5nm is only a marketing term and has no relation to any actual phyiscal feature

> Or is there any objective metrics to determine which one is better, other than MTr/mm2?

There are many, many objective metrics. Many are interdependent, many are tunable or selectable within certain ranges, many are in opposition, most are highly secretive, and all of them influence he type of device being made, the circuit design, logic, in different ways.

Even the experts with all the data for a particular technology and with a particular device in mind can't necessarily tell at a glance whether a change would be worthwhile. And even the meaning of worthwhile varies for the end-user of the device depending where their workload is. Make the middle-performing transistor slightly faster and leakier? Great you can close timing at a few ps faster but the device uses more power and is hotter, good for some customers bad for others.

Even simple density is not that simple in the fine details. Density can be in opposition to leakage, switching speeds, wire RC and propagation delay, etc.

There isn't an easy way to compare and there never will be.

> The term "5 nm" is simply a commercial name for a generation of a certain size and its technology, and does not represent any geometry of the transistor.

And I learnt something new: it seems like MTr/mm² is the metric in use now. I wonder if marketing will find a way to co-opt that?

It's the same as software engineers who don't know what SIMD is or think the CPU in their laptop or smartphone works like a MIPS they studied in college. When something like this is only one click away: https://en.wikichip.org/wiki/intel/microarchitectures/sunny_...