nanoCMB: Minimal CMB Calculator & Concept

• nanoCMB is a compact computational framework that provides a minimal yet accurate Python implementation for unlensed CMB temperature and polarization power spectra.
• It employs optimized line-of-sight integration, RECFAST recombination, and custom non-uniform grids to achieve sub-percent accuracy compared to full-scale Boltzmann solvers like CAMB.
• The nanoCMB concept extends beyond software to include innovative detector engineering and experimental strategies aimed at precise CMB polarization and anisotropic birefringence measurements.

Searching arXiv for nanoCMB and the cited papers to ground the article. The term nanoCMB refers primarily to a compact computational framework for Cosmic Microwave Background power-spectrum prediction and, in a broader usage found in adjacent CMB literature, to a design philosophy that emphasizes minimal, explicit, and nanoscale-engineered approaches to CMB science. In its most concrete and formal sense, nanoCMB is the Python code introduced in “nanoCMB: A minimal CMB power spectrum calculator in Python” (Moss, 26 Feb 2026), a single-script implementation of a flat-$\Lambda$CDM Boltzmann solver that computes the unlensed scalar temperature and polarization spectra $C_\ell^{TT}$, $C_\ell^{EE}$, and $C_\ell^{TE}$. The same label is also used informally in the supplied literature to frame prospective Northern Hemisphere CMB polarization facilities and detector programs whose science case centers on B-mode polarization, anisotropic cosmic birefringence, and thin-film, electron–phonon-engineered sensor architectures (Zhong et al., 2024, Nones et al., 2012, Bernardis et al., 2012). This dual usage makes nanoCMB both a specific software artifact and an organizing concept at the interface of theoretical calculation, instrumental design, and precision CMB inference.

1. Definition and scope

In the software sense, nanoCMB is a minimal but accurate calculator for the unlensed CMB temperature and polarisation angular power spectra of flat $\Lambda$CDM cosmologies (Moss, 26 Feb 2026). Its stated purpose is not to replace large production Boltzmann solvers such as CAMB, but to provide a pedagogical bridge between textbook treatments and research-level Boltzmann solvers, with the entire calculation contained in a single, easily modifiable Python script. The code computes only scalar, unlensed spectra for flat $\Lambda$CDM with $w=-1$, adiabatic initial conditions, and massless neutrinos parameterized by $N_{\rm eff}$; it excludes tensor modes, lensing, curvature, massive neutrinos, isocurvature modes, and exotic dark energy (Moss, 26 Feb 2026).

A broader “nanoCMB” usage appears in the accompanying detector and experimental literature. In that usage, the term denotes a next-generation CMB program that combines ultra-sensitive polarization measurements, large-format detector arrays, wide frequency coverage, strict control of systematic errors, and, in some formulations, nanoscale detector engineering. The review by de Bernardis and Masi describes the experimental and scientific environment from which such a program would emerge, emphasizing B-mode polarization, multi-band foreground separation, cryogenic optics, and mapping-speed scaling with detector count (Bernardis et al., 2012). The NbSi TES work presents an explicitly thin-film, nanomaterial-based detector architecture that the supplied material characterizes as a direct example of a “nanoCMB” approach, because the thermal isolation is achieved by intrinsic electron–phonon decoupling rather than membrane micromachining (Nones et al., 2012).

This suggests that the common denominator of the term is minimality with explicit physical control: in software, a compact line-of-sight Boltzmann pipeline; in instrumentation, compact yet physically transparent detector and survey architectures.

2. nanoCMB as a Boltzmann code

nanoCMB, as introduced in 2026, is written in $\sim$1400 lines of readable Python and implements the full line-of-sight integration method with RECFAST recombination, coupled Einstein–Boltzmann perturbation equations in synchronous gauge, a tight-coupling approximation, precomputed spherical Bessel function tables, and optimally constructed non-uniform grids in wavenumber and conformal time (Moss, 26 Feb 2026). The code calculates the scalar unlensed spectra through the standard transfer-function integral

$$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$

with the primordial curvature power spectrum

$C_\ell^{TT}$0

The background cosmology is expressed in conformal time $C_\ell^{TT}$1 for a flat FRW metric,

$C_\ell^{TT}$2

with a Friedmann-sector implementation written in CAMB-like units. The code defines

