Can a semiconductor have positive temperature coefficient of resistance?

Is there a limit in inverse relation between temperature and resistivity of semiconductors?