Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: steve harris who wrote (163608)	4/6/2002 4:58:12 PM
From: wanna_bmw	Read Replies (4) \| Respond to of 186894

Steve, Re: "I was expecting to learn that long errata lists are why Intel processors are more reliable." Errata lists are neither good nor bad. They are simply a means to communicate problems that haven't been fixed in a given product release. There are plenty of reasons not to fix a given problem. Mostly, it has to do with the severity of the problem, and the ease of the work-around. For problems that do not get fixed, it's probably because the work-around is easier to implement than the fix itself. For more severe problems, the fix is eventually implemented, though perhaps not in the initial steppings. That's because the cost to make an initial stepping is normally quite high for a CPU. Validation is an ongoing process, and many bugs are found after silicon actually ships to market. With any and every product, there is a risk of recall, but the idea is to minimize the risk by fixing the bugs that have the greatest likelihood of happening. Otherwise, with the size and scope of the validation space, it would take forever to completely validate a piece of silicon as complex as a CPU. That's why errata lists are published. It's the responsibility of Intel to give their customers, their developers, and third party manufacturers a list of bugs that have been found, and the work-arounds to prevent those bugs from happening. It is not a burden to work around errata; it's simply a fact of life in the semiconductor industry that all ICs have at some point bugs that have not been fixed, or in many cases, bugs that have not been found. AMD doesn't publish a long list of errata for their CPUs, and I find it ridiculous to believe that the Athlon simply doesn't contain bugs. I was joking before when I said that AMD "lies" about the errata in their CPUs. More likely, they just end their validation cycle before more bugs can be found. With the number of resources that AMD has, and the limited head count, it's very likely that validation engineers get shifted over to new projects before the more obscure bugs can be found. And many bugs in a CPU are obscure, and not likely to affect users over the given life of the CPU. AMD takes this risk, because it's a good risk to take. If they assume a given lifetime for a CPU, they they can have more efficient ways to validate, and shorten the validation cycle of their products. Intel has traditionally been very strong with validation, and often assumes longer product life-cycles relative to the industry. Other than the times they've been pressured to release new products, like the 1.13GHz Pentium III and the MCH, they've been very successful at catching bugs before they reach the market. Intel got several black eyes in 2000 for launching several products that got recalled, which probably reminded them how important the validation process really is. I doubt that they will be making similar mistakes in the near future. AMD has yet to experience the terror of a product recall. It's the bane of any company to admit that they have launched a product that wasn't fit to sell. In reality, all companies launch products that have some degree of uncertainty in their design. It's a level of risk that's associated with every product release. With errata lists, at least a company is able to communicate known bugs and relate work-arounds. But even with these lists, there are often more problems that haven't been found. If you think that AMD is exempt from this, than you are wrong. Each product that AMD launches is capable of being the next CPU recall. How they manage the risk is up to them, and glancing at the errata list is only going to tell you the bugs that they know about, not the ones that haven't been found. wbmw