You will be delighted to hear that electrons are now being taught to go ballistic at great speed:

NEW SPINTRONIC SPEED RECORD. Spintronics is the science devoted to
gaining greater control over digital information processing by
exploiting electron spin along with electron
charge in microcircuits.

One drawback to implementing a scheme of
magnetic-based memory cells for computers has been the relatively
slower speed of spin transistors. Hans Schumacher of the
Physikalisch-Technische Bundesanstalt, Braunscheig, Germany, has now
devised the fastest yet magnetic version of a random access memory
(MRAM) cell, one that switches at a rate of 2 GHz, as good as or
better than the fastest non-magnetic semiconductor memories. The
MRAM architecture is a sandwich, consisting of two magnetic layers,
with a tunneling layer in between.

When the magnetic layers are
aligned (their spin orientation is the same) resistance in the cell
is low; when they are counter-aligned resistance is high. These two
conditions establish the binary 1 or 0 states.

The speed of writing
or reading data to and from the cells has, for MRAMs, been limited
to cycle times of 100 MHz by magnetic excitations in the layers.
This problem has now been overcome, according to Hans Schumacher
(hans.w.schumacher@ptb.de), through a novel approach referred to as
ballistic bit addressing. In the case of the new MRAM architecture,
the influence of magnetic excitations is eliminated through the use
of very short (500 picosecond)
current pulses for carrying out the write operation and that even a
bit whose value will remain the same undergoes a complete 360-degree
precession, whereas a change of status (say, from a 0 to a 1) will
be achieved by pivoting the magnetic status of the cell through 180
degrees.

The 2-GHz switching speed (the rate at which writing can be
accomplished) is faster than static RAM (or SRAM) memories,
currently the fastest memories, can accomplish. Furthermore, the
magnetic memories are non-volatile, which means that the status of
the memory does not disappear
if the computer is shut down.

(Schumacher, Applied Physics Letters,
25 July 2005; and Journal of Applied Physics, August 1;