Official Resources
- Homepage: https://github.com/ETHDMFT/NRG
- Documentation: GitHub repository documentation
- Source Repository: https://github.com/ETHDMFT/NRG
- License: Open-source (check repository for specific license)
Overview
NRG is a numerical renormalization group implementation developed at ETH Zurich for solving quantum impurity problems within dynamical mean-field theory (DMFT). While primarily an impurity solver rather than a ground-state DFT code, NRG is used within DMFT+DFT frameworks to treat strong correlations in materials. It provides highly accurate solutions to Anderson impurity models, which are central to DMFT calculations of correlated electron systems.
Scientific domain: Quantum impurity models, DMFT impurity solver, strongly correlated systems
Target user community: DMFT researchers, strongly correlated materials scientists
Theoretical Methods
- Numerical Renormalization Group (NRG)
- Anderson impurity model
- Quantum impurity problems
- Dynamical Mean-Field Theory (DMFT) solver
- Strongly correlated electron systems
- Spectral functions
- Real-frequency calculations
- Zero-temperature and finite-temperature
Capabilities (CRITICAL)
Category: Open-source impurity solver
Note: Impurity solver for DMFT, not standalone DFT code
- Anderson impurity model solutions
- NRG impurity solver
- DMFT integration
- Spectral function calculations
- Real-frequency results
- High accuracy for static properties
- Quantum phase transitions
- Kondo physics
- Local Green's functions
- Self-energy calculations
Sources: GitHub repository (ETH Zurich)
Key Strengths
NRG Accuracy:
- Highly accurate for ground state
- Excellent for low-energy physics
- Kondo physics specialist
- Quantum phase transitions
- Zero-temperature properties
DMFT Integration:
- Impurity solver for DMFT
- Self-consistent DMFT loops
- Real-frequency calculations
- Strongly correlated materials
- DFT+DMFT frameworks
ETH Development:
- Research-quality code
- Active development
- Open-source
- Academic support
- Community contributions
Inputs & Outputs
-
Input formats:
- Anderson impurity parameters
- Bath discretization
- DMFT self-consistency data
- NRG-specific settings
-
Output data types:
- Impurity Green's functions
- Self-energies
- Spectral functions
- Local observables
- Ground state properties
Interfaces & Ecosystem
-
DMFT Frameworks:
- Integration with DMFT codes
- Triqs potential compatibility
- DFT+DMFT workflows
- w2dynamics connections
-
Related Codes:
- DFT codes (upstream)
- DMFT frameworks
- Impurity solver suite
Workflow and Usage
DMFT Impurity Solver:
# Within DMFT loop
# Solve Anderson impurity model with NRG
nrg_solver.solve(impurity_parameters)
# Obtain impurity Green's function
G_imp = nrg_solver.get_greens_function()
Typical DMFT Workflow:
- DFT calculation (starting point)
- Extract correlated orbitals
- DMFT self-consistency loop:
- Construct impurity problem
- Solve with NRG
- Update self-energy
- Check convergence
- Analyze results
Advanced Features
NRG Algorithm:
- Logarithmic discretization
- Iterative diagonalization
- Wilson's NRG method
- Low-energy accuracy
- Systematic improvements
Spectral Functions:
- Real-frequency results
- High resolution
- Kondo resonances
- Hubbard bands
- Density of states
Correlation Physics:
- Strong correlations
- Kondo effect
- Mott transitions
- Quantum criticality
- Heavy fermions
Performance Characteristics
- Speed: Moderate (iterative NRG)
- Accuracy: Excellent (especially low-energy)
- System size: Single impurity site
- Purpose: DMFT impurity solver
- Typical: Part of DMFT workflow
Computational Cost
- Reasonable for impurity problems
- More expensive than QMC for some properties
- Excellent accuracy/cost for ground state
- Suitable for production DMFT
Limitations & Known Constraints
- Purpose: Impurity solver, not standalone DFT
- High-energy: Limited high-frequency accuracy
- Real-time: Static/equilibrium focus
- Learning curve: NRG methodology expertise
- DMFT required: Must be part of DMFT framework
- Not ground-state DFT: Solves impurity models only
Comparison with Other Impurity Solvers
- vs CT-QMC: NRG better for ground state, QMC for dynamics
- vs ED: NRG handles larger baths
- vs CTMQC: NRG real-frequency, QMC Matsubara
- Unique strength: Low-energy accuracy, real frequencies, Kondo physics
Application Areas
DMFT Calculations:
- Strongly correlated materials
- Mott insulators
- Heavy fermions
- Kondo lattices
- DFT+DMFT studies
Impurity Physics:
- Quantum impurities
- Anderson models
- Kondo problem
- Quantum dots
- Local moment physics
Spectroscopy:
- Photoemission spectra
- Local density of states
- Spectral functions
- Low-energy properties
Best Practices
DMFT Integration:
- Careful bath discretization
- Converge NRG parameters
- Check frequency coverage
- Validate with experiments
- Compare with other solvers
NRG Expertise:
- Understand NRG methodology
- Proper energy scales
- Discretization parameters
- Interpretation of results
Community and Support
- Open-source (GitHub)
- ETH Zurich support
- DMFT community
- Academic collaboration
- Research code
Educational Resources
- GitHub documentation
- NRG literature
- DMFT tutorials
- ETH publications
- Quantum impurity theory
Development
- ETH Zurich
- Active maintenance
- Community contributions
- Open development
- Research focus
Important Note
NRG is an impurity solver for DMFT, not a standalone ground-state DFT code. It must be used within a DMFT framework, which itself interfaces with DFT codes. The workflow is: DFT → DMFT framework → NRG impurity solver → Back to DMFT → Results.
Verification & Sources
Primary sources:
- GitHub: https://github.com/ETHDMFT/NRG
- ETH Zurich DMFT group
- Repository documentation
Secondary sources:
- NRG methodology papers
- DMFT literature
- Quantum impurity theory
- Wilson's NRG papers
Confidence: VERIFIED - Open-source impurity solver
Verification status: ✅ VERIFIED
- GitHub: ACCESSIBLE
- Institution: ETH Zurich
- License: Open-source
- Purpose: DMFT impurity solver (not standalone DFT)
- Category: Open-source DMFT tool
- Status: Maintained
- Specialized strength: NRG impurity solver for DMFT, low-energy accuracy, real-frequency spectral functions, Kondo physics, ETH research code