Block

Block is a highly-optimized DMRG (Density Matrix Renormalization Group) code developed by Garnet Chan's group, designed specifically for quantum chemistry applications. The code implements state-of-the-art DMRG algorithms with a focus on…

3. DMFT & MANY-BODY 3.4 Tensor Networks VERIFIED 1 paper
Back to Mind Map Official Website

Overview

Block is a highly-optimized DMRG (Density Matrix Renormalization Group) code developed by Garnet Chan's group, designed specifically for quantum chemistry applications. The code implements state-of-the-art DMRG algorithms with a focus on ab initio quantum chemistry, providing accurate solutions for strongly correlated molecular systems. Block is known for its efficiency, scalability, and specialized features for chemical applications including spin-adaptation and point group symmetries.

Reference Papers (1)

Full Documentation

Official Resources

  • Homepage: https://sanshar.github.io/Block/
  • Documentation: https://sanshar.github.io/Block/
  • Source Repository: https://github.com/sanshar/Block
  • License: GNU General Public License v3.0

Overview

Block is a highly-optimized DMRG (Density Matrix Renormalization Group) code developed by Garnet Chan's group, designed specifically for quantum chemistry applications. The code implements state-of-the-art DMRG algorithms with a focus on ab initio quantum chemistry, providing accurate solutions for strongly correlated molecular systems. Block is known for its efficiency, scalability, and specialized features for chemical applications including spin-adaptation and point group symmetries.

Scientific domain: Quantum chemistry DMRG, strongly correlated molecules
Target user community: Quantum chemists, strongly correlated molecular systems

Theoretical Methods

  • Density Matrix Renormalization Group (DMRG)
  • Spin-adapted DMRG
  • Point group symmetries
  • Matrix Product States (MPS)
  • Matrix Product Operators (MPO)
  • Perturbative corrections (DMRG-CASPT2)
  • Response properties
  • Excited states

Capabilities (CRITICAL)

Category: Open-source quantum chemistry DMRG

  • DMRG for molecules
  • Spin-adapted formulation
  • Point group symmetries
  • Active space calculations
  • Ground and excited states
  • Multi-reference character
  • Large active spaces (40+ orbitals)
  • Parallelization (MPI + OpenMP)
  • Integration with quantum chemistry codes
  • Perturbative corrections
  • Response properties
  • Production quality

Sources: Official website, GitHub, publications

Key Strengths

Quantum Chemistry Focus:

  • Molecular systems specialist
  • Chemical accuracy
  • Large active spaces
  • Strongly correlated molecules
  • ab-initio calculations

Spin-Adaptation:

  • Exact spin eigenstates
  • Computational efficiency
  • Chemical accuracy
  • Proper quantum numbers
  • Reduced bond dimension

Symmetries:

  • Point group symmetries (D2h, etc.)
  • Computational efficiency
  • Proper state labels
  • Automated exploitation
  • Reduced cost

Performance:

  • Highly optimized C++
  • MPI + OpenMP
  • Large-scale calculations
  • HPC production
  • Efficient algorithms

Inputs & Outputs

  • Input formats:

    • FCIDUMP integrals
    • Input configuration files
    • Orbital ordering
    • Symmetry specifications
  • Output data types:

    • Energies
    • Wavefunctions (MPS)
    • Reduced density matrices
    • Observables
    • Orbital correlations

Interfaces & Ecosystem

Quantum Chemistry Codes:

  • PySCF
  • Molpro
  • Molcas/OpenMolcas
  • GAMESS
  • Q-Chem
  • FCIDUMP standard

Post-Processing:

  • Density matrices
  • Orbital entanglement
  • Correlation analysis
  • Perturbative corrections

Workflow and Usage

Installation:

# Clone repository
git clone https://github.com/sanshar/Block.git
cd Block
# Configure
./configure
make -j8

Input File (dmrg.conf):

nelec 10
spin 0
irrep 1

orbitals FCIDUMP

maxM 1000
maxiter 20
sweep_tol 1e-7

warmup
  M   100 200 400 
  end

Run DMRG:

# Serial
block.spin_adapted dmrg.conf > dmrg.out

# MPI parallel
mpirun -n 16 block.spin_adapted dmrg.conf > dmrg.out

With PySCF:

from pyscf import gto, scf, dmrgscf

mol = gto.M(
    atom = 'N 0 0 0; N 0 0 1.1',
    basis = 'ccpvdz',
    spin = 0
)

mf = scf.RHF(mol).run()

