March 2011: Major cluster expansion!

Welcome to TARDIS cluster web page

TARDIS is a CPU/GPU computational cluster used for numerical simulations of soft matter systems by members of Olvera de la Cruz Research Group in the Materials Science and Engineering Department at Northwestern University. TARDIS cluster was built by Atipa.



TARDIS cluster has a total of 576 CPU cores spread over 36 computational nodes with 36 Fermi graphics cards, one in each node. Nodes are connected through the low-latency Infiniband network making it suitable for running large scale, communication intensive applications like molecular dynamics simulations or linear algebra calculations (e.g., in differential equation solvers).

A wide range of standard simulation tools are available, including molecular dynamics codes like LAMMPS, HOOMD, etc.

What makes TARDIS distinct from other clusters?

Its unique design allows concurrent execution of conventional highly parallel CPU codes and novel GPU programs in a homogeneous environment and even within the same queuing system. Users can utilize the most suitable approach for a problem at hand within a uniform setting, without a necessity to have their data spread over a number of machines and having to deal with different host-specific set ups.


First and foremost our thanks go to Prof. Monica Olvera de la Cruz, whose inspiring mind guides the research of the whole group, far beyond limits of this computer cluster.

We also thank Atipa team who built and installed TARDIS. Especially to Bart Willems, Tim Obiefule, and Dana Chang.

This web presentation is a joint effort of Doris Grillo, Peijun Guo, Prateek Jha, and Rastko Sknepnek. Rastko acts as a system administrator. He has also installed most of the software and set up and tweaked the SGE queuing system.

Special thanks go to Peijun Guo, Vikram Jadhao, Prateek Jha and Ting Li for helping with the cluster assembly.

TARDIS cluster is housed by the Northwestern University Information Technology services.

The funding from the TARDIS cluster is provide by the U.S. Department of Defense National Security Science and Engineering Faculty Fellowship (NSSEFF) awarded to Prof. Olvera de la Cruz.

Categories :