CoChat

Cryptographic transistor for true random number generator with low power consumption

Open access

The DOI matches its Crossref record; 55 of the 55 references it cites were found in a registry or an index; 8 of the 53 sources its claims cite were checked for support.

Abstract

We design a cryptographic transistor (cryptoristor)–based true random number generator (tRNG) with low power consumption and small footprint. This is the first attempt to use irregular and unpredictable operation-induced randomness of a cryptoristor as an entropy source. To extract discrete random numbers with a binary code from the cryptoristor, we developed a noise-coupling analog-to-digital converter. This converter not only converts analog signals to digital random bits but also improves the randomness of the entropy source with low power consumption. The randomness of the cryptoristor is attributed to the random carrier multiplications with the creation and stochastic carrier escape as destruction, which occurs iteratively as long as the input current is fed. The cryptoristor-based tRNG passed 15 randomness test suites of NIST Special Publication 800-22. It is robust to iterative operational stresses and to ambient temperature changes, making it an attractive option for hardware-based security solutions in the Internet of Things due to its low power consumption and small size.

Overview

CoChat Summary

TL;DR

This paper reports a 100% CMOS-compatible true random number generator (tRNG) built from a single cryptographic transistor (“cryptoristor”) plus a noise-coupling ADC, using irregular transistor oscillations as an entropy source. The integrated module produced random bits at 75 kb/s, consumed 10.28 pJ/bit, and passed all 15 NIST SP 800-22 tests without postprocessing.

Key findings

  • The cryptoristor-based tRNG passed all 15 NIST SP 800-22 randomness test suites; the paper reports this both for 100 megabits collected from the cryptoristor and for the integrated tRNG module.
  • The integrated tRNG had a bit generation rate of 75 kb/s and an energy cost of 10.28 pJ/bit; the cryptoristor alone had 3.29 pJ/bit at 100 nA input current and 2.47 V average output voltage.
  • The NC-ADC consumed 524.4 nW in 180-nm CMOS, and the authors state this is roughly 2000-fold less power than other tRNG modules; they estimate 293.95 nW at 28 nm.
  • Under 30,000 s of iterative stress and ambient temperature changes from 300 to 360 K, the paper reports no degradation in randomness quality.

Methods

  • The device is a single cryptoristor fabricated on a 200-mm SOI wafer using CMOS-compatible processing; the paper also built an NC-ADC on a separate 300-mm bulk-Si CMOS chip, and both were mounted on one PCB.
  • The cryptoristor was biased with a constant 100 nA drain current, gate voltage of −1.8 V, and source at ground; the NC-ADC used an 8-bit SAR ADC with noise coupling via charge sharing, plus a low-power floating inverter amplifier.
  • Randomness was evaluated with NIST SP 800-22 tests, including a stress test over 30,000 s and temperature variation experiments from 300 to 360 K.

Limitations

  • The paper itself notes a trade-off between randomness and throughput: longer sampling intervals improved randomness, and fewer output bits improved pass rates, but this reduces bit generation rate.
  • The paper also notes trade-offs in the NC-ADC design between improved randomness and circuit performance, especially die area, power consumption, and linearity.
  • Evident from the text: the work uses two separately fabricated chips on a PCB rather than a single monolithic chip, so the demonstrated module is not yet a fully integrated one-chip implementation.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1126/sciadv.adk6042

