Software Downloads

Which version of NAMD should I download?

The versions of NAMD below are distinguished first by OS, followed by the type of network interface, and whether or not CUDA is supported. If you are installing NAMD on a standalone workstation, we recommend downloading Linux-x86_64-multicore for Linux. If your workstation has a CUDA-capable GPU, you should try downloading Linux-x86_64-multicore-CUDA. If you wish to run multi-copy algorithms, such as replica-exchange MD, you should try the "netlrts" builds, such as Linux-x86_64-netlrts-smp or Linux-x86_64-netlrts-smp-CUDA. Windows users are encouraged to install WSL (Windows for Linux Subsystem) in order to run our most recent Linux builds.

Point release 3.0.2 fixes bugs in the Colvars module that affect not only NAMD 3.0 and 3.0.1, but also earlier NAMD 2 versions.

See announce.txt for details regarding correctness issues in: eigenvector, orientationAngle, orientation, both coordNum and selfCoordNum with pairlist, and distanceZ with dynamic axis. For more information about the bugs in Colvars, please see Colvars Bugfix Announcement.

We encourage all NAMD users to upgrade to version 3.0.2.

Download NAMD:

NAMD is a parallel, object-oriented molecular dynamics code designed for high-performance simulation of large biomolecular systems. Simulation preparation and analysis is integrated into the visualization package VMD. Visit the NAMD website for complete information and documentation.

Selecting an archive below will lead to a user registration and login page. Your download will continue after you have registered or logged in.

Version Nightly Build (2025年10月14日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
MacOS-universal-multicore (Apple silicon and Intel-based)
Source Code

Version 3.0.2 (2025年08月27日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOS-universal-multicore (Apple silicon and Intel-based)
Win64-multicore (Windows 11)
Win64-multicore-CUDA (NVIDIA CUDA acceleration)
Win64-netlrts-smp (Windows 11, multi-copy)
Win64-netlrts-smp-CUDA (NVIDIA CUDA acceleration, multi-copy)
Source Code

Version 3.0.1 (2024年10月14日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOS-universal-multicore (Apple silicon and Intel-based)
Source Code

Version 3.0 (2024年06月14日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOS-universal-multicore (Apple silicon and Intel-based)
Source Code

Version 3.0b7 (2024年05月10日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOS-universal-multicore (Apple silicon and Intel-based)
Source Code

Version 3.0b6 (2024年01月20日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOS-universal-multicore (Apple silicon and Intel-based)
Source Code

Version 2.15 ALPHA Release (2023年12月18日) Platforms:

Linux-x86_64-multicore-AMDHIP (AMD HIP/ROCm acceleration)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Source Code for Intel GPU (Intel oneAPI/SYCL for Intel GPU acceleration)

Version 3.0b3 (2023年04月01日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-AVX512 (x86_64 AVX-512)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts-smp (Multi-copy algorithms, single process per copy)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
Source Code

Version 2.14 (2020年08月05日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-netlrts (Multi-copy algorithms, single host or ethernet)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-KNL-multicore (Intel Xeon Phi KNL processor single node)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86_64-CUDA (NVIDIA CUDA acceleration)
MacOSX-x86_64-netlrts (Multi-copy algorithms)
Win64 (Windows 7, 8, 10, etc.)
Win64-CUDA (NVIDIA CUDA acceleration)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Win64-MPI-smp-CUDA (HPC Server with CUDA)
Source Code

Version 2.13 (2018年11月09日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-smp-CUDA (NVIDIA CUDA with InfiniBand)
Linux-x86_64-netlrts (Multi-copy algorithms, single host or ethernet)
Linux-x86_64-netlrts-smp-CUDA (Multi-copy algorithms, single process per copy)
Linux-x86_64-verbs (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-KNL-multicore (Intel Xeon Phi KNL processor single node)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86_64-CUDA (NVIDIA CUDA acceleration)
MacOSX-x86_64-netlrts (Multi-copy algorithms)
Win64 (Windows 7, 8, 10, etc.)
Win64-CUDA (NVIDIA CUDA acceleration)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Win64-MPI-smp-CUDA (HPC Server with CUDA)
Source Code

