US Army buys $12 million AI-focused “HPC in a Container” from IBM
Written by James Orme Mon 5 Aug 2019

IBM will deliver the mobile supercomputer late 2019
The US Department of Defense (DoD) has bought a containerised supercomputer from IBM that will be deployed at the “tactical edge” and other remote locations.
The $12 million “HPC in a Container” will serve all of the DoD’s agencies and services when it is delivered late 2019 and will initially be based at a US Army supercomputing research laboratory at Aberdeen Proving Ground, Maryland.
The containerised facility will consist of a HPC, capable of six petaFLOPs of single precision performance, housed in an ordinary shipping container, complete with a UPS, chilled water cooling and fire suppression system.
Sally Parsons, Huntsville Center Information Technology Systems Division chief, said the supercomputer will be primarily used for machine learning and data analytics workloads not previously possible with HPCs installed in fixed locations.
To support ML tasks, the system will boast 22 dedicated ML nodes, each with two IBM Power9 processors, 512GB of system memory, 8 Nvidia V100 GPUs with 32GB of high-bandwidth memory, and 15TB of local solid-state storage. It will also support popular machine learning frameworks TensorFlow, PyTorc and Caffe.
“This is a singularly important achievement for its end-users,” Parsons said.
“Because of the sensitive nature of the work involved, we here at Huntsville Center will never know exactly the solutions this unique tool will provide in the field, but I am quite confident that our work will result in both lives saved and problems avoided.”
Earlier this year, the DoD’s CIO, Dana Deasy, revealed the military is exploring ways to use cloud computing and edge infrastructure to deliver powerful applications to fighters in remote locations.
“Imagine a world where we can take that compute power with new applications on top of it, and put the cloud right into the hands of the tactical fighter on the edge,” Deasy said. “That’s why the cloud is so important to us.”
Written by James Orme Mon 5 Aug 2019