Interoperable GPU Kernels as Latency Improver for MEC.
Juuso HaavistoJukka RiekkiPublished in: 6G SUMMIT (2020)
Keyphrases
- heterogeneous computing
- data transfer
- real time
- response time
- kernel function
- digital libraries
- kernel methods
- low latency
- web services
- graphics hardware
- memory bandwidth
- metadata
- compute intensive
- prefetching
- kernel learning
- graphics processing units
- graphics processors
- multiple kernel learning
- parallel implementation
- grid computing
- parallel computation
- parallel processing
- data access
- feature space
- support vector
- multiple kernel
- gpu accelerated
- parallel computing
- gpu implementation
- positive definite
- high dimensional
- feature selection