Carrier freezout actually makes this non-workable -- there's a limit to how cold you can make CMOS devices before they stop functioning. To say nothing of the specific heat of liquid helium, which is miniscule compared to LN2
Yeah, the BladeRF uses an AD9361[1] as the transceiver, which features a pair of ADCs per channel, of which there are two (so, a total of 4 ADCs and 4 DACs in the package). It's got a bunch of other convenient features like built-in PLLs both LO and baseband.
So long as all frequencies you're trying to represent are below the Nyquist frequency, there's still only one band-limited reconstruction of the samples. So yes, it's still possible to reproduce everything accurately so long as all of the individual frequencies (no matter their relationships) fall below 1/2 the sample rate.
I paid $12 on Amazon for 6 Travalo lightning cables on Amazon and I couldn't be happier with them. Lightning interface is legit (no notifications about unsupported lightning connector) and they're more robust than the official Apple cables.
FFTW is one of the seminal examples of the importance of cache awareness, if I remember my history correctly. As I understand it, the reason the fftw authors (Matteo Frigo and Steven G. Johnson) won the Wilkinson prize was essentially because fftw exemplified the importance of cache-aware algorithms, as opposed to blindly optimizing number of operations.
I don't know, but I would guess that intel's implementation post-dates fftw?
Same goes for me when doing CAD and having to constantly convert between standard & metric -- 254, 508, etc show up in numbers a lot and it gives you a nice hint that it's a "clean" value in the other system. Super useful for reading IC footprints off datasheets that only use one of the two systems
Depending on your workload, you may get far better throughput not worrying about distributing the work and data