The bugs are inevitable. It is obvious that no developer wants to find fault with their creations, but sooner or later, the cracks appear. If you are angry enough at the software level, imagine the pain they cause to the users when they are in the hardware . Some have a minimal effect, while others can cripple an entire system. And no matter how great the company or manufacturer, there is no immunity. On this occasion, we will take a look at some of the hardware bugs most famous computer and users have suffered.
Have you bought a new component? A device that has just hit the market? Are you what you would call a "early adopter" ? And then you had problems with that you bought? Welcome to the world of bugs hardware. Because the fact is that yes, the hardware may also have bugs. Errors in the design and manufacture are not something unique to the computer, and history has made tremendous examples, as such as Titanic and low-quality steel. In recent years we have seen bugs that range from the subtle to the rank of unrecoverable catastrophic failure, requiring a complete replacement of the affected components.
 |
| The "Mark II Aiken Relay Calculator" was a failure in September 1947 ... a moth. (Source: Wikipedia) |
There have been cases in which a software patch has been mitigated or solved completely the effects of such errors, but the point remains that nobody likes buy a defective product, and in general has been the most irritating denial or neglect of the companies involved. When we talk business, we are not referring to dark hidden OEMs in Asia. We speak of giants like Intel, AMD, IBM and Nvidia , who heard about how hot the pan can be public. The list we offer today is far from complete, but has some examples that have been it relevant for users and for the market. IBM Deskstar 75GXP
There was a time when IBM, software giant as a few hard drives ago. The Deskstar series was one of the best known, not only for those individuals that worked well, but which crumbled inside computers. The Deskstar 75GXP , launched in March 2000, had some design flaws so severe that the wear on the plates was almost transparent the eye. Approximately fifty percent hard drives under that model failed during the first year of use, which earned them the name "Deathstars ." IBM sold its hard drive division to Hitachi in 2003, but it took four years to resolve the legal conflict that caused this grave error in design.
 |
| can see part of the aluminum chassis across the plate (Source: Flickr / clickykbd / Ryan Gallagher) |
Legend has it that in late November 1999, Intel launched the i820 chipset Pentium III processors, with the intention, inter alia, to introduce market RDRAM memory. The high cost of these reports was that the i820 was available with support for SDRAM, through called "Center for Translation Memory" , also known as MTH . What did happen? The MTH was full of holes, and a of them was so severe that Intel decided to offer motherboard replacement "and" delivery RDRAM memory as compensation. This did not surrender to Intel RDRAM, and tried to enter them with the first Pentium 4, but neither had much luck, since that time, the DDR were everywhere.
 |
| The i820 gave many headaches (Source: hardware-one.com) |
The bug FDIV (Pentium)
In June 1994, when the DOS and Windows 3.x ruled the world, a Lynchburg College professor named Thomas Nicely discovered a calculation error after starting to use one of the then brand new Pentium . It took four months to isolate the error, called "FDIV bug" for "floating-point division" , whose effect is to give erroneous results in floating-point divisions. When Nicely reported it to Intel, not only said they were aware of the problem since May of that year , but just replace the processor users who could prove they were affected by the bug. The media response was relentless, and Intel has changed its position by offering replacements in general. Intel had to spend about $ 475 million less by the bug itself, but for his unfortunate attitude against it.
 |
| The error was not serious, but Intel took over complicate things (Source: Wikipedia) |
The Phenom TLB bug
In late 2007 the community was looking forward to the new architecture AMD K10 . In September they launched the first Opteron, and two months later the Phenom family under the core "Barcelona". However, in the same month of November that the Phenom debuted on the market, it had to suspend production of the review "B2" due to a bug in the TLB or "Translation Lookaside Buffer" that in some cases could lead to a race condition and a possible computer crash. Loa modified their BIOS manufacturers to apply patches that could disable this buffer, but the performance loss could come at the worst, 20 percent . AMD needed five months to fix the bug in the B3 revision, but the damage was done, and Intel did not hesitate to continue gaining market share.
 |
| AMD needed several months to correct the problem of TLB |
Nvidia chips in laptops
Any user likes to have a more or less decent video on your laptop, and both the GeForce and Radeon chips are fulfilling that desire. However, in mid-2008 Nvidia had a serious problem with some of its chips. Basically, the whole family G84 and G86 had problems with overheating. This is compounded by a poor BGA installation in a large number of laptops, affecting brands such as HP, Dell and even Apple. In the best cases, a BIOS upgrade could reduce the clock speed on the chip, causing less heat generated. At worst, you suffer all kinds of video errors, and ended with your laptop useless. The attempt to hide this from Nvidia (aware of the case from November 2007) caused a loss 31 percent in the value of their shares.
 |
| The cost of a particular repair for this issue may elevarsre several hundred euros (Source: nbbbs.zol.com.cn) |
chipsets for
Sandy Bridge processors based on the Sandy Bridge architecture were slaughtering each benchmark that were on their way, however, was not a bug in them which stopped them, but an error in the chipsets under which they operate. Intel issued a withdrawal order for all motherboards with chipsets P67 and H67 support Sandy Bridge, due to a bug in the built SATA 3 Gb / s that can not be corrected by software. While it is estimated that the affected machines in a span of three years not exceed five percent , the bug causes a degradation in the SATA links, losing performance over time, and arriving at a possible malfunction detection units . Manufacturers have already enabled the replacement programs for consumers, and is expected to further review of the chipsets will be available in April, but for safety, it is reasonable to wait until at least May to ensure avoidance of the bug. A very ugly trip to Sandy Bridge, which no doubt is being used by AMD.
 |
| Companies like Asus and Gigabyte are already offering replacements for defective chipsets plates |
The plague of capacitors (capacitors)
Do you remember our gallery capacitor (or capacitors, as some prefer) ? Is that true horror show deformed and releasing chemicals? That was nothing more than one nuclear equivalent to a hardware bug . The manufacturers did not take the necessary steps to determine the origin of these components, focusing mainly on cost reduction, and both users and systems around the globe have paid (and continue to pay) the consequences. All that was saved manufacturers and more should be spent replacing components and by honoring the existing warranty periods. Without going any further, I recently found a Dell OptiPlex GX270 computer , and guess what the problem was ... ?
 |
| Without words ... |
Of course, the list goes . From the 72-pin connector on the Nintendo NES and its verification error 10NES chip to Red Ring of Death Xbox 360 , to recent things such as the decrease in signal iPhone 4 . Design problems manufacturing errors, and all other kinds of problems and will continue to plague hit our devices. After all, those who create are human, and something always going to escape. Good luck!
Source:
http://www.neoteo.com/los-peores-bugs-de-hardware.neo
0 comments:
Post a Comment