Once upon a time, in order to get sound on your Windows computer you’d have to buy and install a sound card of which the most popular were the Sound Blaster line of cards from Creative Labs. Back in those days, The CPUs inside your computer had more than enough to do running Windows while offloading things like sound and video to secondary cards. I remember trying to play an MP3 file on our computer with a Pentium 120MHz CPU inside and WinAMP would max out the CPU trying to play the audio leaving your computer little ability to do anything else.
Over time though, CPUs became more powerful and we saw things like dialup modems, audio processing and video processing being run in software to cut down on hardware costs. These days, we have heaps of things built into motherboards including multiple network interfaces and USB controllers in addition to 5.1 or 7.1 surround sound audio. However, sometimes all of this stuff in software can interfere with each other and, in the case of audio, you might hear glitches, stuttering or small gaps despite having a more than capable of CPU.
Basically, there are two things to try in this situation.
Firstly, you should consider checking for updated drivers for not only your integrated audio devices but also your integrated network interface devices. As was the case with softmodems, early drivers could be unreliable while later revisions ironed out instabilities particularly as the bug reports flooded in from unsuspecting early adopters. In particular, Realtek network chipsets with older drivers seem to cause problems with audio.
The other option would be to invest in a proper hardware solution for either your audio or network interface (the more cost effective solution might be a sound card). You can still pickup Sound Blaster cards (such as the X-Fi which can be had for well under $100 for a PCIe slot). It’s certainly a far simpler option particularly if you use two or more of the built in network interfaces.
One other thing I am looking at is the default sound sampling rate and bit depth. Since a lot of audio is sampled at 44100Hz you may hear gaps or glitches if you have your sampling rate set at another bit rate that does not evenly divide into it. So outputting 44100Hz audio at 48000Hz (which results in 0.91875) could potentially be problematic. I need to conduct some more testing before I determine this to be a credible course of action though.