Queue Benchmarks

[polybeandip]

As described by calyxir/calyx#2276, this thread is a running log of benchmarks for our hardware.

[polybeandip]

Sep 13th, 2024

We have cycle counts, synthesis results (clock period = 7ns), and graphs comparing our two round-robin and strict implementations.

For the graphs below, we compute total_time with formula

total_time = cycle_count * (1000/(7 - worst_slack))

All this data comes from calyxir/calyx#2276.