droneb
Well-Known Member
- Joined
- Aug 28, 2009
- Messages
- 1,072
then i would suggest a minimal environment for these kinds of test, on a raw state the processor and data busses will be dedicated on the sole purpose of tx/rx the nfo so if it archieves the expected speed the we get a baseline (the chip and its connection works), then we need gradual testing in order to check if something else is lowering the performance, or is it that we get a processor or data I/O and or any other bottle necks.
i really recommend MW to get back to the minimal environment for the tests, instead of adding variables to the testing like having a processor heavy app like the navigator or any aplications doing I/O access to any memory involved.
i really recommend MW to get back to the minimal environment for the tests, instead of adding variables to the testing like having a processor heavy app like the navigator or any aplications doing I/O access to any memory involved.