Version 2.12 (2016年12月22日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-smp-CUDA (NVIDIA CUDA with InfiniBand)
Linux-x86_64-netlrts (Multi-copy algorithms, single host or ethernet)
Linux-x86_64-verbs (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-KNL-multicore (Intel Xeon Phi KNL processor single node)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86_64-CUDA (NVIDIA CUDA acceleration)
MacOSX-x86_64-netlrts (Multi-copy algorithms)
Win32 (Windows XP, etc.)
Win64 (Windows 7, 8, 10, etc.)
Win64-CUDA (NVIDIA CUDA acceleration)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Win64-MPI-smp-CUDA (HPC Server with CUDA)
Source Code

Version 2.11 (2015年12月22日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-smp-CUDA (NVIDIA CUDA with InfiniBand)
Linux-x86_64-multicore-MIC (Intel Xeon Phi coprocessor acceleration)
Linux-x86_64-netlrts (Multi-copy algorithms, single host or ethernet)
Linux-x86_64-verbs (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp (InfiniBand, no MPI needed, supports multi-copy algorithms)
Linux-x86_64-verbs-smp-CUDA (InfiniBand, no MPI needed, supports multi-copy algorithms)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86_64-CUDA (NVIDIA CUDA acceleration)
MacOSX-x86_64-netlrts (Multi-copy algorithms)
Win32 (Windows XP, etc.)
Win64 (Windows 7, 8, 10, etc.)
Win64-CUDA (NVIDIA CUDA acceleration)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Win64-MPI-smp-CUDA (HPC Server with CUDA)
Source Code

Version 2.10 (2014年12月11日) Platforms:

Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-smp-CUDA (NVIDIA CUDA with InfiniBand)
Linux-x86_64-multicore-MIC (Intel Xeon Phi coprocessor acceleration)
Linux-x86_64-netlrts (Multi-copy algorithms, single host or ethernet)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86_64-netlrts (Multi-copy algorithms)
Win64 (Windows 7, 8, 10, etc.)
Win64-CUDA (NVIDIA CUDA acceleration)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Source Code

Version 2.9 (2012年04月30日) Platforms:

Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-x86-TCP (TCP may be better on gigabit)
Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-smp-CUDA (NVIDIA CUDA with InfiniBand)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-x86 (Mac OS X for Intel processors, fails on 10.7 "Lion")
Solaris-x86_64
Win32 (Windows XP, etc.)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Source Code

Version 2.8 (2011年05月31日) Platforms:

AIX-POWER-lapi (IBM POWER clusters)
AIX-POWER-multicore (IBM POWER single node)
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-x86-TCP (TCP may be better on gigabit)
Linux-x86_64-multicore (64-bit Intel/AMD single node)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
Linux-x86_64-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-CUDA (NVIDIA CUDA with InfiniBand)
MacOSX-x86_64 (Mac OS X for 64-bit Intel processors)
MacOSX-PPC (Mac OS X for PowerPC)
MacOSX-x86 (Mac OS X for Intel processors, fails on 10.7 "Lion")
Solaris-x86_64
Win32 (Windows XP, etc.)
Win64-MPI (Windows HPC Server, multi-copy algorithms)
Source Code

Version 2.7 (2010年10月15日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-x86-TCP (TCP may be better on gigabit)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, not for Omni-Path, no MPI needed)
Linux-x86_64-CUDA (NVIDIA CUDA acceleration)
Linux-x86_64-ibverbs-CUDA (NVIDIA CUDA with InfiniBand)
Linux-Itanium-Altix (original SGI Altix, not Altix UV)
MacOSX-PPC (Mac OS X for PowerPC)
MacOSX-x86 (Mac OS X for Intel processors, fails on 10.7 "Lion")
Solaris-x86_64
Win32 (Windows XP, etc.)
Source Code

