facebook rss twitter

IBM GXP75 Failures!

by David Ross on 7 August 2001, 00:00

Tags: Hitachi (TYO:6501), IBM (NYSE:IBM)

Quick Link: HEXUS.net/qagr

Add to My Vault: x

IBM Failures!

Well this is an issue IBM! Oh Big Blue what are we going to do with you?

Well if you have been dead for the past 18 months then you wont know about this issue but there is a clear one. This evolves around their latest range of GXP75s and GXP60 hard drives. This could be crippling to there IDE division.

The hard drives are getting worse too! I have a lot of people mail me and ask for advice especially with their RAID Arrays when they have a problem like when it collapses. We have about 11 IBM drives used in the Hexus systems. So far 4 have failed, and one last night corrupted itself.

I have spoken to some local retailers to me here, and we are getting reports that they are getting 15 failures a week and they only sell around 60 systems a month, this is nearly a 100% failure. Most companies are suffering from 65 -> 70% failure. Not all drives will fail but a HIGH percentage are failing. Hasten to say they have now moved to Quantum drives again, which they had moved away from after the Maxtor buyout.

Whilst speaking to him this sums it up: -

Retailer "I have a 45gb IBM in my personal pc, it died last week, I lost nearly all my data!" This shows it isn’t an End user fault it is down to the drives plain and simple.

The biggest issue, which faces these companies is that, their customers don’t expect JUST a new hard drive they want more. They want all their data recovered. This is a "NIGHTMARE" for them, they thought that IBM had solved the issue with the introduction of the GXP60 range but these are dieing even more so now.

The main fault is down to the IBM GMR heads; this stands for 'Giant Magneto Resistive'. The point is manufacturers want as much data per platter as they can get, as this leads to cheaper drives with less plates. BUT! The data is too densely packed, and this causes heat problems. The constant heating/cooling cycles contaminates the data on the adjacent plate, and this means the head ends up reading / writing the wrong track, then you get the kkktch-click -click-lcik.

This happens as it seeks for the other data, and as it cant find it, then it causes more and more problems with the drive.

Some retailers knew this was happening a while back but they didn’t know the scale of the issue. This means they have to deal with hundreds of HD failures.

One of them spoke to an IBM Rep about this issue and this is the response.

Customer: 'Are your IDE GXP60/75s failing a lot at the moment?'

IBM: 'NO!' she said. 'No more than any other drive on the market'

There is an advantage though. some distributers are sick of it and have decided to credit the retailer the buying price when they fail, so a new drive for the same capacity costs less. They actually make money on failed hard drives (but loose when you count the labour in putting it right). The person I spoke to got £138 credit when they returned a 20GB drive, this is nice but not nice enough!

The 3 Hours data recovery causes costs of around £100 per drive this is costly especially under warranty. Most of the time it costs more than the initial drive in the system.

People can low level format the drives every 2 months but this doesn’t fix the problem it is a temporary measure. What we can say is that all the retailers we have spoken to have been very good about RMAs and are replacing bad drives quickly.