Assuming reverse engineering that reads the circuit by removing a chip's layers one at a time and imaging them with an electron microscope, or by imaging its interior with X-rays, which is an appropriate design approach?

Answer: A. Mask ROM values and wiring shapes can be read from images, so secret keys are not built into circuit shapes but made unique per device and protected together with physical measures

Mask ROM values are built into the shape of the circuit as the presence or absence of transistors or contacts, so they can be read by removing layers and imaging them. Moreover, every chip made with the same mask contains the same values, so analysing one leaks the secrets of every device. In the 2020s, research has also reported using X-ray imaging to capture internal wiring in fine detail without destroying the chip. Secrets are therefore made unique per device rather than relying on circuit shapes; for storage whose programmed state may be visible, such as eFuses, measures like storing keys encrypted with a key derived from a PUF are used, read-out paths are limited and tamper-resistance measures are combined.

• A. Correct. Secrets do not rely on circuit shapes but are protected by per-device keys and physical measures.
• B. Even a complex layout can be read with enough time and money. Relying on hiding shapes for secrecy is dangerous.
• C. It has been shown that X-rays can image internal structures, so merely hardening the surface is not enough.