If the piston is physically attached (screwed in) to the top platen, you need to apply a vertical compensation load equal to the cell pressure times the area of the piston. You need to do this because the cell pressure pushes up on the piston, trying to blow it out of its bearing. If it is not attached to the top platen, then you are actually putting the sample in tension.

This upward force is not insignificant at high confining pressures. We had an old setup that used air pressure to apply the compensation load. One night while saturating a sample under high back pressure (and therefore high cell pressure), the pressure system blew a leak and the sample sheared in extension. No fun.

If your piston is not attached (the piston sits in a hole in the top platen) then you are getting a true all-around pressure, because if you should be leaving a small gap between the platen and the piston (not in contact).

Correction to last sentence of first paragraph: If the piston IS attached to the top platen, then you are actually putting the sample into tension.

Depending on the style of the triax chamber, you might need only compensate for the area of the push rod on the top of the sample. The loading plate on the top of the sample will apply the fluid pressure of its exposed area. If the test is a drained test, the pressure on the bottom of the sample will equalize, assuming saturation. If an undrained test, you will have to add load to the top of the sample to compensate for the area not pressurized at the bottom of the sample.