As delivered, our NVidia K20m GPGPUs were in GPU Operation Mode 1/COMPUTE. If one tries to start the Xorg server,
[ 134.516] (II) NVIDIA: Using 4095.62 MB of virtual memory for indirect memory [ 134.516] (II) NVIDIA: access. [ 134.523] (EE) NVIDIA(0): Failed to allocate 2D engine [ 134.523] (EE) NVIDIA(0): *** Aborting *** [ 134.523] (EE) NVIDIA(0): Failed to allocate 2D objects [ 134.523] (EE) NVIDIA(0): *** Aborting *** [ 134.619] Fatal server error: [ 134.619] AddScreen/ScreenInit failed for driver 0
nvidia-smi --gom=0 and rebooting the node, the GPU Operation Mode can be set to 0/ALL_ON. Now, starting the Xorg server succeeds (and VirtualGL should work for remote visualization).
According to the documentation for NVidia driver 325.15, setting the GPU Operation Mode is only supported on “GK110 M-class and X-class Tesla products from the Kepler family”.