This is going to leave a mark … looks like there is an HPC component as well …

I saw this a few days ago and ignored it at first. Vendor bashing pieces are nothing new from the media.
To paraphrase Mark Twain: Rumors of Dell’s demise are greatly exaggerated.
Dell is, and continues to be a powerhouse at pushing machines out. Their innovation is, basically customized machines in volume. There is no magic in their machines, they use the same parts from the same sources as others do.
Which means, if some of those parts are defective, they ball is in their court to assess the issue, and rectify it. This is what we do. Just had a customer with a DeltaV unit that simply wouldn’t behave correctly. So we sent them a replacement and they moved their data. We have the old unit back in, and will try to understand what happened with it, but its our responsibility, as codified in the warranty, to make sure that what we sell does the job.
It seems though, that Dell got bit by the capacitor fiasco a while ago. And it hit a cluster.

According to the Times, when the University of Texas complained that its Dell PCs were failing, the company said the school’s math department had pushed them too hard, making them solve difficult calculations.


The claim was that one shouldn’t use a desktop system for server work, and that the user pushed the system too hard. Rather than fess up and say, ya know, we have these bad capacitors, so if you see problems, send your unit in and well fix it … they told the customer that it was effectively the customers fault the machine died.
Anyone can get hit with bad parts. Which is why we stress our builds hard in our lab. We want them to fail here … not in the field. When we ship something out, we want to know it is going to work. We don’t want any uncertainty.
Unfortunately, this is not a part that you would see failure in the initial tests.
And the right thing to do is to own up to that.
If a customer wants to use their iPhone for their calculations, more power to them. There are pragmatic limits to what you can do on less powerful platforms. But as Doug Eadline keeps pointing out, a cycle is a cycle is a cycle, and you can in many cases minimize the cost of a cycle. So you can use desktop machines/processors for scientific computing. Though in many cases you can’t use ECC ram, which in my own experience, isn’t great (that you can’t use it). We’ve seen RAM errors from the strangest things and ECC helps fix some of them.
Apart from that, the cycles are still just cycles.
The problem I see is how they came across to the customers. I’d say a mea culpa is in order. Not sure if the lawsuit makes sense though.
The columnist makes the point

Even that would have been forgivable if Dell owned up and really fixed the inevitable problems.

The article’s title is The ignominious death of Dell. Dell isn’t dead, and I am not sure its even on the ropes.
But I’d like to remind people of another ascendant PC company named Gateway. Anyone remember them?
They were large while Dell was growing. Their fall came for many reasons.
Call it an object lesson in failure. I don’t think Dell is going down, and I don’t think they are going the way of Gateway. They had a problem, and I think they addressed it … I hope so. But they undoubtedly face challenges ahead.