$C_\ell^{TT}$3

so that

$C_\ell^{TT}$4

It also computes the sound horizon

$C_\ell^{TT}$5

The perturbation evolution follows the synchronous-gauge Einstein–Boltzmann hierarchy in the CDM rest frame, with photons, baryons, CDM, and massless neutrinos. Photon temperature and polarization multipoles, neutrino multipoles, baryon velocity and density, CDM density, and the synchronous metric variables are evolved in conformal time. The code uses a line-of-sight formulation rather than direct evolution of photon multipoles to large $C_\ell^{TT}$6, so the hierarchy is truncated at $C_\ell^{TT}$7 and then projected to the full angular range through source functions and spherical Bessel kernels (Moss, 26 Feb 2026).

The temperature transfer function is written as

$C_\ell^{TT}$8

and the polarization transfer function as

$C_\ell^{TT}$9

For numerical stability, the temperature source is re-expressed in three channels multiplying $C_\ell^{EE}$0, $C_\ell^{EE}$1, and $C_\ell^{EE}$2, with explicit source terms

$C_\ell^{EE}$3

$C_\ell^{EE}$4

For E-mode polarization,

$C_\ell^{EE}$5

with $C_\ell^{EE}$6 and $C_\ell^{EE}$7 the visibility function (Moss, 26 Feb 2026).

3. Recombination, thermodynamics, and numerical design

A central feature of nanoCMB is that it includes a compact implementation of RECFAST. Hydrogen recombination is treated with a Peebles-type equation,

$C_\ell^{EE}$8

with the Peebles factor

$C_\ell^{EE}$9

while the matter temperature evolves through Compton coupling and adiabatic cooling: $C_\ell^{TE}$0 Helium recombination is handled by the RECFAST-style effective three-level treatment, including singlet and triplet channels and hydrogen continuum opacity corrections (Moss, 26 Feb 2026).

Reionization is modeled with the CAMB $C_\ell^{TE}$1 parameterization in $C_\ell^{TE}$2 with width $C_\ell^{TE}$3, plus a second helium reionization at $C_\ell^{TE}$4. The Thomson opacity and visibility are then constructed as

$C_\ell^{TE}$5

The diffusion damping scale is also computed: $C_\ell^{TE}$6

The numerical architecture is deliberately explicit. nanoCMB uses two non-uniform $C_\ell^{TE}$7-grids: an ODE grid of about 200 modes for perturbation evolution and a fine grid of about 4000 modes for line-of-sight and $C_\ell^{TE}$8 integration. The ODE grid is built from the weight

$C_\ell^{TE}$9

and the fine grid from a related weight that includes Gaussian windows around $\Lambda$0. The conformal-time grid is also non-uniform, concentrated around recombination, the acoustic source region, and reionization. Heavy numerical work is delegated to SciPy ODE solvers, trapezoidal quadrature on these tailored grids, and cached spherical Bessel tables extending to $\Lambda$1 (Moss, 26 Feb 2026).

4. Accuracy, validation, and relation to CAMB

The principal empirical claim attached to nanoCMB is that, despite its brevity, it reaches sub-percent agreement with CAMB across the multipole range $\Lambda$2, with a runtime of $\Lambda$310 seconds on a modern laptop (Moss, 26 Feb 2026). For a Planck-like fiducial cosmology, the reported ratios of nanoCMB to CAMB for $\Lambda$4 and $\Lambda$5 are close to unity, with typical scatter at the $\Lambda$6 level across most of the multipole range. Across a 50-cosmology Latin hypercube sample spanning $\Lambda$7 of the Planck 2018 posterior, the RMS residual in $\Lambda$8 is typically $\Lambda$9 for $\Lambda$0, rising to about $\Lambda$1 in the highest multipole bin, while $\Lambda$2 is generally at the $\Lambda$3–$\Lambda$4 level except at very low $\Lambda$5, where the signal is extremely small and fractional metrics become unstable (Moss, 26 Feb 2026).

