3 ms·
I call all those posh sounding "platform security" types being much like "antivirus" makers, trying to sell a non-solution for a problem to technical illiterate
by baybal2 7y ago
I call all those posh sounding "platform security" types being much like "antivirus" makers, trying to sell a non-solution for a problem to technical illiterates.
What they claim sounds impressive to CEOs and such, but everybody in the industry knows that anybody with straight hands can probe a chip, and somebody with ion beam equipment can do pretty much anything he wants with it.
For example, chips from credit cards, said to be "physically impossible" to tamper with, are known to be copied by somebody in Eastern Europe.
If somebody becomes a victim of credit card cloning, proving the bank that it was not you who withdrew money on other side of the globe becomes nearly impossible without going to a court.
- rkangel 7y agoIn the real world, security is not an absolute. Nothing is perfectly secure. What is important is the resources required to break the security - what that bar needs to be depends on your product and your business goals. In the video he specifically talks about an economic threshold - it must cost more than 10 games ($600) to mod a console. Above that price it's not worth it, so direct chip beam attacks aren't relevant to their threat model. That is the direct 'resource requirement' as a result of their business goal - prevent piracy.
- zxcmx 7y agoI agree with your general point but computers are weird. Attacks and exploits can have scaling properties much like other software. We can do something expensive (say $200k attack cost) to break the platform and sell 10k $100 mod chips. The defender's threat model can't discount expensive attacks; the constraint is that an expensive one-off attack must not enable a cheap bypass. This is hard!
- baybal2 7y agoThis would not do anything about the seed for DRM key being extracted. Those types would only need to get a single key to decrypt all downloads.