I can't correlate this to anything, but I am seeing an increase and staying high. Times somewhat match on a couple metrics.
CPU spike 11:45-12 sept 6 and then again 15:45
memory spike 16:14 sept 6
garbage collection 11:45 sept 6 then again at 15:45
garbage activations 12:30 and again 17:45
Have you check the frontend logs in the system information dashlet? If you hit the drop down next to the server instance name -> support information -> then log files, you should see multiple frontend server logs, the most recent one should cover this timeframe and show any error that could be causing this.
Another point of investigation would be a spike in users of the client/requests from users of the environment. If you check the server health dashboard you will see at the top right a chart that shows the number of clients.
Also, you could use the audit logs to investigate the dashboards that were created on that day. If someone created a dashboard that pulls in massive amounts of data (many dashlets) and has a low auto refresh rate, it could cause frontend server issues as it is responsible for servicing client requests.