If you are looking for glucosyl/galactosyl ceramides, these can be fractionated from total lipid extract using normal phase silica column. Simply, dissolve the total extract in 100% chloroform and load onto the silica column (cartridge) conditioned with chloroform. Then you can elute cermides by 2 mL 9:1 chloroform:methanol and then in subsequent elution using 2 mL 8:2 chloroform:methanol glucosyl/galactosyl ceramides can be fractionated.
Good advice from Abdul above on sample prep. For a simple HPLC separation method, you may find the link below useful. "A Rapid Method For Phospholipid Class Separation by HPLC Using an Evaporative Light-Scattering Detector".
Article A Rapid Method For Phospholipid Class Separation by HPLC usi...
Good answers above, the silica cartridges work well. A simple, low-cost alternative; use solvent-resistant mini-columns or plug glass Pasteur pipets with a small plug of glass wool. Add 100-200mg dry silica gel. Suspend the gel with chloroform until air bubbles are removed. This solvent elution has good selectivity for the neutral glycosphingolipids up to 4 sugars; load the sample dissolved in chloroform and wash with 2 mL chloroform (unbound and wash contains neutral lipids such as cholesterol). Elute glycosphingolipids with acetone:methanol 9:1, 10-20 column volumes. Elute phospholipids with methanol. See http://www.cyberlipid.org/fraction/frac0010.htm