What are the best GaN Amplifiers available today?

@markmuse Thanks for the details. 

@greg_f Thanks for the correction on Aavik. Are Pascal modules a GaN design or no? I didn’t think they were. They sound great regardless. 

@blisshifi No problems, the Pascal modules (UMAC) don't use GaN technology, one of their features that makes them different to Hypex, Pufiri, IcePower and the rest of the classD designs is that they use a sinusoidal waveform instead of the more widely used triangular waveform to produce the PWM so theoretically the output filter design is easier and should produce less RF noise. That's the theory.

You say they sound great, I have never heard any Aavik amplifier in real life, it will be very useful to know how they compare to a good AB integrated like the Gryphon, VItus, Pass Labs, T+A, etc. Any pointers?

I recently did a week's test with the Orchard Audio Starkrimson amp with upgraded caps and power supply ($3500 retail).  

I really liked it !  The headline is speed and microdynamics as if to reveal notes and timing in the music that were probably hidden with previous amplifiers.  

Tonally neutral and overtly smooth, there is no warmth yet it's clean enough not to matter.  Midrange presence and transparency are a bit lacking compared to other amplifiers in its class but these are minor issues.  

I totally competent powerful amp in a very friendly package.  

I should also say that I am looking for very high performing GaN amplifiers to perform against some of the best / most high-end units out there, like the T+A A 3000 HV amplifier (which I own along with the T+A PS 3000 HV power supply), Audionet amps as @pennfootball71 mentions, or other Class D leaders like Aavik or Bel Canto Black. High bandwidth, control, transparency, and resolution are all important, but just as important is a completely organic, natural presentation with the proper use of harmonics and lifelike presence/immediacy and dynamics.

@greg_f I unfortunately haven't AB/ed Aavik against other amplifiers with the same speakers in the same room to be able to comment. I would say that Aavik 580 series is a cleaner sound, but still incredibly musical, than my reference T+A HV stack. The T+A has richer harmonics and a tad less control in the lows (Aavik is known for its insane damping factor and that is a large contributor to making Borresen speakers sing the way they do). My T+A is more natural, a bit less clinical, but at the expense of not being as lightning fast as the Aavik. Keep in mind that this was the same set of speakers, but in different rooms (I brought mine back to Next Level Hifi to compare) and different cables (Synergistic Research Galileo SX in my room vs their Ansuz DTC-2 loom). Gryphon has richer harmonics still over T+A, and I prefer T+A, which is why I became a dealer for them. Pass generally has less rich harmonics than Gryphon or T+A but has a warmer character to them. Vitus, I've only heard once, with Estelon Extreme at AXPONA this year, and though I've heard a few Estelon, I'm not familiar enough with what the amp was doing vs what the speakers were doing. In that room, it was a very burly, natural sound, but not as immediate and transparent as Aavik / Borresen combos I've heard time and time again.

@blisshifi Thank you for taking the time to form such a detailed reply, it is exactly what I was looking for. I have been using Class D amplifiers for almost 20 years and the technology keeps on improving, IMO they are the future of amplification. It seems like the Achilles heel of the Class D amplifiers (the top end) is becoming less of an issue with the latest generation. Of course regardless of the topology used, the design of the amplifier is the most critical factor. My speakers (Vienna acoustics Mahler) require the amp to have good grip of the bass units and be fast otherwise they can sound lazy, so they work well with class D amplifiers. 

Even though GaN transistors were not designed for the HiFi industry, they could suit class D topology since they are switching much faster (decreasing dead time, easier filter design, etc) but again it all depends in the implementation of the amplifier, you can't say because an amplifier uses GaN transistors is better than one that uses MOSFETs or bipolars, we need to talk about specific amplifiers.