This talk, by Michele Weiland, was given at the Intel Collaboration Hub at ISC-HPC 2016 in Frankfurt. The talk outlines how the project is using the new 3D XPoint NVRAM technology to address the key issue of I/O bottlenecking in the design of a new system working towards Exascale computing.
Slides from this talk are available below.
Supporting Documents:
Addressing the I/O Bottleneck 339.3 Kb [5 download(s)]