The validation is multi-layered. The background expansion history agrees with CAMB at the $\Lambda$6 level; the thermodynamic quantities $\Lambda$7, $\Lambda$8, and $\Lambda$9 agree to $w=-1$0; representative perturbation variables such as $w=-1$1, $w=-1$2, $w=-1$3, $w=-1$4, and $w=-1$5 agree at about the $w=-1$6 level or better across relevant $w=-1$7-modes; and the transfer functions $w=-1$8 are in sub-percent agreement over the $w=-1$9-ranges that dominate the final spectra (Moss, 26 Feb 2026).

Its limitations are equally explicit. The code computes only unlensed spectra, so it is not a tool for lensed $N_{\rm eff}$0 analysis, delensing studies, tensor-mode inference, or beyond-$N_{\rm eff}$1CDM precision forecasting. This makes nanoCMB best understood as a reference implementation or sandbox rather than a production inference engine. A plausible implication is that its scientific value lies less in breadth of functionality than in transparency: approximations such as $N_{\rm eff}$2, tight-coupling thresholds, and the reionization prescription are exposed directly rather than hidden behind configuration layers (Moss, 26 Feb 2026).

5. Pedagogical and research roles

nanoCMB occupies a specific niche between textbook derivations and large Boltzmann codes. Textbooks derive the harmonic formalism,

$N_{\rm eff}$3

and the spin-2 decomposition of polarization,

$N_{\rm eff}$4

but they usually do not spell out a complete implementation of recombination, tight coupling, source construction, non-uniform sampling, and stable line-of-sight projection. The CMB review by de Bernardis and Masi provides the broader observational framework in which such calculations sit: the blackbody CMB at $N_{\rm eff}$5, temperature anisotropy of order $N_{\rm eff}$6 rms, polarization of order $N_{\rm eff}$7 rms, and a rotational component below $N_{\rm eff}$8 rms (Bernardis et al., 2012). nanoCMB makes these textbook observables calculable within a compact code base (Moss, 26 Feb 2026).

The code is therefore valuable in at least three technical roles. First, it is a reference implementation of the scalar CMB pipeline, from background cosmology through recombination, perturbation evolution, visibility construction, source assembly, and line-of-sight integration. Second, it is a prototype platform for methodological modifications such as alternative recombination models, different tight-coupling schemes, or extensions to new species, because the state vector and right-hand sides are compact and explicit. Third, it is a didactic benchmark against CAMB: because every approximation is exposed, users can assess how specific numerical choices propagate into $N_{\rm eff}$9 differences (Moss, 26 Feb 2026).

This suggests that nanoCMB is especially well suited for technically literate cosmologists who need to understand, alter, or audit a Boltzmann pipeline at the level of equations and numerical design rather than merely consume its outputs.

6. The broader nanoCMB program: detectors, platforms, and precision polarimetry

Outside the software paper, the supplied literature uses “nanoCMB” to describe an experimental program aimed at next-generation CMB polarization and fundamental-physics measurements. The review literature identifies the central scientific drivers: direct access to the early universe through the CMB blackbody and anisotropy fields, precision measurements of linear polarization at large and intermediate angular scales, and the need to reduce systematic effects and foreground contamination as detector sensitivity improves (Bernardis et al., 2012). It emphasizes that once experiments become background-limited, mapping-speed gains scale primarily with the number of detectors,

$\sim$0

which motivates large arrays of TES, MKID, or related detectors (Bernardis et al., 2012).

The detector paper on high-impedance NbSi TES sensors provides a materially concrete expression of a nanoCMB-style instrumentation philosophy. It describes a bolometer architecture based on a 50 nm amorphous Nb$\sim$1Si$\sim$2 film patterned into meanders, in which absorber, thermometer, and thermal link collapse into a single thin film. Instead of membrane-supported thermal isolation, the design uses intrinsic electron–phonon decoupling: $\sim$3 with

$\sim$4

The reported device characteristics include $\sim$5, $\sim$6–$\sim$7, $\sim$8, $\sim$9 at the operating point, absorption efficiency $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$0 for polarization parallel to the meanders, and optical NEP $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$1 in the demonstrated configuration (Nones et al., 2012).

The experimental review places such detectors in a larger system context: ground, balloon, and space platforms; the need for cold optics and low sidelobes; polarization modulators such as half-wave plates; and multi-band observations to separate synchrotron, free–free, dust, and extragalactic foregrounds (Bernardis et al., 2012). In this broader sense, nanoCMB denotes an integrated technological trajectory in which thin-film sensor physics, multiplexed readout, cryogenic optics, and polarimetric control are all directed toward sub-$$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$2 CMB signals.

