The M0-5 has a longer lifetime by design. The nature of silicon dictates that it will last virtually forever when driving a non-damaged load. When there is a faulted load the device protection is basically current limited to current shutdown which can stress the die considerably. As an example, the VNH5050 has a 30A minimum current limit with an actual limit somewhere between 30 and 60A. A possible application would be 50A on a 12V supply for 600W of power. This would cause a lot of stress on the die. To remain in a shorted load condition like that means that the output will be toggling off and on repeatedly driving high current and remaining at very high junction temperatures. Doing this for any length of time will cause metal migration; all kinds of damage to happen to the surface of the die. High current limit is desired because of the need to deal with in-rush currents. Stall current in a motor is present at start up. Motor torque is essentially proportional to the current flowing through the motor. To get the torque so the motor can start smoothly and quickly, the driver should provide for the start-up, or in-rush, phase with a high current limit. If the device drives high current long enough it enters into thermal shutdown – not power limitation but thermal shutdown. So once it hits thermal shutdown, then the current limit drops down to some more reasonable level. This level is still above what the part can handle normally in a normal application but still much lower than the in-rush phase current limit level. One other improvement to faulted load reliability in the M0-5 process is switch to passive bond pads for the outputs. Passive pads allow the current to be more evenly spread across the power section of the die. This more even spread of current evens out the temperature gradients across the die during faulted loads. This reduces the stress considerably.