Hillbilly said:
The AD Sharc DSP used in UAD-2s is a 32 bit processor. My guess is before 64 bit plugs, hardware will need to change........
The "bit" issue confuses people, especially non-developers. Really, it's not that complex, it's just how large numbers can be. 32-bit hardware can process 64-bit numbers, or 128-bit numbers, 64-bit hardware can generate 32-bit, or 1-bit numbers etc etc. It's what the 64-bit numbers *apply* to, where the complexity arises.
The driver is just there to shuffle data from the computer to the card and back again. You can compile the drivers for 64-bit operation, but you shouldn't necessarily need to change how the driver interfaces with the card. While I have no idea of the UAD2's internals or drivers, the thought that you'd need 64-bit UAD hardware to use a 64-bit driver doesn't make sense. For example, my audio interface, a cheap little thing bought five years ago, was designed *way* before 64-bit was on the cards, and yet I now have 64-bit drivers for it. That didn't require new hardware, so the same probably applies to the UAD.
(This is not the same as plugins - the main issue in this case is that the plugin needs to operate *within* the host, which means for technical reasons it needs to be compiled the same way as the host - this is why 32-bit plugins won't work in a 64-bit host.)
So my (somewhat informed) guess is that the delay is most likely because UA probably want to do it *properly*, which means updating their frameworks, plugin code across multiple plugins, and various drivers, across multiple plugin formats and platforms, and thoroughly testing and debugging them - this is *not* by any stretch a minor task - even more so if there are technical hurdles or the frameworks need to be junked and redeveloped from scratch.
Non-developers are kinda seeing the number "64" everywhere without understanding that the 64 represents different things in different contexts. And that something compiled as 64-bit code doesn't mean it can't talk to other things using smaller numbers - it's not like everything has to be a 64-bit number - working with different data sizes and representations is something fundamental to computer programming. Address space and code compilation is different to user data and mathmatical precision.