Re: What about auto-updates?
>>Previous SD Card specs have lacked wear levelling and other niceties found in enterprise SSDs
Neither does the NVMe spec nor usb mass storage. There is no enterprise "ssd spec"
The sd card spec is not a flash specification, it is an interface specification - dealing with electrical, physical and interoperability specs. It also isn't remotely trying to address the enterprise market.
Wear levelling, gc, sector/cluster/block mapping is the flash controller's job. It has nothing to do with the bus spec. The sd card manufacturer is free to choose flash technology, nor, nand, slc,mlc etc.
and will need to decide the appropriate wear levelling/endurance/spare block and thus cost/profit balance they wish to target.
This is what is different in the experiences reported by others about their cards. It isn't completely about brands, the older cards were typically better made, more expensive but do last longer as a result. Nowadays it's fewer ecc bits, fewer spare blocks, etc.
I think the business model has changed where consumers expect less (as replacement cycles are far shorter and accepted) and brand marketing is more cost effective (to catch those replacement purchases).
Biting the hand that feeds IT
