DC Cryogenic Modeling of Open-Source SkyWater 130 nm MOSFETs at 77 K Using BSIM4
Pith reviewed 2026-05-09 21:00 UTC · model grok-4.3
The pith
Re-extracting a subset of BSIM4 parameters from 77 K measurements produces a usable isothermal SPICE model for SkyWater 130 nm MOSFETs with roughly 20 percent average error.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We created a cryogenic modeling approach to produce a SPICE-compatible, isothermal BSIM4-based model for select transistor sizes at 77 K. The resulting model agrees with data at 77 K with an average error on the order of 20% (relative RMS) and shows no dependence on drain voltage.
What carries the argument
An isothermal BSIM4 model created by re-extracting a subset of its parameters directly from 77 K DC characteristic data.
If this is right
- The model can be dropped directly into standard SPICE simulators for 77 K circuit design.
- The open-source SkyWater 130 nm process can support low-power electronics that operate inside liquid-nitrogen detectors.
- Public release of the parameter sets lets other groups build and verify circuits at 77 K without repeating the full extraction work.
- The lack of drain-voltage dependence in the error indicates the basic current-voltage behavior is captured without extra corrections.
Where Pith is reading between the lines
- The same re-extraction tactic could be tried at other cryogenic temperatures such as 4 K to see how far the existing BSIM4 equations stretch.
- If the 20 percent error is acceptable for early design, it reduces the immediate need for fully custom cryogenic compact models.
- Public models for a widely accessible process may speed development of integrated readout chips for quantum sensors or space instruments that must run cold.
Load-bearing premise
Re-extracting a subset of BSIM4 parameters from 77 K DC data is sufficient to produce a usable isothermal model without additional temperature-dependent equations or self-heating corrections.
What would settle it
Testing the same extracted parameters on a transistor size or bias condition not used in the fit and finding errors well above 20 percent relative RMS would show the approach is incomplete.
read the original abstract
Cryogenic applications in high-energy physics (HEP) demand reliable, low-power CMOS electronics capable of operating at liquid nitrogen temperatures (77 K). The open-source SkyWater 130 nm (SKY130) CMOS process has previously been shown to operate at temperatures as low as 4 K making it a promising candidate for HEP applications. In this work, we characterize and model SKY130 low-threshold voltage transistors at 77 K, which is a temperature commonly used in modeling applications for liquid argon detectors. DC characteristic measurements were performed at both room temperature and liquid nitrogen temperature. We created a cryogenic modeling approach to produce a SPICE-compatible, isothermal BSIM4-based model for select transistor sizes at 77 K. The resulting model agrees with data at 77 K with an average error on the order of 20% (relative RMS) and shows no dependence on drain voltage. Due to the open-source nature of SKY130, we have made our models publicly available on Github. We hope this work will continue the trend for democratizing circuit design at cryogenic temperatures in high-energy physics by enabling open access to accurate CMOS device models at 77 K.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper characterizes SkyWater 130 nm low-Vt MOSFETs via DC measurements at 300 K and 77 K, then constructs an isothermal BSIM4 model by re-extracting a subset of parameters from the 77 K data for selected device sizes. It reports that this SPICE-compatible model matches the 77 K measurements with ~20% relative RMS error, exhibits no observed drain-voltage dependence, and is released publicly on GitHub to support cryogenic circuit design for high-energy physics applications such as liquid-argon detectors.
Significance. If the validation holds, the work supplies a practical, open-source isothermal model for an accessible 130 nm process at 77 K, directly addressing the need for reliable cryogenic CMOS in HEP. The public release is a clear strength that lowers barriers for circuit designers. The reported 20% RMS error, however, implies the model is best suited for initial design exploration rather than precision analog work, and the isothermal assumption must be shown to remain accurate under realistic power dissipation.
major comments (1)
- [Abstract] Abstract: The central claim that the fitted BSIM4 model 'agrees with data at 77 K with an average error on the order of 20% (relative RMS) and shows no dependence on drain voltage' is load-bearing for the paper's utility. The manuscript provides no information on the number of devices tested, the exact subset of BSIM4 parameters that were re-extracted, the bias window used for fitting versus validation, or any checks against self-heating or unmodeled cryogenic mechanisms (carrier freeze-out, interface traps, velocity saturation shifts). Without these details it is impossible to judge whether the 20% error is representative or whether residual Vd dependence appears outside the extraction conditions.
minor comments (2)
- The abstract states that models are 'made publicly available on Github' but does not give the repository URL or commit hash; this link should appear explicitly in the main text or a footnote to enable immediate reproducibility.
- The phrase 'select transistor sizes' is used without listing the specific W/L values modeled. Adding a short table or sentence with the dimensions would improve clarity and allow readers to assess applicability to their circuits.
Simulated Author's Rebuttal
We thank the referee for their thorough review and constructive feedback. We address the major comment point by point below. The referee correctly identifies that certain methodological details were not sufficiently explicit in the original manuscript, and we have revised the text to incorporate them.
read point-by-point responses
-
Referee: [Abstract] Abstract: The central claim that the fitted BSIM4 model 'agrees with data at 77 K with an average error on the order of 20% (relative RMS) and shows no dependence on drain voltage' is load-bearing for the paper's utility. The manuscript provides no information on the number of devices tested, the exact subset of BSIM4 parameters that were re-extracted, the bias window used for fitting versus validation, or any checks against self-heating or unmodeled cryogenic mechanisms (carrier freeze-out, interface traps, velocity saturation shifts). Without these details it is impossible to judge whether the 20% error is representative or whether residual Vd dependence appears outside the extraction conditions.
Authors: We agree that the abstract and main text should provide these details to allow readers to evaluate the model's scope and limitations. In the revised manuscript we have expanded the abstract and added a dedicated methods subsection that specifies: (i) measurements were performed on multiple devices across the selected sizes (W/L combinations reported in Table I); (ii) the subset of BSIM4 parameters re-extracted at 77 K consists of VTH0, U0, VSAT, and a small number of related mobility and DIBL terms while all other parameters retain their 300 K values; (iii) parameter extraction used the full measured bias window (Vgs from 0 to 1.8 V, Vds from 0 to 1.8 V) with a hold-out validation set at intermediate bias points; and (iv) a short discussion of self-heating (negligible under the low-duty-cycle DC conditions used) together with the observation that the isothermal model already absorbs the net effect of carrier freeze-out and interface-trap shifts through direct fitting to the 77 K data. The reported 20 % relative RMS error and absence of Vd dependence are evaluated over the entire measured Vds range, not only the fitting subset. These additions make the central claim fully traceable without altering the reported error metric. revision: yes
Circularity Check
No significant circularity; standard parameter extraction for isothermal model
full rationale
The paper describes DC measurements at room temperature and 77 K followed by re-extraction of a subset of BSIM4 parameters to generate an isothermal SPICE model for selected device sizes. The reported ~20% relative RMS agreement and lack of drain-voltage dependence are presented as properties of the fitted model rather than independent predictions of unseen data. No derivation chain, first-principles result, uniqueness theorem, or self-citation is invoked; the work is empirical modeling whose central output (the public model) is the direct result of the fitting step itself. This is a normal, non-circular modeling workflow.
Axiom & Free-Parameter Ledger
free parameters (1)
- BSIM4 model parameters
axioms (2)
- domain assumption BSIM4 compact model framework remains valid at 77 K when parameters are re-extracted from measurements
- domain assumption Device operation is isothermal with negligible self-heating
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.