I have many options to play around with hardware like having more than one GPU per machine or upgrading CPU's but it doesn't reflect in the Measured floating point speed or the Measured integer speed on each machine. Is there a problem? I do check my GPU's and CPU's and due to the loading I have had to throttle usage back to 80% just to not stress my systems. they are crunching. Being a Newbie what am I missing here?

Each project updates their stats on their own periodic schedule. MW last updated their stats 2021-10-12 15:37.
https://milkyway.cs.rpi.edu/milkyway/stats/

The benchmarks that BOINC runs are based on rather old Whetstone and Dhrystone tests from a time when a cpu only had one core and rather slow clock speed. You can't correlate the benchmark results from BOINC from the actual performance on current hardware and applications.

You have to figure out what kind of loading you can put on your hardware by experiment and observation on each project's applications. Or ask in the forums from people with similar hardware what their configuration settings are set for.

---

Thanks for the info, thought something like that, looks like I have to do my performance tweaks as I see it not via Milkyway@home.