The DPRK apparently told China before its 2006 test that they were aiming for a yield of about 4 KT but they only acheived a yield around ½ KT. This implies that there was a significant problem with that design. ( Their announcement of this test seems to indicate that they are acknowledging that.) If this time they had gone with a test of a 20 KT design, as the Russians are apparently saying, it means that they were changing designs to ensure they got a big bang and avoided the publicity problems they went through after the 2006 test: of having their first test a failure.

If its yield really is closer to 4 KT, it is, of course, still possible that it was a 20 KT design that failed (which seems rather hard to believe since 20 KT is the “easiest” yield to get—that’s why it was used in the first plutonium bomb during WWII). But its also possible they were once again aiming for a 4 KT yield. That would mean they kept the basics of their old design and just corrected whatever the problem was. This would show a certain level of sophistication of project management that is not obvious in their missile development program. (That assessment, of course, depends in part on what the 2006 Tae’podong II looked like.)

If they had gone with the “fail safe” WWII design, it would probably mean it was too heavy to mount on a missile. They would be making a political bomb that would undoubtedly use a lot of high explosive to ensure it got a good compression of the plutonium pit. The 4 KT bomb, however, might very well fit on a DPRK missile. If they have stayed with this design, it probably indicates that weaponizing it is even more important than ensuring a successful test.