Reliability problems, chronic instabilities or poorly designed technologies … there are many errors and problems with hardware devices that have existed throughout history, and in this article we are going to expose the most important ones. Surely many will ring a bell!
Obviously, in this article we do not intend to be exhaustive, but we want to review the biggest and most important errors that have occurred in the hardware industry over the years, which are many. Some could be described as authentic “epic fails”, and others surely do not ring a bell, but surely you also know some more that is not on this list, so we will be happy to read your comments about it.
The GeForce 8600M GT
This blissful graphics card has been tarnishing NVIDIA‘s image for a long time, and is actually still in the memory of many people. This mid-range graphic was used extensively between 2007 and 2008, and it had a problem with poor quality solders between the GPU and PCB that literally caused the GPU to detach from the board.
Because of this tremendous production error in the hardware for this notebook graphics, NVIDIA had to invest several million dollars to exchange them for new graphics without problems.
The 8MB error on Intel SSDs
It is likely that this problem sounds familiar to you, and it is that a game of Intel SSD 320 suddenly went from having several hundred GB of capacity to just 8 MB, also causing loss of all the data stored inside.
The problem arose in May 2011, and it was not until the end of August that Intel managed to fix the problem with new firmware. But of course, it had hundreds of users for three months with an SSD of only 8 MB, and also having lost their data.
Intel Pentium hardware errors
In 1994, shortly after Intel released the first Pentiums, an error was discovered: The FDIV division operation gave an incorrect result. Intel initially denied this bug, saying it was very rare. Finally, under pressure from users and with AMD eating ground, the company ended up assuming the ruling and changing the processors.
The Pentium 60 and 66 (P5 800nm) and some P54C (Pentium 75 at 100 MHz) were affected by this same error, although finally Intel corrected it in subsequent revisions of the chip.
Hardware errors for IBM 75GXP disks
In 2000 IBM launched its 75GXP hard drives, with a PATA interface and capable of reaching 55 MB / s. They offered 75 GB of capacity with 5 plates of 15 GB each (hence its name), but in the end the launch was a disaster because they had almost no reliability, forcing users to manage the warranty within a few days of starting to use it.
So much so, that IBM received 10 times more RMA management of these hard drives than any other, making it the worst product in its history in this regard, and many speculate that it was the turning point for which they ended up selling their division. of storage to Hitachi, given the poor image that the brand had.
Red Ring of Death on Xbox 360
Xbox 360 console users always lived in fear that one day their console would stop working, with the dreaded red ring of death (RRoD). This problem, very common in models with Falcon board, was due to the fact that the technique used to solder the GPU and the CPU to the board used a new type of solder that followed the RoHS directive, and was of poor quality because it ended up desoldering to cause of heat.
The scope of the problem was such that Microsoft had to develop a motherboard (Opus) exclusively for after-sales service, and thus be able to fix all Xbox 360 with the red rings of death that users took to technical service.
Seagate 7200.11 hard drives
Possibly the blackest stain in the manufacturer’s history. Hundreds of users began to experience problems with these discs as of January 2009, which suddenly became unusable without warning. The disks entered a “protected” state and prevented access to the data, although in some cases they were recovered by accessing the disk in alternative ways, such as connecting it to an external case.
So much so that Seagate ended up discontinuing this version and launching the 7200.12, with no problems, although not before launching a new firmware version that solved the problem of these discs.
Hardware errors in Intel chipsets
The Intel series 6 (Cougar Point) chipset sets for Sandy Bridge processors (P67, H67, Q67, HM67, HM65, QM67, QS67, H61 and B6) had a rather annoying error, and that is that their SATA controller It could fail, something really important given that many of these boards only had two SATA ports.
Many manufacturers released “V2” versions of these motherboards with the problem already fixed, or even installing third-party SATA controllers, not the Intel one.
The VIA KT133A chipset
This chipset has been a tremendous source of errors for users. It was used for quite some time in conjunction with AMD Athlon processors, replacing the AMD 76x (which was released by AMD simply to have a chipset at the time of the processors launch). In short, the problems included the inability to restart the computer, corruption in transfers between IDE hard drives, incompatibility with drives and even refused to work with many ADSL modems.
Fortunately, AMD stopped using this chipset on its motherboards as soon as they realized the myriad of problems they were carrying under their arms.
Everest controllers for SSD
In 2011, the manufacturer OCZ launched its Octane SSDs with SATA 2.0 interface. OCZ used an Everest controller based on the Indilinx design, but it turned out to have quite… problematic reliability. Apparently, after a few months of use these SSDs began to have performance problems that in many cases came to see it reduced by 50%.
Even the next model, the Vector incorporating Barefoot 3 controller, had some issues and OCZ indicated a rate of return over 10% in that model. Many speculate that it was the trigger for Toshiba to buy the company.
The GeForce FX 5800
Failed releases are rare in the world of graphics cards, but this one is almost iconic. It was a high-end graphics card released in 2003 that had two main problems: it was slower than its direct competitor, the ATI Radeon 9700 Pro, and it was extremely noisy. Like an airplane engine, something incredible.
On top of that, this graph was over-consuming and NVIDIA had a lot of trouble proposing viable variants to its NV3x family. Keep in mind that the GeForce FX were the first PCI-Express graphics of the brand, and with the PCX variants that integrated an AGP bridge to PCi-Express.
Crucial and 5,200 hours
The Crucial M4 SSDs, despite their high level of sales, also had some serious errors in their hardware, and that is that when the SMART controller exceeded 5,200 hours of use (more specifically 5,184 hours), the SSD became unstable and started to give all kinds of reading errors and data corruption.
The error was reported as soon as it was detected, but of course, those 5,184 hours had to pass. Fortunately for users, Crucial acted on time and released new firmware that fixed the bug.
AMD Phenom hardware errors
The AMD K10 processor, also known as Phenom, also does not escape this list. As soon as it was released, problems with L3 cache management appeared, especially when used with virtualization software, forcing AMD to roll back the release and fix its microcode.
Then, with the Phenom II, the error was already corrected and in fact it was one of the most successful launches of the brand. However, as for these first Phenom, well imagine what it meant for AMD to have to back down just after launching it to the market. A total mistake.