Fujitsu admits 4.9 million (potentially) defective HDDs
By Keith Warburton, PC Association
Posted: 05/11/2002 at 09:16 GMT
theregister.co.uk
The papers lodged with the court in California, in the $50 million action between Cirrus Logic and Fujitsu (of Japan), confirm what observers had already guessed, that – contrary to the blasé denials of any notable problems - there is indeed an officially acknowledged problem with the MPG3xxx series of drives. And Fujitsu knew the massive scale and scope of the problem at least 18 months ago.
Cirrus Logic's case against Fujitsu for breach of contract was filed on 19th October 2001 – over a year ago, as was Fujitsu’s counterclaim for breach of contract & breach of warranty.
Fujitsu’s documents make it clear that it knew Cirrus Logic had supplied defective chips as long ago as July 2001, saying that it started to receive complaints about failures from May 2001. It promptly informed Cirrus Logic and requested information regarding the nature and extent of the problem and proposed remedies, but says that this information was not forthcoming.
The blame is being laid at the door of the supplier of the epoxy mould compound used in the manufacture of Cirrus' Himalaya 2.0 and Numbur chips. It is claimed that in the summer of 2000 the supplier of the epoxy, a Cirrus Logic sub-contractor - made the first of several changes to its product, and it was this that ultimately caused the chips to fail by short-circuiting. Fujitsu claims that Cirrus Logic should have known of these changes – which it says were "significant" – and that it should have warned Fujitsu.
At time of lodging the court papers Fujitsu estimated that approximately 4 million of the approximately twelve million Himalaya 2.0 chips it bought, and all of the approximately 900,000 Numbur chips are defective; that these defects have already caused a substantial number of these chips – and thereby the drives in which they are used - to fail; that the time before failure is variable and unpredictable, and that it is "highly likely" that a significant number of additional chips will short and cause the drives to fail. Based on field data, Fujitsu determined that the failure rate of the chips (and thereby, we suggest, the failure rate of their own drives) is "in excess of all reasonable industry standards".
Fujitsu says that Cirrus refused to deal fairly with it because it failed to take prompt and thorough action and didn’t provide it with full information, thereby harming Fujitsu’s customer relationships –a critical Fujitsu asset. This claim will no doubt bring an outraged response from Fujitsu’s trade customers and end-users, who have been trying to get an open response from the company for many months. Fujitsu's stance has been either an outright denial of any problem (stand up, Fujitsu Germany, Fujitsu Canada & Fujitsu in Australia) or a grudging statement that "some customers have reported problems".
That Fujitsu has consistently failed to take prompt and thorough action to remedy the problems being suffered by its customers, and indeed has denied any such problem, when there have been documents in the public domain (albeit well hidden in court filings) admitting the problem, is a shameful reflection upon the company and its claimed care for its customers.
By Fujitsu’s own statement potentially 4.9 million of their drives will fail outwith normal life expectation. This admission could provide valuable ammunition in world-wide legal actions against the company. |
