So I was looking on the TI website and found this chip.
http://www.ti.com/product/RF430CL331H/description
It is in a 3mmx3mmx1mm package add a SPI flash chip and you could in theory make a flat device that is capable of being implanted in the arm that can larger amounts of data. if this was adopted into a ASIC type device that had the core of both the flash and CPU on the same die the size could be reduced to an extremely small package. any thoughts on this concept?
Lol nevermind. 
I read the datasheet and the FAQ pdf and I don’t know that it’s a good application for implantable devices. It seems to be a prepheral rather than host device so it needs a microcontroller to do most of the memory / storage work if you want to use flash etc, and I think that’s going to require external power.
The bottom line here is that, if you could store a ton of data… the bandwidth over NFC is so shitty, it would take forever to do any kind of file operations. At a certain point, practicality plays a roll in whether 5 people would want to buy this device or 50,000… and when we’re playing in such a niche field as biohacking implants, making things that most of your tiny target market would want to buy is the top priority when allocating R&D resources.
Any sort of file transfer device generates heat, thus I don’t recommend using an implant as a means for constant reading and writing.
My recommendation for file transfers if you wanted to transfer information to another location would be to upload the data to a host and use an NFC implant as a means to store the URI along with any information required to read or decrypt the data.
Having the Data on a server defeats the purpose of having secured data on your person. Servers can be hacked a device that can only be accessed by being 1-2cm from it is by nature more secure.
I certainly understand the scale of manufacture and the market size considerations. I read a White paper regarding a possible data rate of 4Mbits/s but this was using a higher frequency.
https://ieeexplore.ieee.org/document/4519345
If that becomes a standard then this may be possible. As for heat concerns if the surface area was a modest size then you could in theory use the body as a heat sink. Naturally the device cannot emit more than a small amount of heat by design.
This option has a MSP430 MCU built in and is RF powered. I know GB of storage are not realistic due to band width issues but a few MB. Also there is a integrated ADC for sensor data.
Of course having it in “cold storage” completely on your implant is more secure, but it’s also just super not practical to do any sort of large (even into just megabytes) file storage on an implant. Why not just encrypt the file using VeraCrypt or the 256 AES option in 7zip, upload the file to a server, and then store a super secure password on your chip? Or, even more simply, use a service like Mega and just store the key on your implant?