Page 1 of 2

Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 8:26 am
by josephclemente
Press release: http://newsroom.intel.com/community/int ... g-solution
SANTA CLARA, Calif., Jan. 31, 2011 - As part of ongoing quality assurance, Intel Corporation has discovered a design issue in a recently released support chip, the Intel® 6 Series, code-named Cougar Point, and has implemented a silicon fix. In some cases, the Serial-ATA (SATA) ports within the chipsets may degrade over time, potentially impacting the performance or functionality of SATA-linked devices such as hard disk drives and DVD-drives. The chipset is utilized in PCs with Intel's latest Second Generation Intel Core processors, code-named Sandy Bridge. Intel has stopped shipment of the affected support chip from its factories. Intel has corrected the design issue, and has begun manufacturing a new version of the support chip which will resolve the issue. The Sandy Bridge microprocessor is unaffected and no other products are affected by this issue.


The company expects to begin delivering the updated version of the chipset to customers in late February and expects full volume recovery in April. Intel stands behind its products and is committed to product quality. For computer makers and other Intel customers that have bought potentially affected chipsets or systems, Intel will work with its OEM partners to accept the return of the affected chipsets, and plans to support modifications or replacements needed on motherboards or systems. The systems with the affected support chips have only been shipping since January 9th and the company believes that relatively few consumers are impacted by this issue. The only systems sold to an end customer potentially impacted are Second Generation Core i5 and Core i7 quad core based systems. Intel believes that consumers can continue to use their systems with confidence, while working with their computer manufacturer for a permanent solution. For further information consumers should contact Intel at www.intel.com on the support page or contact their OEM manufacturer.


For the first quarter of 2011, Intel expects this issue to reduce revenue by approximately $300 million as the company discontinues production of the current version of the chipset and begins manufacturing the new version. Full-year revenue is not expected to be materially affected by the issue. Total cost to repair and replace affected materials and systems in the market is estimated to be $700 million. Since this issue affected some of the chipset units shipped and produced in the fourth quarter of 2010, the company will take a charge against cost of goods sold, which is expected to reduce the fourth quarter gross margin percentage by approximately 4 percentage points from the previously reported 67.5 percent. The company will also take a charge in the first quarter of 2011which will lower the previously communicated gross margin percentage by 2 percentage points and the full-year gross margin percentage by one percentage point.

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 9:26 am
by Tzupy
This news is kind of welcome for me, since it stops my 'Sandy Bridge itch', at least for a while. :lol:
Of course, there is nothing funny for the people who already bought SB motherboards, I hope there will be some BIOS workaround too.
I wonder how can someone test for this degradation over time of the SATA ports. Could SPCR, already having tested several SB motherboards, prove the existence of this flaw?

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 12:18 pm
by continuum
Fortunately it's only the PCH's 3Gbps ports... rather than all of them.

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 1:04 pm
by whispercat
This news is all over the Mac boards as well. Apparently, most of those chips were destined for Mac machines:

http://forums.macrumors.com/showthread.php?t=1088682

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 1:29 pm
by ces
Tzupy wrote:This news is kind of welcome for me, since it stops my 'Sandy Bridge itch', at least for a while. :lol:
I know what you mean. I was already even getting started on an Ivy Bridge itch... it will be only about 12 months before it arrives.

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 2:13 pm
by MikeC
http://www.pugetsystems.com/blog/2011/0 ... idge-flaw/

"First, let’s clarify the issue. Puget Systems has extensively qualified the Sandy Bridge platform, and in our testing, we have not encountered this issue. We have also had no reports of this issue from any of our customers, and were not informed of this issue until today. The issue is a long term performance degradation, and it does not damage your storage device. The [potential] problem is on ONLY the 3Gb/s SATA ports on the Intel chipset controller. If a motherboard has SATA ports on a 3rd party chipset (such as Marvell), or on a PCI-E SATA or RAID controller, they are not affected."

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 2:45 pm
by Tzupy
I don't understand why this SATA malfunction is a 'design issue'. A design issue should have been spotted when Intel tested the chipset.
I believe Intel didn't need to redesign the 3Gbps SATA already used in the X58, and which performs well.
Maybe they tried to cut some corners and the incriminated redesign - which should have saved a couple million $ - turns out to cost a lot more?
I would have expected the 6Gbps SATA to be not perfectly reliable, not the 3Gbps one.
Since I had an issue with the SATA in my previous X58 mobo, I don't know if I can trust the new 6Gbps ports.
Anyway, the user reports on the new Sandy Bridge mobos are riddled with problems, as expected with early BIOS / UEFI versions.

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 3:23 pm
by cloneman
Why not offer people a free PCI-E SATA6 device instead of recalling the boards? Wouldn't that be cheaper for them?

