CrossBar Reimagines ReRAM Technology for Physically Unclonable Functions
As the need for hardware security increases, companies are searching for ways to use PUF. CrossBar claims its newest resistive-RAM technology could be suited for a new class of PUF applications.
As the world becomes more digitized, security becomes more and more critical. One of the more popular hardware security devices today is the Physically Unclonable Function (PUF). A highly researched field, PUFs are seeing new techniques and approaches being developed constantly.
The latest group to add to the slew of PUF techniques is the resistive-RAM (ReRAM) company CrossBar. Earlier this week, it announced the use of its ReRAM devices for PUF applications.
CrossBar hopes to bring ReRAM PUFs to embedded devices. Image from CrossBar
This article will briefly discuss PUFs, ReRAM, and how CrossBar’s new technology may work.
A Brief Intro to PUFs
PUFs are a technique in hardware security that exploits inherent device variations to produce an unclonable, unique device response to a given input. PUFs are often used in ICs for security purposes, including cryptographic key generation, random number generation, and device authentication.
A comparison of PUF-based key generation (a) and hardware entangled (b) for cryptography. Image used courtesy of Maes and Verbauwhede
Unlike software solutions, PUFs can create a genuinely unpredictable IC identifier or cryptographic key by exploiting true randomness in nature. This unpredictability increases security since keys cannot be predicted based on some deterministic or quasi-deterministic process.
A Brief Intro ReRAM
A ReRAM (or RRAM) is a device with a resistance value that can be set by applying a control signal in external voltage/current. They are a type of non-volatile memory, offering uniquely high switching speeds, uniquely small size, and uniquely low power consumption.
RRAM working principle. Image used courtesy of Meena et al
RRAM devices typically consist of two metal electrodes around a resistive oxide layer. When appropriate voltages are applied, the RRAM forms conductive paths called filaments, resulting from metal migration and physical defects, which is the "on" state of the device. To reverse the state, you'll need to apply a different external voltage to form an "off" state.
Notably, the resistance of ReRAM is highly subject to cycle-to-cycle variation due to defects in the filaments and the thermal voltage fluctuations.
Now that a basic understanding of ReRAMs is understood, let's see what CrossBar's ReRAM looks like.
How CrossBar’s ReRAM Works
Recently, CrossBar has announced that its ReRAM can be used for PUF applications. The most available information on its ReRAM's PUF states that its PUF can be either a single ReRAM cell or dual ReRAM cell configuration.
With the lack of available information, luckily, there is a plethora of academic research available that we can pull from to make an educated guess.
CrossBar’s ReRAM PUF is either a single or dual ReRAM cell. Image from CrossBar
One researched type of ReRAM PUFs is based on a 1T-1R bit cell and crossbar array. In a 1T-1R architecture, the challenge is a column multiplexer that randomly selects and reads two cells of a row. A current sense amplifier generates the response by comparing the resistance of two cells.
Another known ReRAM PUF is a crossbar array PU. This type of PUF implements an arbiter PUF by serially connecting the ReRAM devices in a crossbar array. Their responses are XORed to generate a final response.
In general, PUFs are a great option in hardware security. Seeing new techniques being developed in the industry is always a good sign, with CrossBar's announcement showing both the increasing ubiquity of PUFs and the versatility of ReRAM devices.
Interested in other ReRAM related news? Read more in the articles down below.
French Researchers Exploit Non-Idealities of Memristors to Bring AI to the Edge
Choosing Between 2D and 3D Materials to Set Off the Commercialization of Next-Gen Semiconductors
Materials Spotlight: How Perovskite Was a Hotbed of 2020 Research