As a result of the huge performance potential of multi-core microprocessors, HPC infrastructures are rapidly integrating them into their architectures in order to expedite the performance growth of the next generation HPC systems. However, as the number of cores per processor increase to 100 or 1000s of cores, they are posing revolutionary challenges to the various aspects of the software stack. In our research, we endeavor to investigate novel solutions to the problem of extracting high-performance.
In this paper, we advocate for the use of virtualization as an alternative approach to the traditional operating systems for the next generation multicore-based HPC systems. In particular, we investigate an efficient mechanism for shared-memory communication between HPC applications executing within virtual machine (VM) instances that are co-located on the same hardware platform. This system, called Vshmem, implements low latency IPC communication mechanism that allows the programmer to selectively share memory regions between user-space processes residing in collocated virtual machines. Our contributions addressed