NAMD 2.15 Alpha, x86 AVX-512 Test Builds

This page contains special developmental early "alpha" test builds of NAMD 2.15 intended to provide very early testing access to the NAMD user and developer communities.

STMV benchmark Frontera scaling

The new NAMD 2.15 AVX-512 feature provides up to 1.8x the
performance of prior versions on x86 hardware.

This alpha NAMD 2.15 build for AVX-512 can speed up overall simulation performance by up to 1.8x on CPUs that support AVX-512F instructions, such as Skylake, Cascade Lake, Ice Lake, and Tiger Lake. This version of NAMD is already available on the TACC Frontera supercomputer. Please be aware that the best scaling on computers with recent Infiniband networks like Frontera requires building NAMD with UCX-enabled Charm++.

These AVX-512 optimizations in NAMD speed up the CPU non-bonded calculation, so cannot be used together with GPU acceleration. The new "tiles" algorithm is a port of the existing CUDA kernels for short-range non-bonded force calculation to AVX-512, implemented by Michael Brown from Intel.

What MD Simulations Are Well-Suited to NAMD 2.15 AVX-512 Alpha Versions?

Since the AVX-512 tiles optimization is a port of CUDA algorithms, its use has most of the same feature restrictions as CUDA-accelerated NAMD. Generally speaking, simulation methodologies that depend on modification to the standard non-bonded interaction potentials are not supported, such as alchemical free energy methods, locally enhanced sampling, pressure profile calculation, and Drude force field when using the screened Coulomb correction of Thole.

Use of AVX-512 tiles should be compatible with all other methodologies, including basic dynamics, minimization, and most enhanced sampling methods including those implemented through the Colvars module.

NAMD 2.15 AVX-512 downloads

This code is still evolving and test builds will be updated as needed. Stay tuned!

For more information about NAMD and support inquiries:

For general NAMD information, see the main NAMD home page http://www.ks.uiuc.edu/Research/namd/

For your convenience, NAMD has been ported to and will be installed on the machines at the NSF-sponsored national supercomputing centers. If you are planning substantial simulation work of an academic nature you should apply for these resources. Benchmarks for your proposal are available at http://www.ks.uiuc.edu/Research/namd/performance.html

The Theoretical and Computational Biophysics Group encourages NAMD users to be closely involved in the development process through reporting bugs, contributing fixes, periodical surveys and via other means. Questions or comments may be directed to namd@ks.uiuc.edu.

We are eager to hear from you, and thank you for using our software!