Isn't there a bit of contradiction in saying "TEE AND. NOT TEE OR." and "[...] software-based cryptography [...] *winning* in the long run (7-10 years)."? Not sure I would agree with the latter if winning means NO TEE.
assumption is that computational overhead of FHE comes down to such an extent that is becomes competitive with TEEs. FHE wins because it will be more flexible and upgradable. so the 7-10 years is more of a prediction of the progress of FHE for mass-market confidential compute applications. Maybe I am underestimating the AND complementary scenario here though?
Nice article, just a small nit: guess the "cloud computing consortium" is meant to read "confidential computing consortium"
edited! thanks for the feedback. anything in here you disagree with?
Isn't there a bit of contradiction in saying "TEE AND. NOT TEE OR." and "[...] software-based cryptography [...] *winning* in the long run (7-10 years)."? Not sure I would agree with the latter if winning means NO TEE.
assumption is that computational overhead of FHE comes down to such an extent that is becomes competitive with TEEs. FHE wins because it will be more flexible and upgradable. so the 7-10 years is more of a prediction of the progress of FHE for mass-market confidential compute applications. Maybe I am underestimating the AND complementary scenario here though?