Testing Round 1

Initial Plan

1. establish a baseline by measuring performance of an HPC machine (ccu and gpu?) during pulsar observation.
2. run just the CS software by itself first and measure it's performance (in sim mode, since we don't know how to run dspsr against guppi_daq shared memory without screwing up guppi_daq)
3. then try to run them both at the same time and measure performance (with CS in sim mode, since it can't share the same Net buffer with guppi_daq right now)

Timing and Testing Comments

Results

node_exporter

memory available

GPU

memory used

temperature