# DMRG-SCF calculation
mc = dmrgscf.DMRGSCF(mf, 8, 8)  # 8 orbitals, 8 electrons
mc.fcisolver.maxM = 1000
mc.kernel()

print("DMRG-SCF energy:", mc.e_tot)

Advanced Features

Perturbative Corrections:

  • DMRG-CASPT2
  • NEVPT2 variants
  • Dynamic correlation
  • Chemical accuracy
  • Production quality

Excited States:

  • State-averaged DMRG
  • State-specific calculations
  • Multiple states
  • Excitation energies
  • Spectroscopy

Response Properties:

  • Transition densities
  • Dipole moments
  • Response DMRG
  • Properties calculations

Orbital Optimization:

  • DMRG-SCF
  • DMRG-CASSCF
  • Orbital rotation
  • Active space optimization
  • Self-consistent calculations

Performance Characteristics

  • Speed: Highly optimized, HPC-ready
  • Accuracy: Chemical accuracy achievable
  • System size: Large active spaces (40+ orbitals)
  • Purpose: Production quantum chemistry
  • Scalability: Excellent MPI scaling

Computational Cost

  • Active space dependent
  • Bond dimension scaling
  • Efficient for quantum chemistry
  • HPC production
  • Hours to days per calculation

Limitations & Known Constraints

  • Quantum chemistry focus: Not general tensor network tool
  • 1D ordering: Orbital ordering matters
  • Active space: Limited to moderate sizes
  • Learning curve: Quantum chemistry expertise
  • 2D systems: Not designed for lattices

Comparison with Other DMRG Codes

  • vs ITensor: Block QC-specialized, ITensor general
  • vs CheMPS2: Block more established, similar focus
  • vs ORCA DMRG: Block standalone, ORCA integrated
  • Unique strength: Spin-adapted QC DMRG, large active spaces, point group symmetries, Chan group development

Application Areas

Strongly Correlated Molecules:

  • Transition metal complexes
  • Metalloproteins
  • Singlet-triplet gaps
  • Bond breaking
  • Multi-reference systems

Quantum Chemistry:

  • Active space calculations
  • CASSCF/CASPT2 alternatives
  • Large active spaces
  • Spectroscopy
  • Reaction mechanisms

Materials Chemistry:

  • Molecular magnets
  • Catalysis
  • Electronic structure
  • Properties prediction
  • Benchmark calculations

Best Practices

Active Space Selection:

  • Chemically relevant orbitals
  • Natural orbitals preferred
  • Orbital ordering optimization
  • Test smaller spaces first

DMRG Parameters:

  • Appropriate M (bond dimension)
  • Convergence criteria
  • Sweep schedule
  • Truncation tolerance
  • Warmup strategy

Symmetries:

  • Use point group symmetries
  • Spin-adaptation crucial
  • Proper irrep labels
  • Computational efficiency

Community and Support

  • Open-source (GPL v3)
  • Chan group (Caltech)
  • GitHub repository
  • Active development
  • Scientific publications
  • User community
  • Quantum chemistry focus

Educational Resources

  • Official documentation
  • Publication list
  • Example inputs
  • Quantum chemistry DMRG literature
  • User contributions
  • Workshop materials

Development

  • Garnet Chan group (Caltech)
  • Sandeep Sharma (lead developer)
  • Active research
  • Ongoing development
  • Feature additions
  • Performance optimization
  • Quantum chemistry focus

Research Impact

Block has enabled numerous quantum chemistry calculations on strongly correlated molecules with large active spaces, advancing understanding of transition metal chemistry, bond breaking, and multi-reference systems.

Verification & Sources

Primary sources:

  1. Homepage: https://sanshar.github.io/Block/
  2. GitHub: https://github.com/sanshar/Block
  3. Publications: J. Chem. Phys. 142, 034102 (2015)

Secondary sources:

  1. Quantum chemistry DMRG literature
  2. User publications
  3. DMRG review papers

Confidence: VERIFIED - Quantum chemistry DMRG code

Verification status: ✅ VERIFIED

  • Website: ACCESSIBLE
  • GitHub: ACCESSIBLE
  • License: GPL v3 (open-source)
  • Category: Open-source quantum chemistry DMRG
  • Status: Actively developed
  • Institution: Caltech (Chan group)
  • Specialized strength: Spin-adapted DMRG for quantum chemistry, large active spaces, point group symmetries, strongly correlated molecules, ab-initio calculations, perturbative corrections, HPC-optimized, production quality, chemical accuracy, Chan group development

Related Tools in 3.4 Tensor Networks