If the purpose is sterilization, you have to choose a radiation that kills the germs without killing the CMOS. And that radiation has to be cheaply provided. Generally the thinner the memory layer the less likely the radiation would deposit damaging energy.
@resistion: You have made an excellent point!! If there is a way to control the radiation such that the radiation serves its purpose without damaging the semiconductor while keeping the cost lower than what it would be with the CBRAMs, that should make the case for it.
Are the CBRAMs equivalent to FLASH functionally? Currently what kind of memory chips being used in space equipments, for which the radiation threats are similar? I have heard of radiation hardened FPGAs being used but do not have much idea about the memory.
I remember I've read article on Scientific American IBM at that time was investigated MEMS memory - essentially atomic-size abacus. I never heard what happend to that technology, probably dead-ended. Sad, if they've kept the development, it could be highly tolerant to radiation memory.
My experience in the medical device is that the components command good premium compared to the consumer market. So, this can be done. I know for example in MEMS, you can sell a unit of XL for $60 compared with $0.60 you can get in the consumer market.
Another challenging issue is the need for medical devices to tolerate cremation (or accidents) without releasing toxic materials even after the patient dies. In the instance of plutonium powered heart pacemakers this was a serious issue. While the dead patient wouldn't care, the release of plutonium into the environment during cremation was a serious concern. Physical shielding accounted for a significant portion of the bulk of the device.