[djsumdog]

Well this type of attack has been theoretical for years. The Project Zero referenced some papers from the mid-2000s that talked about it. But the implementation, even today, isn't exactly trivial.

Modern processors are insanely complex systems. Branch prediction, out of order execution, hardware virtual memory management, hardware virtualization, etc. Not to mention that these are side-channel attacks. It's not a direct vulnerability, it requires executing some code and measuring timing very precisely; similar to and oscilloscope and a very expensive safe.

Of course Intel is going to be spinning this however they can for damage control. That's what PR departments do. I still doubt engineers at Intel really thought this attack was plausible, or else they wouldn't have been engineering chips this way for the past decade.

[wyager]

> Modern processors are insanely complex systems.

And until we align their market incentives properly, silicon vendors are going to continue to ignore this fact when it comes to verification. Intel is especially bad here; they’ve had an unreasonable number of hardware bugs in recent years.

[krylon]

> The Project Zero referenced some papers from the mid-2000s that talked about it.

http://www.daemonology.net/papers/htt.pdf