As we all know, nothing in IT is 100% secure. No material, and even less that no software is completely secure, since it can include all types of vulnerabilities and some may be available from the initial design phase. These types of vulnerabilities, called Zero-Day, are the most dangerous because it is not possible to know when they have been exploited by other people’s friends.
Intel and AMD, as well as Microsoft through Windows, offer different layers of security to prevent outsiders from accessing information stored on the equipment, protecting all phases, from power on to power off, through through the data transmission phases. When they detect a problem, they block access.
However, it is not the only method, even if it is the most effective, especially in business. Last week the International Semiconductor Circuits Conference (ISSCC) was held, where a processor designed by the University of Vermont was presented, which automatically destroys itself when it detects that the security of the equipment it it manages been compromised and prevents information from being stolen.
Processors destroyed when unauthorized access is detected
In this way, it prevents data from being manipulated and/or hacked depending on the environment in which it is implemented, which is particularly useful when it comes to protect intellectual property.
To carry out this project, the University of Vermont benefited from the collaboration of Marvell Technology, an experienced semiconductor manufacturer. The work team at this university is led by Eric Hunt-Schroeder, who is also an employee of Marvell Technology.
The idea for this project came from a publication claiming that it was possible to clone SRAM-based non-clonal physical functions (PUFs) using a simple electron microscope.
How it works
This processor works using physical unclonable functions (PUFs), a solution to generate and store keys easily and inexpensively. This is responsible for creating unique fingerprints for each processor. Unclonable physical functions are divided into two types, types which have been implemented on this chip to provide two different forms of self-destruction.
The first way to destroy this chip is electromigration, a process in which electricity moves atoms from their place, causing open circuits and generating a vacuum. The second self-destruct method implemented by this chip subjects transistors operating under 1 V to 2.5 V. In this way, an aging effect called dielectric breakdown is created, which causes a series of short circuits which end up frying the processor.
At the same conference where this chip configured to be destroyed was presented, other security systems aimed at avoiding computer attacks were also presented, such as a method to mask the electromagnetic emissions of processors when they are in operation, broadcasts which, with the proper knowledge, perhaps they can decrypt and steal information important.
