Very unlikely. 64-bit is mostly an advantage if you have to use a lot of memory or are doing processing on actual 64-bit integers.

For the kind of software like Opera (and Firefox and Chrome), the additional registers in 64-bit mode help a little, but that tends to be offset by most pointers now being 64-bits and the L1/L2 caches hence being less effective.

On my computer most memory hungry apps are browsers.

And by the way I use opera to open a lot of tabs and it handles those very well (compared with other browsers)

Right, but 64-bit computing isn't a magic bullet that speeds everything up. As gcp stated, there are both benefits and drawbacks.

Although it does break that annoying 4,294,967,296 open tabs limit!

This is nowhere near realistic. Opera's performance degrades linearly after the first 10000 tabs.

Do you have the 16bit build?

There are more registers in x64, so there may be some speed gain. Or a loss of speed due to larger pointer size and more frequent cache misses.

Registers, more instructions guaranteed (like SSE2) etc. It's up to the compiler to take advantage of that tough.

Did you just repeat exactly what I said?

There is an official 64-bit Firefox nightly built for Windows.

I tried it and to be honest I can't tell the difference.

Also bear in mind that it needs 64-bit Flash and Java plugins which are notoriously problematic.

Opera 12 uses a separate process for plugins like Flash and I believe it can actually use the 32-bit Flash within the external process while still displaying the content within the 64-bit browser.

Java and Flash are notoriously problematic (not to mention insecure) on 32-bits too.

64-bit is not really useful for browsers with multi-process architecture like Chrome. It is very unlikely that a single tab will need 64-bit address space.

Except once you have enought tabs Chrome starts putting multiple tabs in a single process. The cutoff seems to be somewhere in the low double digits. And at that point, you can in fact start running out of address space (not to be confused with running out of memory) without too much trouble, especially because system libraries on 64-bit OSes have a tendency to be mmap-happy, since they think they have lots of address space to work with.

64 bit address space significantly improves the effectiveness of ASLR, 64 bit tagged pointers can store way more data (7 byte strings, dunno if anyone actually does that, though).

I generally find Chrome likes to start falling over around the 3.5GB mark (as measured by chrome://memory) - larger images stop loading, pages stop rendering properly, and eventually it can end up crashing. Address space fragmentation I'm guessing.

This isn't helped much by Chrome's aggregate memory use being about 10x higher than other browsers. Takes a good 2-300 tabs in Opera to break 4GB, takes more like 20-30 in Chrome.