Which is why I want to think about (or compare against) an implementation where a decimal64 would contain { significand; exponent; } instead of { bid64; }
When you first asked me about decimal 2 or so years ago I toyed with an implementation that skipped the combination field: https://github.com/mborland/decimal/blob/main/include/boost/decimal/decimal3.... It sacrificed some of the range of the exponent to make that happen.
Regarding comparison to Intel's implementation (bid based) which is used in both GCC's decimal64 and also in Bloomberg, that's especially important because it used some impressively large tables to get performance (which also bloats binaries)
So far we don't rely on giant tables. Chris added an STM board QEMU to our CI that checks our ROM usage among other things. I'll work on some benchmarks and see if it's worth building out the above non-IEEE754 decimal32. Matt