The group speculated (and Phison later confirmed) that this early firmware was the real cause of the instability. Unlike the finalized firmware shipping on retail drives, the engineering preview caused crashes under stress. To prove the point, Phison ran the same stress tests (100GB to 1TB sustained writes) on consumer-available SSD models and reported no failures or crashes.
That turned out to be because of some prerelease ssds with nonfinalized firmware that shouldn’t have been in the wild.
https://www.techspot.com/news/109370-windows-11-cleared-all-charges-killing-ssds-real.html