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1. Integrated tRNG including cryptoristor and NC-ADC. (A) Photograph of the chip enclosing the cryptoristor and the chip including the NC-ADC. They were connected to a PCB by wire bonding. The cryptoristor supplied by a current source is connected to the NC-ADC with a connector. (B) Operational schematic of the integrated cryptoristor and NC-ADC. (C) Scanning electron microscopy image of the fabricated cryptoristor. (D) Optical microscope photograph of the fabricated NC-ADC (left) and corresponding layout (right). (E) Measured oscillating output voltage (Vout) from the NC-ADC. The inset is a magnification of Vout in the blue box.
Fig. 1. Integrated tRNG including cryptoristor and NC-ADC. (A) Photograph of the chip enclosing the cryptoristor and the chip including the NC-ADC. They were connected to a PCB by wire bonding. The cryptoristor supplied by a current source is connected to the NC-ADC with a connector. (B) Operational schematic of the integrated cryptoristor and NC-ADC. (C) Scanning electron microscopy image of the fabricated cryptoristor. (D) Optical microscope photograph of the fabricated NC-ADC (left) and corresponding layout (right). (E) Measured oscillating output voltage (Vout) from the NC-ADC. The inset is a magnification of Vout in the blue box.CC BY-NC · PubMed Central
Fig. 2. Stochastic oscillation behavior in the fabricated cryptoristor. (A) Optical microscope photograph of the fabricated cryptoristor with its cross-sectional schematic. (B) Measured ID-VD characteristics to show the STL with VLU. When Iin is within the oscillation zone, oscillating Vout is generated. (C) Measured Vout with respect to the time elapsed under the condition of Iin = 100 nA, VS = 0 V, and VG = −1.8 V of less than VT. (D) Close-up view of the measured Vout in (C). One period is composed of a charging and discharging process. The time difference between one Vout,bot and the next Vout,bot is referred to as cycle time (tcycle). The voltage difference between Vout,top and Vout,bot is defined as the peak-to-peak voltage (Vp-p). (E) Distribution of tcycle for 1 s of oscillation. (F) Distribution of Vp-p during 1 s of oscillation.
Fig. 2. Stochastic oscillation behavior in the fabricated cryptoristor. (A) Optical microscope photograph of the fabricated cryptoristor with its cross-sectional schematic. (B) Measured ID-VD characteristics to show the STL with VLU. When Iin is within the oscillation zone, oscillating Vout is generated. (C) Measured Vout with respect to the time elapsed under the condition of Iin = 100 nA, VS = 0 V, and VG = −1.8 V of less than VT. (D) Close-up view of the measured Vout in (C). One period is composed of a charging and discharging process. The time difference between one Vout,bot and the next Vout,bot is referred to as cycle time (tcycle). The voltage difference between Vout,top and Vout,bot is defined as the peak-to-peak voltage (Vp-p). (E) Distribution of tcycle for 1 s of oscillation. (F) Distribution of Vp-p during 1 s of oscillation.CC BY-NC · PubMed Central
Fig. 3. Working principle of Vout oscillation. Oscillations of triangular Vout and IS are caused by the iterative charging and discharging processes from (1) to (4). When VD reaches VLU, as shown in Fig. 2B, impact ionization for the charging is triggered to drive Vtop and the peak of IS, and out-diffusion for discharging is enabled to lead Vbot and the base of IS. The number of carriers generated by impact ionization fluctuates randomly, and the number of carriers that disappear due to out-diffusion, as quantified with IS, also varies randomly, as shown in fig. S2.
Fig. 3. Working principle of Vout oscillation. Oscillations of triangular Vout and IS are caused by the iterative charging and discharging processes from (1) to (4). When VD reaches VLU, as shown in Fig. 2B, impact ionization for the charging is triggered to drive Vtop and the peak of IS, and out-diffusion for discharging is enabled to lead Vbot and the base of IS. The number of carriers generated by impact ionization fluctuates randomly, and the number of carriers that disappear due to out-diffusion, as quantified with IS, also varies randomly, as shown in fig. S2.CC BY-NC · PubMed Central
Fig. 4. Schematic circuit of the NC-ADC. (A) Overall configuration of the tRNG module including the cryptoristor and NC-ADC. The sampled voltage of the cryptoristor and the residue voltage from the NC-ADC with the previous sample are added (i.e., noise coupling) to Csam by charge sharing, and the noise-coupled voltage of Csam is amplified by floating inverter amplifier (FIA) and used as the next input for the NC-ADC. The residue voltage is continuously generated by CDAC. Random digital bits are generated through the comparator and successive approximation register (SAR) logic. (B) Timing diagram of the NC-ADC. (C) Schematic of the generation of Vres acting as the quantization error. VIN, which is the input node voltage of CDAC, is quantized sequentially for the 8-bit decision. We applied 1.2 and 0 V for the reference voltages, and the common mode voltage (VCM) is the median of the two reference voltages. (D) Input network of the cryptoristor module. The output voltage of the cryptoristor (Vout) is sampled on the small capacitor Csam during Φ1. The noise-coupling process is accomplished by charge sharing between Csam (Vout) and CDAC (Vres), producing Vcouple during Φ2. Then, Vcouple is amplified by FIA and transmitted to CDAC through Φ3. The symbol, which is a circle enclosing a parallelogram, implies the cryptoristor. In the parallelogram, the lower base denotes an HRS, the upper base represents an LRS, and both oblique sides express latch-up by impact ionization and latch-down by out-diffusion.
Fig. 4. Schematic circuit of the NC-ADC. (A) Overall configuration of the tRNG module including the cryptoristor and NC-ADC. The sampled voltage of the cryptoristor and the residue voltage from the NC-ADC with the previous sample are added (i.e., noise coupling) to Csam by charge sharing, and the noise-coupled voltage of Csam is amplified by floating inverter amplifier (FIA) and used as the next input for the NC-ADC. The residue voltage is continuously generated by CDAC. Random digital bits are generated through the comparator and successive approximation register (SAR) logic. (B) Timing diagram of the NC-ADC. (C) Schematic of the generation of Vres acting as the quantization error. VIN, which is the input node voltage of CDAC, is quantized sequentially for the 8-bit decision. We applied 1.2 and 0 V for the reference voltages, and the common mode voltage (VCM) is the median of the two reference voltages. (D) Input network of the cryptoristor module. The output voltage of the cryptoristor (Vout) is sampled on the small capacitor Csam during Φ1. The noise-coupling process is accomplished by charge sharing between Csam (Vout) and CDAC (Vres), producing Vcouple during Φ2. Then, Vcouple is amplified by FIA and transmitted to CDAC through Φ3. The symbol, which is a circle enclosing a parallelogram, implies the cryptoristor. In the parallelogram, the lower base denotes an HRS, the upper base represents an LRS, and both oblique sides express latch-up by impact ionization and latch-down by out-diffusion.CC BY-NC · PubMed Central
Fig. 5. Reliability of the tRNG module to iterative stress and robustness of cryptoristor to temperature variations. (A) Vout through the NC-ADC before the stress test. (B) Vout through the integrated tRNG module after iterative stress lasting for 30,000 s. (C) Averaged Ppass of Vout through the NC-ADC for nine brief sub-suite tests of NIST SP 800-22 between a fresh and aged condition. (D) Directly measured STL characteristics (ID-VD) from the cryptoristor for various ambient temperatures at VG = −1.6 V. (E) Vout oscillation for different temperatures. (F) Averaged Ppass of Vout from the cryptoristor for nine brief sub-suite tests of the NIST SP 800-22 under various temperature conditions.
Fig. 5. Reliability of the tRNG module to iterative stress and robustness of cryptoristor to temperature variations. (A) Vout through the NC-ADC before the stress test. (B) Vout through the integrated tRNG module after iterative stress lasting for 30,000 s. (C) Averaged Ppass of Vout through the NC-ADC for nine brief sub-suite tests of NIST SP 800-22 between a fresh and aged condition. (D) Directly measured STL characteristics (ID-VD) from the cryptoristor for various ambient temperatures at VG = −1.6 V. (E) Vout oscillation for different temperatures. (F) Averaged Ppass of Vout from the cryptoristor for nine brief sub-suite tests of the NIST SP 800-22 under various temperature conditions.CC BY-NC · PubMed Central

