docs: Add details and examples on optimal GPU setup (#2531)

* docs for GPU optimizations

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* improve time reporting and improve execution

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* fix standard pipeline

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* tune examples with batch size 64

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* add benchmark results

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* improve docs

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* typo in excluded tests

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

* explicit pipeline in table

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

---------

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>