I got a question about IPv6 subnetting.
A question similar to the following is often included in exams in Germany:

“From the internet provider, the XY Agency receives a letter stating that the public address space 2001:db8:abc::/48 will be allocated to it.
From this range, create four equally large subnets. Use the entire address space for this purpose. Please indicate the network addresses of the four subnets.”

This is the suggested solution for the 4 subnets.

1. 2001:db8:abc::/50
2. 2001:db8:abc:4000::/50
3. 2001:db8:abc:8000::/50
4. 2001:db8:abc:C000::/50

This is a valid solution, but is it really necessary to do it that way? In the end, we have /64 nets in IPv6 anyway. Otherwise things like SLAAC won’t work.
So in my opinion, this is also a valid solution for this task:

1. 2001:db8:abc:1::/64
2. 2001:db8:abc:2::/64
3. 2001:db8:abc:3::/64
4. 2001:db8:abc:4::/64

Those are 4 IPv6 Subnets and all have equal size.
What do you think? Do I miss something?