Verification

Claims the paper makes

4 Valid · 2 Not valid · 6 Not checked · 52 Own results and interpretations

33 more claims

Create a free account to see all 64 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 11
  • It may state two separate claims as one. 20
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Citations

Cites

Show all 55 references

Cited by

See all 33

Metadata

Authors
Kim, S.-I., You, H.-J., Kim, M.-S., An, U.-S., Kim, M.-S., Lee, D. H., Ryu, S.-T., & Choi, Y.-K.
Affiliations
  • Korea Advanced Institute of Science and Technology (KR)
  • Hanbat National University (KR)
Corresponding author
Choi, Yang-Kyu
Venue
Science Advances
Published
2024-02-23
Volume
10
Issue
8
Pages
eadk6042-eadk6042
ISSN
2375-2548 (ISSN-L)
OpenAlex
W4392094046
Cited by
33
References
46
Citations by year
2026: 10 · 2025: 21 · 2024: 2
FWCI
10.24
Type
journal article
Language
en
Open access
Licence
cc-by-nc
Locations
  • Science Advances (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
Funders
  • National Research Foundation (RS-2023-00217888, 2022M3F3A2A01072851)
  • National Research Foundation of Korea (4120200113769, RS‐2023‐00217888, 2022M3F3A2A01072851, RS‐2023‐00260637)
  • Ministry of Science and ICT, South Korea (2022M3F3A2A01072851)
Topics
  • Physical Sciences › Computer Science › Hardware and Architecture › Physical Unclonable Functions (PUFs) and Hardware Security
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Analog and Mixed-Signal Circuit Design
  • Physical Sciences › Computer Science › Artificial Intelligence › Cryptographic Implementations and Security
Keywords
true random number generator, hardware randomness, entropy source, carrier escape, low power consumption, Internet of Things security, NIST SP 800-22 randomness tests

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Cryptographic transistor for true random number generator with low power consumption. https://app.cochat.ai/papers/10.1126/sciadv.adk6042
The paper
Kim, S.-I., You, H.-J., Kim, M.-S., An, U.-S., Kim, M.-S., Lee, D. H., Ryu, S.-T., & Choi, Y.-K. (2024). Cryptographic transistor for true random number generator with low power consumption. Science Advances. https://doi.org/10.1126/sciadv.adk6042

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.