Official Resources
- Homepage: https://www.acellera.com/acemd
- Documentation: https://software.acellera.com/acemd/
- License: Commercial (free for academics)
Overview
ACEMD is a high-performance molecular dynamics engine specifically designed for GPU acceleration. Built on OpenMM, it provides an optimized and user-friendly interface for biomolecular simulations with exceptional speed on NVIDIA GPUs.
Scientific domain: Biomolecular simulations, drug discovery, GPU-accelerated MD
Target user community: Pharmaceutical researchers, computational biologists
Theoretical Methods
- Classical molecular dynamics
- Langevin dynamics
- Multiple force fields (AMBER, CHARMM)
- PME electrostatics
- Implicit solvent (GBSA)
- Replica exchange
Capabilities (CRITICAL)
- Ultra-fast GPU MD simulations
- AMBER/CHARMM force field support
- Implicit and explicit solvent
- Replica exchange MD
- Metadynamics integration
- HTMD workflow integration
Key Strengths
GPU Performance:
- Optimized for NVIDIA GPUs
- Microsecond timescales routine
- Multi-GPU support
- Exceptional speed
Integration:
- HTMD workflow
- PlayMolecule platform
- Automated setup
Inputs & Outputs
-
Input formats:
- PDB structures
- AMBER prmtop
- CHARMM PSF
-
Output data types:
- XTC trajectories
- DCD trajectories
- Restart files
Interfaces & Ecosystem
- HTMD: High-throughput MD
- PlayMolecule: Web platform
- OpenMM: Backend engine
Advanced Features
- GPU optimization: CUDA-optimized kernels
- Metadynamics: Enhanced sampling
- Replica exchange: REMD support
- Adaptive sampling: With HTMD
- Free energy: Alchemical methods
Performance Characteristics
- Among fastest GPU MD codes
- Optimized memory usage
- Excellent for long simulations
- Multi-GPU scaling
Computational Cost
- GPU provides 100x+ speedup
- Microseconds per day achievable
- Efficient for large systems
- Overall: Industry-leading GPU performance
Best Practices
- Use latest NVIDIA GPUs
- Enable mixed precision
- Use HTMD for workflows
- Validate force field choice
Limitations & Known Constraints
- Commercial license
- NVIDIA GPU required
- Less flexible than OpenMM
- Biomolecular focus
Application Areas
- Drug discovery
- Protein dynamics
- Membrane simulations
- Long timescale dynamics
- Adaptive sampling
Comparison with Other Codes
- vs OpenMM: ACEMD optimized/streamlined, OpenMM more flexible
- vs GROMACS: ACEMD faster single-GPU, GROMACS better multi-node
- vs Desmond: Both commercial GPU-focused, Desmond in Schrödinger ecosystem
- Unique strength: Extreme GPU speed, HTMD integration, adaptive sampling workflows
Community and Support
- Commercial support (Acellera)
- Documentation
- Tutorials
- Email support
Verification & Sources
Primary sources:
- Website: https://www.acellera.com/acemd
- M. Harvey et al., J. Chem. Theory Comput. 5, 1632 (2009)
- M. Harvey & G. De Fabritiis, J. Chem. Theory Comput. 5, 2371 (2009)
Secondary sources:
- HTMD documentation
- PlayMolecule tutorials
- Published drug discovery applications
Confidence: VERIFIED
Verification status: ✅ VERIFIED
- Commercial with academic license
- Academic citations: >500
- Active development: Acellera
- Industry adoption: Pharmaceutical companies