Opening this to track a perf regression shown on the plots page in jenkins. This perf regression is a bit of an odd one and I suspect not code related but may be worth investigating to understand the root cause.

The reason I think its not perf related is that multiple version all show the same regression on the same date. March 22nd. Suggesting a potential change to the perf machine.

I queried Luke C to see if there was a known change at that date but he has stated that he wasn't aware of one.