A rash of catastrophic head crashes suffered by several US users of IBM’s 3380K disk drive has brought to light a problem that arose in manufacturing last autumn, Technology News of America reports from New York. IBM says that the source of the fault was isolated and corrected in the plants, but it is unclear how many bad disk drives went out before the problem was spotted. IBM says the problem was corrected early last autumn, but refused to say whether it affected all or only a batch of the drives made before it was discovered, so it is difficult to determine the likely cost to IBM – and the user community – of rectifying the problem. One New York stockbroker says that IBM replaced all the head-disk assemblies in a vast disk farm of suspect 3380Ks, and the biggest embarrassment for IBM, which claims between 18 and 20 years mean time between failures on its big disk drives, is that – according to its own figures, 80% of the world’s disk capacity is in just 200 sites – all of which are highly likely to be users of the 3380K, as well as being IBM’s most valued customers. Details and background – see page five.
