diff --git a/careers.rst b/careers.rst index 5d9ca5e..ec9e851 100644 --- a/careers.rst +++ b/careers.rst @@ -206,6 +206,27 @@ They usually have strong skill overlaps with Network Engineers (with so much modern storage being network-attached in some manner) and usually have strong skills in capacity planning and performance management of infrastructure. +High Performance Computing Engineer +=================================== + +High Performance Computing (HPC) involves large scale computing infrastructure, +which is at the cutting edge of presently available technology, oftentimes making it at TOP500.ORG + +Typically, HPC platforms are used for scientific computing, big data, complex models +and may have applications in fields as diverse as physics, finance, medicine, defence, economics etc. +for instance: Meteorology and Climate prediction may appear first as a tiny slice of the bigger picture +yet they tend to be applicable to several aspects of human life and scientific efforts to improve it. + +HPC engineers are expected to master an array of high-end technologies in fields such as +networking (InfiniBand, multi-Gigabit Ethernet), computing (several computer architectures), +parallel storage/filesystems/filers (Lustre, GPFS, Isilon, NetApp, PanaSAS), as well as +be able to give advice on a number of software components (gnu/intel/pgi compilers, +debuggers, mpi stacks, FFTW, linear algebra and other optimized math libraries etc etc). + +Most importantly, HPC engineers should be able to interface with other operations engineers, +each a specialist in their own field, in order to let all systems run both at top performance +and within the maximum range of their reliability envelope, since HPC downtimes come at a high cost. + .. _how-to-become-an-operations-engineer: ************************************