sPIN: High-Performance Streaming Processing in the
Network
SessionPerformance Tuning
Event Type
Paper
TimeWednesday, November 15th4:30pm -
5pm
Location301-302-303
DescriptionOptimizing communication performance is imperative for
large-scale computing because communication overheads
limit the strong scalability of parallel applications.
Today’s network cards contain rather powerful processors
optimized for data movement. However, these devices are
limited to fixed functions, such as remote direct memory
access. We develop sPIN, a portable programming model to
offload simple packet processing functions to the
network card. To demonstrate the potential of the model,
we design a cycle-accurate simulation environment by
combining the network simulator LogGOPSim and the CPU
simulator gem5. We implement offloaded message matching,
datatype processing, and collective communications and
demonstrate full-application speedups of up to 3.7%.
Furthermore, we show how sPIN can be used to accelerate
redundant in-memory filesystems. Our work investigates a
portable packet-processing network acceleration model
similar to compute acceleration with CUDA or OpenCL. We
show how such network acceleration enables an eco-system
that can significantly speed up applications and system
services.
Download PDF:
here