7. Fundamental-physics applications and anisotropic birefringence

A further extension of the nanoCMB concept in the supplied literature concerns anisotropic cosmic birefringence. The 2024 forecast paper models a Northern Hemisphere mid-latitude CMB polarization experiment and is described in the provided material as “effectively a nanoCMB-like forecast” because it maps closely onto the science case of a Northern Hemisphere CMB polarization facility (Zhong et al., 2024). The observable is a rotation field

$$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$3

with the anisotropic component treated as a statistically isotropic Gaussian random field with power spectrum $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$4. Under a scale-invariant assumption,

$$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$5

The analysis uses a quadratic estimator on simulated polarization data to reconstruct anisotropic cosmic birefringence and forecast constraints on $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$6 (Zhong et al., 2024).

For a wide winter scan with effective sky coverage of approximately $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$7, the reported constraints depend strongly on aperture and multi-frequency combination strategy. With a small-aperture telescope operating at 95/150 GHz, the $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$8 upper bound for $$C_\ell^{XY} = 4\pi \int d\ln k\;\mathcal{P}_\mathcal{R}(k)\,\Delta_\ell^X(k)\,\Delta_\ell^Y(k), \qquad X,Y\in\{T,E\},$$9 can reach 0.017 under the low-noise scenario when merging multi-frequency data in the map domain, while merging multi-frequency data in the spectrum domain tightens the limit by about 10%. A large-aperture telescope at the same bands is found to be more effective, tightening the $C_\ell^{TT}$00 upper limit to 0.0062 (Zhong et al., 2024). The supplied material interprets these results as directly relevant to a nanoCMB deployment at a similar Northern Hemisphere site, with a key lesson that large-aperture, high-resolution polarization information is particularly powerful for anisotropic birefringence reconstruction.

In theoretical terms, this science target is linked to Chern–Simons-type couplings of the form

$C_\ell^{TT}$01

which induce polarization rotation and hence off-diagonal $C_\ell^{TT}$02 correlations. The appearance of this topic in the nanoCMB literature indicates that the term is not confined to software minimalism or detector nanofabrication; it also denotes a science program focused on Lorentz- and CPT-sensitive observables in CMB polarization (Zhong et al., 2024).

8. Interpretation, coherence, and misconceptions

Because the term appears in multiple contexts, a common misconception is to treat nanoCMB as referring only to one software repository or only to one detector concept. The supplied literature supports a more differentiated interpretation. Strictly defined, nanoCMB is the Python Boltzmann code of Moss, publicly available as nanocmb.py and designed to compute unlensed scalar $C_\ell^{TT}$03, $C_\ell^{TT}$04, and $C_\ell^{TT}$05 spectra in flat $C_\ell^{TT}$06CDM (Moss, 26 Feb 2026). Contextually, however, the same term functions as a shorthand for a minimal-yet-complete CMB research program spanning theoretical calculation, detector nanophysics, and precision polarization cosmology (Nones et al., 2012, Bernardis et al., 2012, Zhong et al., 2024).

Another misconception is to infer that “minimal” implies physically crude. The software paper argues the opposite: nanoCMB includes the full line-of-sight machinery, RECFAST recombination, tight coupling, and physically optimized grids, while maintaining sub-percent agreement with CAMB (Moss, 26 Feb 2026). Similarly, the detector paper shows that architectural simplification can coexist with nontrivial condensed-matter and optical design, as in the use of NbSi thin films, high-impedance TES operation, and direct electronic absorption (Nones et al., 2012).

A final misconception is that nanoCMB, in any of its senses, is already synonymous with a fully comprehensive CMB analysis ecosystem. The evidence in the supplied papers does not support that. The code does not include lensing or tensors (Moss, 26 Feb 2026). The detector architecture is a proof-of-concept whose present NEP is still above the most demanding future space-mission targets (Nones et al., 2012). The birefringence forecast is based on simulations that do not explicitly include beam systematics, polarization angle miscalibration, or time-domain filtering (Zhong et al., 2024). The coherence of the nanoCMB concept therefore lies in transparent, extensible minimality, not in exhaustive closure of every experimental and theoretical challenge.