. Strange to me too. Unfortunately this is one of those loops I definitely can’t share 🙂 and it’s quite involving so I am painstakingly looking at every vi, node, method etc within  that loop. Will let you know when I find it. In the meantime I was hoping anyone knows something about possible LabVIEW, windows settings ( something specific with the LV 2017 64-bit? Etc etc.

[BertMcMahan]

Are you dynamically calling VI's from a path? If so the editor could be trying to recompile them each time they're opened. (That's not a 2017 thing, that's a "new version" thing).

[blessedk]

OK. At which point Would it then stop re-compiling every time then? Would a mass compile nake a difference?

[BertMcMahan]

I'm not the best person to answer that, but I think they'd have to recompile each time you call them until you save them. A mass compile should update them to the current version and you should be OK.

(Just a guess on my part... hopefully if I'm wrong someone will correct me.)

[Michael_Munroe]

You don't mention any timing VIs, but the symptom of a perfect doubling of the loop rate sounds like a metronome function.  If you place a metronome set to 140ms and your code runs 130ms, your loop timing is 140ms.  If your code now runs 10% slower at 143ms, then your loop rate jumps to 280.

[blessedk]

I have confirmed that this is not a LabVIEW issue. It’s a windows system issue. So for those out there who say  “That’s why I don’t upgrade” etc etc. There is nothing wrong with updating. LabVIEW 2017 works fine 🙂

Thanks guys

[Intaris]

OK, so that begs the question.... what actually caused your differences in timing? Which kind of "windows system issue" was at fault?