>>>It's chatter...hardware chatter...regarding tech (non)support...in no way VFP-related, but...
>>>
>>>I'm talking to Buffalo tech support about a LinkStation Duo 2TB NAS drive that I was having some problems with. I asked the rep about why it was taking 3 days to do a disk erase and she claimed that the unit would slow down over time due to "wear and tear on the processor". Yeah, really, she said that. I guess all the electrons rushing through the processor tear it up. I should have asked for a performance degradation prediction curve (or maybe it's linear). What can I expect? 1 GHz when new, .75 GHz after the first year, .50 GHz after the second year, and so on. Maybe I'll get to zero GHz eventually, right? (These are made-up numbers, I have no idea at the moment what the speed of the processor is.)
>>>
>>>Can you believe this? What would you say to this woman? I told her it was simply not true and we discussed this, during which time she kept insisting, and I eventually asked to talk to someone else. She said she could put me back in the queue, which I agreed to, and then - coincidentally, I'm sure - I got disconnected.
>>>
>>>Now I know her statement to be untrue, but I did a little research anyway and found this: "Now then, as for mechanical degradation- in the solid-state world we call it "electromigration" ie the electrons very slowly migrate to try and make the atoms neutrally charged. This can create very bad things like shorts inside the CPU that simply cannot be fixed and your CPU just wont work anymore. It takes a very long time for this to happen (it's also a factor of materials used, size of the mask used, and voltage applied) and so you usually dont have to worry about it. Your CPU wiill not slow down due to elecromigration, it will simply go caput."
>>>
>>>Caput is something we can all probably relate to. Mainly I've had drives go caput, but solid-state parts have failed on me, too. But this is, of course, not what she was saying. She was claiming the processor would slow down over time due to wear and tear. Where do they find these people?
>>
>>I hear ya on the tech support... I've had similar things happen a few times, but never quite that blatant.
>>
>>As for your slow erase problem - I can think of a couple of things:
>>
>>- If you're doing the erase from a workstation, you might try from the NAS config/management software (if it has an erase option there) (or vice-versa)
>>
>>- If you *are* using an erase option in the config/management software, by default it may do a "secure" erase i.e. overwrite all of the disk with zeros or random values (may even do so multi-pass), so data recovery software won't work. In that case, depending on the secure erase algorithm used and number of passes selected, it could easily take several days to erase a TB-class drive.
>
>I thought about this little thread I started today, when I saw this quote from an article about Intel's recall of some chips:
>
>"Specifically, Intel found a design problem in a support chip, the Intel 6 Series, which is code-named Cougar Point. In a nutshell, chipsets with Serial-ATA ports could degrade over time and hurt the performance of hard drives and DVD drives."
>
>I wonder what causes this degradation over time, since this thread and the research I did indicate that performance degradation over time does not happen with chips/firmware/SSDs? The original article is here:
>
>http://www.zdnet.com/blog/btl/intel-hit-with-chipset-design-flaw-in-sandy-bridge-rollout/44257?tag=nl.e539
>
>Maybe it's just a poor wording/explanation of what is really happening. Whatever is happening, they said it will cost them $700 million to fix the issue and replace the components. Ouch!!

If Intel's explanation is accurate, my guess would be a physical design flaw in the chipset. If a circuit is taking more current than designed or runs unexpectedly hot then it may physically degrade over time. This might manifest as an increase in error rate, which could lead to excessive retries on communications links to attached devices such as SATA in this case, which would in turn decrease performance.

UPDATE: from http://techreport.com/discussions.x/20326 :

"The issue is a circuit design problem resulting in a gradual degradation over time of SATA connectivity on the affected ports, manifesting itself as high bit-error rates on those ports and eventually as total device disconnects."