I think you are getting your “64s” & “32s” a bit out of context and confusing yourself as a result. When you are dealing with PCM, the DAC display will show a maximum of 768kHz PCM and the number of bits should be handled transparently by the audio application software. Anything above that will be a DSD sampling rate and is technically 1bit (as opposed to the 16, 24 or 32bit structure of PCM) but at a rather higher sampling rate, starting at 2.82MHz.
“64” & “32” relate, in all likelihood, to the computer operating system and it is what it is i.e. nothing directly to do with audio.
Don't think I am confused at all, PCM are usually 16, 24 or 32 bit, and I am fully aware DSD are 1 bit with super high rate. So it's not me who is getting this out of context, the link clear states:
"converts from 1-bit DSD to 64-bit PCM at 1/8th of the DSD sample rate. The total amount of data from this conversion grows by 8x, so the process is effectively lossless / perfect. Once you have PCM, it will be 64bit @ 352.8 kHz for DSD, and 64bit @ 705.6 kHz for 2×DSD. "
So no, that does not answer my question at all, and be honest I think it's you who are taking it out of context and randomly added operating system bits into this.
My best guess now is it is indeed 64 bit 352.8 khz PCM at first, then they perform another conversion at later stage so DAC can do it's job. Now benefit of two step conversion instead of one is beyond me.
What still puzzle me still is in the link the second down sample step implied as optional, but I can't find any DAC can do 64 bit 352.8 khz PCM on market.