Better drivers is indeed one potential improvement in the performance offered by a good sound card, also more streamlined functionality. Onboard sound chips will have processing capabilities for use in gaming, movie/DVD play and 5.1/7.1 that are not utilized by a plain stereo music signal. But it's pretty likely that they still exist in the signal path, using resources, IRQs etc, even though their instantaneous function is "do nothing to this signal", then the next one "do nothing to this signal"...etc

An external audio interface will only have the functions necessary to fulfil its purpose and is likely to need less dialog with the system to achieve this.

This also points up another way that changing to an external interface may resolve an issue - it necessarily changes the allocation of resources such as IRQs, which result can be cumbersome to achieve otherwise, especially on a laptop. If a resource conflict such as IRQs (which interrupt the CPU - being an Interrupt ReQuest) is contributing to or causing the problem, using the alternative audio device might be enough to resolve the issue.

The external audio card does not process any audio.

I'm not sure what you mean by this. The purpose of an external audio device is for it to be handling the audio duties, digital to analog conversion. Are you thinking of it as a MIDI interface?

Another question: how would using an external audio processor work?  If the laptop has to receive the MIDI signal from the piano in order for pianoteq to do its magic, how does using something like focusrite take load off of the processor?  Isn't the focusrite only getting the audio signal once it has left the laptop?  Confused about this....

The external audio card does not process any audio. I'm not sure about it, but I believe the better performance is due to optimized drivers. By the way, I have used ASIO4all and Windows audio exclusive mode with good results.
I think your processor should be capable of handling PT since I used to run it (v5) without glitches on a Pentium 2020m.
If you want to be sure your PC is able to run pianoteq, I suggest you to run it from a GNU/Linux distro like Ubuntu Studio (you don't need to do anything with your data, it can run from a pendrive). On a old laptop, I used to have better performances under Ubuntu Studio than under Windows. The procedure is quite easy, you will not need another activation slot, since modartt detects that the computer is the same.

When I look at the audio load panel as I replay a piece, there are occasional red spikes --- they don't seem to be associated with pops during playback,

Bolding added by me. This seems to add to the likelihood that digital noise injected into the audio is being misidentified as pops from latency dropouts, as well as being the result of a ground loop. This is also suggested by the quote below:

Loading an application produces bursts of digital noise, as data is squirted from the drive into RAM, screen is loaded etc. There may also be some effects present from latency, buffer underrun etc, but it's possible they are only say 20% of what is being heard, and digital noise is 80% of it.

I guess what I'm wondering is why a quadcore 7th gen i5 CPU isn't good enough to handle things

An i5 7300U isn't quad core, it's a fairly modest 2.6 GHz 2 core CPU with hyperthreading. Even allowing for hyperthreading capability, a 2 core CPU will be kept busy just running Windows, especially Windows 10 Pro, and a time critical application like PianoteQ has to compete for execution time.

The question I can't figure out the answer to, is, what background process is holding things up?

Can you not use the Windows Task Manager to monitor which processes are running? Even possible malicious malware lurking in the attic of your laptop should be visible there. If something else than PT is taking up a lot of cpu -power, it will show.

Try disabling WiFi. I have a laptop (old i7 4710mq) in which there are pops and cracks if WiFi is active. In the old days it had to do with the IRQs...

Thanks, but changing to airplane mode makes nary a difference. The question I can't figure out the answer to, is, what background process is holding things up?  If I click on internet explorer (in other words to start internet explorer, even when wifi is not on), I get a mega series of pops, indicating that the pops are due to processing of data.  But I can't figure out what random background driver/program is creating the pops.  I can say that if I play a regular VST (can I say Kontact free hammersmith piano?) that I get no pops of any kind.  That implies to me that it is processor hiccups, not disk drive access, which causes the issue. 

Latencymon gave me tons of gobbledeegook.

Another question: how would using an external audio processor work?  If the laptop has to receive the MIDI signal from the piano in order for pianoteq to do its magic, how does using something like focusrite take load off of the processor?  Isn't the focusrite only getting the audio signal once it has left the laptop?  Confused about this....

Try disabling WiFi. I have a laptop (old i7 4710mq) in which there are pops and cracks if WiFi is active. In the old days it had to do with the IRQs...

Same experience here - I run Pianoteq on an old MacBook Air (i5-4250U, 1.30GHz), and that works perfectly fine (including a sustain pedal down stress test with keyboard wide arpeggios and lots of bass) as long as wifi is not connected. Check also that your computer isn't occupied with some other background process.

If you can increase the pianoteq latency at all, that will help as well.  The idea of an external sound card is a very good one, but there's quite a quality range in the market.  You want something that has enough processing power (built-in chipsets) and good enough software that it actually offloads most of the work from the CPU and motherboard, most $20-40USD ones won't do that (but a few do), while most $100-200 ones almost certainly will.  The external soundcards should be something like these: https://www.bestproducts.com/tech/gadge...pc-or-mac/

Thanks, I'm looking at those....
I guess what I'm wondering is why a quadcore 7th gen i5 CPU isn't good enough to handle things.... Now, if the issues is that Realtek drivers are notoriously latent and/or eat up CPU power, is it possible to just sidestep those and then rely on the CPU and ASIO4ALL to do the work that Realtek is getting in the way of?  And if so, would a cheap sound solution work (again, because even if it isn't powerful, it's better than Realtek?)  for example the $30 soundcards?

I definitely want to solve this, but I don't want to spend that much, as I know that I actually am planning on getting another laptop shortly, so if merely switching to a different brand is all I need to make things work, I'd rather not accumulate more connections and gear.....

OH, one other question. WHICH $20-40 external sound cars are likely good enough?

Personally I wouldn't try to run PianoteQ or any serious music applications on the onboard Realtek audio in a laptop, you may well find that both this and the ground loop problem could be cured by the use of a respectable USB audio module incorporating MIDI. Do you know anyone who has one you might be able to borrow and try?

If the Realtek Audio remains as a WDM (Windows Driver Model) device, that can be due to onboard audio chipsets such as this utilizing software or partially software codecs, with the sound rendering being done by the CPU, rather than in dedicated hardware, which would take that load off the CPU.

Here's what I don't understand (and a lot of what I wrote above is gobbledeegook for me, so I really mean that I don't understand!).....  Pianoteq says it's using the Asio 4 all Driver.  It's only when I click on that box that Asio4all brings up the Realtek audio.  (1)  Is there a way to run it through my laptop and bypass Realtek as the WDM, if that is the problem?  If so, how? (2) The laptop is not a super powered one but SHOULD be way more than adequate for pianoteq, at least based on the specs that Pianoteq publishes and the FAQ.  Is it normal for me to have to reduce the load when it's so powerful, and gives me powerteq indexes in the 95-120 range?  Is the spiking likely indicative of some kind of an undiagnosed bottleneck?  Where would I look for that?  (3)  Is this something specific to Dell chipsets?  If so, what kind of inexpensive windows laptop would do better? (4) What are examples of USB audio modules?  DO you mean something like the Roland Um-One (That didn't work)?  Links would be great, especially if they are for Amazon (generally easy to return if needed) and on the less expensive side.....

I know that's a lot of questions.  Any help is much appreciated