To be announced
Machine Learning and AI programming on the Micron Hybrid Memory Cube and FPGAs
Wednesday, February 15, 2017, 1PM-4PM
The Science Center, 4th Floor, The Graduate Center
Both Intel and Micron are developing systems incorporating 3D Crosspoint (3DXP) and FPGAs for machine learning and artificial intelligence. These systems have very large memory specifically designed for high-speed processing of unstructured semantic data for machine learning and AI applications as well as other applications (including computational intensive applications). This technology has been percolating for a few years and is now at the point of market readiness.
In this workshop, Micron will present on its module-based architecture —Hybrid Memory Cube, FPGAs, and X86 processors -- that provides extremely high-memory bandwidth and provides super high-performance for classification, ML and AI applications.
Topics to be covered include:
- System architecture
- Application space
Micron will provide programming instruction and demonstrate its OpenCL programming development flow for the FPGAs incorporated into the architecture.
A number of universities (Georgia Tech, U of Missouri, U of Virginia) are actively engaged with Micron in exploiting the capabilities of this class of devices. There may be an opportunity for CUNY researchers to become similarly engaged, if there is sufficient research interest. A prototype system from Micron is installed in the CUNY HPC Center and is network available for evaluation.
Seats are limited and will be filled on a first come basis