Version 2.6 (2006年08月31日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
BlueGeneL (bypasses MPI for better scaling)
BlueGeneP-MPI (Blue Gene/P)
Linux-amd64-Clustermatic5-TCP (Clustermatic 5)
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-x86-TCP (TCP may be better on gigabit)
Linux-x86_64 (64-bit Intel/AMD with ethernet)
Linux-x86_64-TCP (TCP may be better on gigabit)
Linux-i686-Clustermatic4 (Clustermatic 4 or 5)
Linux-i686-Clustermatic4-TCP (TCP may be better on gigabit)
Linux-i686-Scyld29 (Scyld Beowulf 29)
Linux-i686-Scyld29-TCP (TCP may be better on gigabit)
Linux-Itanium (Itanium)
Linux-Itanium-Altix (original SGI Altix, not Altix UV)
MacOSX-PPC-xlC (Mac OS X for PowerPC, needs IBM libraries)
MacOSX-x86 (Mac OS X for Intel processors, fails on 10.7 "Lion")
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
Tru64-Alpha
Tru64-Alpha-Elan (AlphaServer SC with Quadrics)
Tru64-Alpha-MPI (HP MPI, no Quadrics or Elan)
Win32 (Windows XP, etc.)
Source Code

Version 2.5 (2003年09月29日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
HPUX
HPUX-PA2.0
Linux-Alpha
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-x86-TCP (TCP may be better on gigabit)
Linux-i686-Clustermatic (Clustermatic 3)
Linux-i686-Clustermatic-TCP (TCP may be better on gigabit)
Linux-i686-Scyld (Scyld Beowulf)
Linux-i686-Scyld-TCP (TCP may be better on gigabit)
MacOSX-PPC (Mac OS X for PowerPC)
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
Solaris-Sparc-MPI (Sun HPC ClusterTools)
T3E
Tru64-Alpha
Tru64-Alpha-Elan (AlphaServer SC with Quadrics)
Tru64-Alpha-MPI (HP MPI, no Quadrics or Elan)
Win32 (Windows XP, etc.)
Source Code

Version 2.4 (2002年03月11日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
HPUX
HPUX-PA2.0
Linux-Alpha
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-i686-Scyld (Scyld Beowulf)
MacOSX-PPC (Mac OS X for PowerPC)
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
Solaris-Sparc-MPI (Sun HPC ClusterTools)
T3E
Tru64-Alpha
Tru64-Alpha-Elan (AlphaServer SC with Quadrics)
Win32 (Windows XP, etc.)
Source Code

Version 2.3 (2001年08月02日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
HPUX
HPUX-PA2.0
Linux-Alpha
Linux-x86 (32-bit Intel/AMD with ethernet)
Linux-i686-Scyld (Scyld Beowulf)
MacOSX-PPC (Mac OS X for PowerPC)
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
Solaris-Sparc-MPI (Sun HPC ClusterTools)
T3E
Tru64-Alpha
Tru64-Alpha-Elan (AlphaServer SC with Quadrics)
Win32 (Windows XP, etc.)
Source Code

Version 2.2 (2000年09月29日) Platforms:

AIX-POWER (IBM POWER)
AIX-POWER-MPI (IBM POWER clusters)
HPUX
HPUX-PA2.0
Linux-Alpha
Linux-x86 (32-bit Intel/AMD with ethernet)
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
T3E
Tru64-Alpha
Win32 (Windows XP, etc.)
Source Code

Version 2.1 (1999年11月11日) Platforms:

AIX-POWER (IBM POWER)
Linux-x86 (32-bit Intel/AMD with ethernet)
Origin2000 (any SGI, shared-memory only)
Origin2000-MPI (SGI MPI)
Solaris-Sparc
T3E
Tru64-Alpha
Source Code