How serious can it be?

Let's take a different look at the problem. Everyone agrees that it's serious, but are there reasons to panic?

Intel states that the error has been localized to the specific clock generator responsible solely for the 4 SATA-300 ports. If this is true, we shouldn't forget that this chipset series offers 6 SATA ports in total, and the remaining 2 are SATA 6Gbps and have their own clock generator. One exception is lower-end solutions like B65 or H61, but those haven't really been shipped yet, so we might as well disregard them for the time being. So, this means that if you only have 2 drives, you might avoid the problem completely by simply using the SATA-600 ports. Most PCs only have always had a single hard drive, and optical drives are becoming optional. As for notebook buyers, they most likely don't have to worry at all, unless dual-drive models with additional eSATA ports are involved &mdash the latter may be in trouble (as well as desktop PCs with multiple storage drives). Although even eSATA isn't that necessary thanks to USB 3.0.

But, as we have already stated, "may be" doesn't mean "will be." Whether you'll experience the problem or not will also depend on how SATA ports will be used. These are just assumptions, but we believe that occasional use of potentially defective ports (e.g. for a DVD-RW drive or an external eSATA hard drive) will hardly cause chip wear. All the more so, if the drive is only used to read data. In this case even the actually occurring problem doesn't seem critical. Why, even drives under heavier load might not suffer, or at least not in the near future. The most pessimistic forecasts state that only up to 15% of PCs will experience problems within 3 years. And that's not the 15% of all PCs based on these chipsets, just the 15% of those which may experience problems. This equals to a few percent of sales or even less. And then one has to consider the possibility of such a motherboard breaking down for other reasons.

So, not everyone may experience this problem. Some users just may be unlucky, while most users won't even have a chance to experience it. The 'vulnerable group' will only include those who use multiple drives and that's closer to being a PC enthusiast, not a regular user. Ironically, PC enthusiasts were the first to buy the new LGA1155-based solutions. The rest either took their time or wanted to wait for the more affordable Core i3 CPUs slated for the late February or Pentiums scheduled for Q2'2011. But even among enthusiasts the majority doesn't have or need a lot of hard drives on a 24/7 basis. Of course, this doesn't mean that we should all just disregard the whole issue. There's a real chance of losing data. But that's not the end of the world, mind you.

What should regular users do?

Now that we've sorted out who's to blame, we need to understand what to do next. We know what Intel will do. We know what some motherboard makers will do. For example, Gigabyte: as a rough approximation, they'll follow in Intel's footsteps. They stopped shipping LGA1155-based motherboards and will resume those as soon as they get enough fixed chipsets. If you have already purchased such a board from Gigabyte, they should replace it without further ado, although you'll have to wait until the end of April. At that you don't have to just sit and wait until the fixed model hits the stores. If it doesn't go on sale late in April, you should tell them your product's serial number and wait for a notification. Other motherboard makers will most likely do the same.

What should regular users do? That depends which group they're in. If you haven't planned on buying an LGA1155-based PC, grab some popcorn and watch how major industry players get out of this scrape. If you have already upgraded your PC, you have a few options, none of those simple and trouble-free. You'll have to watch the news to find out whether you'll be able to have your motherboard replaced or not. We believe you should do it, if you have a chance. Why take another risk? Besides, even if you don't experience the problem, you'll have a hard time selling it, when you decide to buy something new.

If you have just two drives, connect those to SATA-600 ports. If you have more than two, use an auxillary drive controller, given your motherboard has one. Even if it results in a certain performance drop. Obviously, you shouldn't connect hard drives with vital data to potentially faulty ports (2nd to 5th). Even if only 0.01% of users lose their data, it won't be nice to be among those anyway. And, of course, make backups. Although that's something you should do anyway.

If you have been planning on buying a Sandy Bridge-based rig, you should probably wait a bit longer. After all, the fixed motherboards will be arriving soon and you shouldn't be in a hurry. If you do need to buy or upgrade your PC now, there are alternatives in the market. And none of those suffer from the problem described above. Why, they might have own issues, but the today's problem may only happen to PCs with Sandy Bridge processors (Core i5-2000, Core i7-2000). If you have something else, don't be bothered by this storm in a teacup.

If you like to take risks, you may buy an LGA1155-based motherboard now and replace it free of charge later. Either way, the ending should be happy. For most users. Over time.

Conclusions

We would all like to live in a perfect world, where every product is great and not susceptible of wear. Well, we're not, so errors happen, be it art or microelectronics. Fortunately, this doesn't happen too often, because developers manage to fix most of serious bugs before products go on sale. Not all of them, that's true. You may remember the popular TLB bug in the first AMD Phenom processors (it didn't show too often, mind you, although the processors turned out to be not as decent as expected). Or you may remember the issue with the first Marvell SATA-600 controllers — just a year and a half ago manufacturers had to urgently fix high-end LGA1156 motherboards right before the rollout.

Intel has its share of skeletons in the cupboard too. That error in the math coprocessor of the first Pentiums, for example. It remained unclear how high were chances to actually experience it, but the company didn't want to risk it and replaced the processors. Now there's a chipset problem. But, fortunately, they found it early enough, so all they need to do is replace about 8 million chips. Could be worse — they could've discovered the issue in, say, the third of fourth quarter of this year. Meaning that the number of defective motherboards could've been tens of millions. Or it could be better, too, if the chipset was faultless. But as we all know, he is lifeless who is faultless. Well, Intel is quite lively, so their chances of making mistakes are higher.

We believe that in the end all the hype will just result in somewhat lower income of companies involved and the slower advance of the LGA1155 platform. Some users will have to worry, too. And that's it.

Write a comment below. No registration needed!