I’m banging my head on the wall with the total lack of clarity on the FP64/FP32 throughput ratio for the models listed:
I already know all the other specs, but finding out their behaviour in DP over the internet (before actually buying any of them) is a treasure hunt. On some post here we read 1/8 somewhere on the internet 1/12, 1/24…
imho it is unacceptable that a company does not state ALL features clearly. Nor am I going to listen to “advice” not based on numbers or tests.
I’m asking owners of any of the above cards to write here their values for “single precision float” and “double precision float” so that everyone can couple this info with other specs (e.g. TDP) in order to choose the appropriate middle-end card for a home-assembled laptop (scientific computations with home-made codes).
Thanx in advance