From ISSCC to GOMACTech, we’re presenting new and exciting technology solutions that will benefit the aerospace and defense industry.
For anyone who’s watched the 1990s cartoon, Inspector Gadget, the phrase “This message will self-destruct” is bound to be familiar. Anytime Inspector Gadget is given a new assignment, he receives a message containing critical information and instructions. Inspector Gadget must act quickly because shortly after opening, the message is promptly blown to smithereens.
Recently, I had the opportunity to present self-destructing encryption key technology within microchips during the Institute of Electrical and Electronics Engineers (IEEE) International Solid-State Circuits Conference (ISSCC) hosted in San Francisco from February 18 to 22. If the chip or keys become compromised, the keys self-destruct. Essentially, we’ve created security on the fly—the Inspector Gadget way. We received a significant amount of media attention from publications such as IEEE Spectrum and Tom’s Hardware.
But we didn’t stop with ISSCC. Shortly after, I had the opportunity to attend the GOMACTech conference where I presented our reconfigurable physically unclonable function (PUF) and random number generation technology—all in one chip. This technology promises to be critical for the aerospace and defense industry.
Pushing technology forward through collaboration
The reconfigurable PUF and random number generation technology was created as part of an industry-academia collaboration. I’m a senior staff manager at Marvell working on my Ph.D. at the University of Vermont (UVM). UVM is a Tech Hub for semiconductor research under the CHIPS and Science Act, which authorized $10 billion in funding for the Tech Hubs Program. This funding is helping companies build strong relationships with local flagship universities to deepen collaboration and push technology forward.
Two separate concepts, one seamless design
Any secure system requires keys from static and dynamic entropy. Conventionally, these capabilities come from two standalone blocks: The PUF which provides the static entropy, and the true random number generators, or TRNGs, which provide the dynamic entropy. Traditionally, static and dynamic entropy comes from two dedicated areas, both of which function independently. Instead, we are proposing a new design, which implements both functions in one, reducing area and power while increasing efficiency. At the flip of a switch, the user can easily change between the static and dynamic entropy modes.
This new design will also allow users to easily change the static PUF output response. If the PUF key becomes compromised or needs to be refreshed, users can reconfigure the PUF output response and generate a completely new key. Alternatively, when the design operates in dynamic entropy mode it creates an unpredictable and random set of data in the same silicon, allowing the design to operate as a TRNG. Once done, they can easily switch back to static mode, which generates the new stable, repeatable key. By generating a new random key to reconfigure the PUF, they can continue using the same chip, which bypasses long approval and procurement processes.
Caption: A reconfigurable PUF has multiple keys and fingerprints.
What Makes a Reconfigurable PUF?
A conventional silicon PUF has one key and one fingerprint. Our proposed technology will allow us to decompose the cell in a pull-up or pull-down network and regroup the devices, allowing users to change the static key and get an alternate, repeatable key—or multiple fingerprints. We wanted to get a lot of re-use in our design, so we’re sharing the circuits. Ultimately, users can invalidate the original key and generate a new key that is the only valid option going forward, creating an added layer of protection.
To create a TRNG, we needed to convert the stable (repeatable) bitcells into unstable (unrepeatable) bitcells. When we drive these devices into a high-gain state, small deviations in voltage show up big in the operating response and output. When we randomly activate multiple bitcells with variable overdrive that are very near their threshold voltages, the result is very high instability. Once that threshold is reached, the random number generation begins, flipping the variables to 1s or 0s.
Caption: The unstable bit region enables random number generation.
We are extremely excited about our new proposed technology and believe that the application opportunities are endless. By enabling static and dynamic entropy within the same chip, we’ve created a flexible, secure solution that will allow our customers to protect their information in a faster and more efficient way—all without changing the chip.
We are thrilled to continue to develop innovative solutions that will push the government microelectronics industry forward. Stay tuned for more updates.
This blog contains forward-looking statements within the meaning of the federal securities laws that involve risks and uncertainties. Forward-looking statements include, without limitation, any statement that may predict, forecast, indicate or imply future events or achievements. Actual events or results may differ materially from those contemplated in this blog. Forward-looking statements are only predictions and are subject to risks, uncertainties and assumptions that are difficult to predict, including those described in the “Risk Factors” section of our Annual Reports on Form 10-K, Quarterly Reports on Form 10-Q and other documents filed by us from time to time with the SEC. Forward-looking statements speak only as of the date they are made. Readers are cautioned not to put undue reliance on forward-looking statements, and no person assumes any obligation to update or revise any such forward-looking statements, whether as a result of new information, future events or otherwise.
Copyright © 2024 Marvell, All rights reserved.