1. 26
  1.  

  2. 5

    IBM has a long history of lying to people. It goes back more than 30 years. My dad was a tech for a very long time and he was constantly told to lie to customers from the beginning, even when it was obvious to everyone involved that IBM was shipping defective hardware.

    1. 3

      Darn it, I have gotten comfortable defaulting to Crucial SSD’s as a nice, cozy point on the price/performance curve. Good thing I bought my last new SSD a year and a half ago. I guess I’ll have to find another brand that lies less. Anyone have recommendations?

      1. 4

        Well, I’m a happy Samsung customer. I’ve never had a part fail (yet) or have performance seriously degrade. They’ve delivered a smooth experience, although I haven’t really pushed the devices to see how fast they can go or how long they might last.

        That said, I think the point of the article is that no matter how good a past experience might be, there’s always a risk of a vendor doing some short sighted stupid thing that undermines consumer trust. All you can do about that is keep an eye on news reports and/or reviews to see if changes are happening that are incompatible with your usage.

        1. 1

          I have a few dead-ish samsung 840/850/940 SSD’s here…. though they worked fine for their warranted period, they do die after writing way more then the TBW value guaranteerd by samsung ;)

          First performance starts to degrade rapidly, and if you keep using them, at some point they report 0 size, or just completely dissapear from your sata bus :)

        2. 1

          Depends a bit on what you’re using the drives for. I’d be tempted by the Crucial drives if the QLC is significantly cheaper than TLC for my NAS. Even in the slow mode, it can handle 100 MB/s sustained transfers, in a RAID-Z configuration that means 200 MB/s, so the bottleneck will be the network (unless I upgrade to 10GigE, which won’t happen any time soon).

          The problem, as the article points out, is the lack of transparency. QLC SSDs look as if they’re still performing a lot better than spinning rust, so I’d be quite happy to buy them for this kind of use, if the marketing material didn’t lie about performance.

          1. 1

            Note that QLC also has many fewer erase cycles in its lifetime. It’s an interesting product for read heavy workloads, but a very different beast to TLC.

        3. 2

          Why lying is bad period.

          1. 1

            Within the culture shared by IBM’s customers.

            1. 1

              Lying is “bad” within any culture.

              1. 1

                Some cultures.

                Others have it bad only when done to certain people, or about certain kinds of things.

                1. 1

                  What do you mean by “have it bad”?

                  1. 1

                    Consider it a moral problem to lie.

                    1. 1

                      Assuming there is such a thing called “morals”.

          2. 2

            I haven’t seen that name in a long time. I can still feel it :D

            1. 3

              I never had those, but I bought a number of Maxtor 20-40GB drives at the same time and I don’t think any of them lasted more than two years. IBM handled it very badly, but it was a pretty awful time for spinning rust. When the first laptops with SSDs came out and people were talking about the write limit of SSDs, a back-of-the-envelope calculation suggested that they were much more reliable than any spinning-rust disks that I’d had over the prior 10 years.

              Disks have become a lot more reliable since the early 2000s. I don’t think I had a single disk last more than 3 years between about 1995 and 2005. I hope that the DeathStar was a bit of a wake-up call for the industry.

              1. 1

                You know those sick guys who freeze themselves hoping that there may be a cure in the future. Well, I kind of do that with all my old disks! Keeping them, hoping that one day, maybe, that they can be restored :D

            2. 1

              Way back around that time, I was a kid interested in programming. I’d written a bunch of toys on my own and I was interested in getting involved with a “real” project. I ended up working on this website (it was kind of cool, albeit kind of doomed; sort of a combination link-aggregator / blogging / social-network thing aimed at the Linux-nerd crowd). The site ran on PHP/PostgreSQL, on a small cluster of servers the site’s founder built himself, sitting in his basement (which he had a T1 pulled to). I didn’t get paid for any of my work, but every now and then he would send some hardware my way, which I considered a decent deal at the time. So when he was building a new database server with a big fast RAID, he bought one extra drive and sent it to me, and I promptly popped it into the family PC and transferred everything over; it was like 5 times bigger and faster than what I had before.

              Then the DB server had a drive failure and the RAID needed rebuilding.

              Then it had another one a month later.

              Then my drive started throwing errors, and I didn’t have RAID or any real backups. Luckily I convinced my parents of the gravity of the situation, and we went out to the store, bought a new drive, and I got 99% of everything copied over before the “old” (like 8 months) one quit for good.

              Needless to say, they were all 40GB DeskStars.