Even with molecular sieve storage you could be disappointed in the water peak amplitude.
It turns out NMR tube glass traps H2O via interstitial bonding sites. Neither heat nor molecular sieves remove this source of H2O. In H-1 measurements for extremely dilute samples, the water peak is still a nuisance.
The solution is to soak NMR tubes in D20 overnight and store those tubes in a sealed container with a small beaker full of D20. Eventually almost all of the interstitial H2O becomes D2O. Now when the interstitial water is released into a deuterated solvent, the HDO residual peak is much less annoying..