Real life NEXUS INTRUDER PROGRAM

[nn_step]

For those of you who are unfamiliar with the "Reflections on Trusting Trust" paper basically it is an attack that uses high level languages abstraction features against the programmer. Making it important in the world of trusted compilers to know on the assembly level that the compiler is trusted. However, I've chosen to take it into another far more interesting direction. I've created an attack that infects several Verilog implementations and thus puts a custom backdoor circuit inside of every single new circuit that exceeds a predefined transistor count [to keep it hidden], which is currently set at 14,000 transistors because it requires a little over 1,400 transistors to perform all of the cooperation and control logic that I was aiming for. It also specifically targets microcode and randomly wired control logic because it is the optimal place to hide.

Here is an example of an infected circuit in the attachment. [A 32bit CPU core with 32,000 transistors]
I hope to publish a complete version of it in ACM, once I've worked out several unusual quirks

[Rman Virgil]

.

Facinating and scary.

Stuff like this always makes me think of Grid vulnerability and the consequences within a week or two, of one brought down.
.

[nn_step]

.

Facinating and scary.

Stuff like this always makes me think of Grid vulnerability and the consequences within a week or two, of one brought down.
.

Now the really interesting thing is that if someone combined it with say a smart enough program that replaces or modifies the local operating system and a microcode infecter, it could spread to every single computer on the planet in less than a decade. The worst part of it all is there is nothing one can to do stop something like that short of starting back with 1970s tech and building chains of trusted designs.

For those unfamiliar microcode infecter programs update the microcode on modern microprocessors with a version that contains a backdoor or similar exploit. With the added benefit of that the only way to remove it is to take the infected CPU and throw it into the trash can.

And don't get me started on the cluster f**k of flash based BIOS when it comes to similar style attacks.

[Tenoh]

yeah,hehe... he gonan publish it and then militery gonna hang him by the ballz or otherwise he will dissupear lol!

[nn_step]

yeah,hehe... he gonan publish it and then militery gonna hang him by the ballz or otherwise he will dissupear lol!

actually from the looks of things. Most people just seem to completely ignore it and ACM has rejected it and refused to comment on why. Perhaps the concept of hardware integrity just makes people too uncomfortable to discuss it. [Largely because it is an unfixable problem for 99.99% of the world's population ]