Let the customer choose which option they prefer

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 3:42 pm
by Rebellious
AM3 890GX mobos will be hard to find for a while.

Re: Sandy Bridge, Part 6: Recall

Posted: Mon Jan 31, 2011 6:35 pm
by josephclemente
EDIT: I'll wait to see what kind of solutions will be available. Knowing that only the four 3Gbps ports are affected isn't horrible. Plus the eSata port on my motherboard is controlled by Marvell.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 1:53 am
by Tzupy
A plausible explanation for the bug, at Anandtech:
http://www.anandtech.com/show/4143/the- ... t-sata-bug

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 3:37 am
by ame
josephclemente wrote:EDIT: I'll wait to see what kind of solutions will be available. Knowing that only the four 3Gbps ports are affected isn't horrible. Plus the eSata port on my motherboard is controlled by Marvell.
I'm using the same motherboard and not having any problems. I don't think the eSATA is Marvel though. I would have noticed the device driver.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 3:51 am
by tim851
I know this is not exactly helpful, but one of the things that popped into my mind when every other topic here became about people's new Sandy Bridge builds, was how "careless" (for lack of a better word) I thought it was to jump on new tech so quickly. Obviously I didn't expect a failure of this magnitude, but I think it's never a good idea to jump into things before they had time to mature. I thought of unfinished BIOSes and flaky drivers. There is a reason why Apple is almost never first to market with new technology. They always seem to wait a couple of months, until Intel/ATI/nVidia got some consumer feedback.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 4:16 am
by nutball
tim851 wrote:I know this is not exactly helpful, but one of the things that popped into my mind when every other topic here became about people's new Sandy Bridge builds, was how "careless" (for lack of a better word) I thought it was to jump on new tech so quickly.
There are two elements to this in my mind.

1. It's great and all to be cautious and wait to tech to mature but in many cases, particularly where failures are due to in-the-field wear-and-tear, if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing! It might be great for you personally to know that someone else has ironed out the bugs by taking the pain of the failure, but if everyone had that mindset no-one would ever buy anything. The in-the-field testing would never happen and the problems never be found.

2. In this case, which looks like a problem with design and in-house life testing, I find it quite difficult to believe that Intel only found out about this yesterday. Sandy Bridge launched three weeks ago. Intel seem to have already gone a long way down the route to identifying the problem at the transistor level, designing a fix, getting the board to authorize a large pot of cash to pay for recalls, got OEMs on board... and so on. So basically what I'm getting at here is that I'd lay money that at least someone at Intel knew about this problem before SB launched. From that perspective, yeah I can see that waiting is a good thing to do :)

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 4:56 am
by ame
nutball wrote:
tim851 wrote:I know this is not exactly helpful, but one of the things that popped into my mind when every other topic here became about people's new Sandy Bridge builds, was how "careless" (for lack of a better word) I thought it was to jump on new tech so quickly.
There are two elements to this in my mind.

1. It's great and all to be cautious and wait to tech to mature but in many cases, particularly where failures are due to in-the-field wear-and-tear, if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing! It might be great for you personally to know that someone else has ironed out the bugs by taking the pain of the failure, but if everyone had that mindset no-one would ever buy anything. The in-the-field testing would never happen and the problems never be found.

2. In this case, which looks like a problem with design and in-house life testing, I find it quite difficult to believe that Intel only found out about this yesterday. Sandy Bridge launched three weeks ago. Intel seem to have already gone a long way down the route to identifying the problem at the transistor level, designing a fix, getting the board to authorize a large pot of cash to pay for recalls, got OEMs on board... and so on. So basically what I'm getting at here is that I'd lay money that at least someone at Intel knew about this problem before SB launched. From that perspective, yeah I can see that waiting is a good thing to do :)
I agree with 1.
I'm typically an early adopter, but only when I feel comfortable and want the product badly 8) . I'm also a beta tester and a system builder, so SNB was just perfect for me. Had my system been non functional my supplier would have replaced it at once no questions asked. Apple are very smart to wait, but they are already invested in this, just not released.

However I disagree with 2. If Intel knew about this PRE release they would have saved hundreds of millions of dollars on shipping alone. Read Anand's article about it, you might appreciate the fact that the cost of repair and labor is minor compared to the cost of shipping the parts back to OEMs and back to customers. According to Anand's report they have first learned about the problem a few days ago and have been testing to find the exact cause. I'm not saying its not possible but its highly unlikely IMO.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 5:00 am
by alecmg
Yeah, its been at least a week since Intel got an idea about scale of problem
Their stock market actions point to that.

I'm comforting myself with idea that even if the controller fails, CRC error checking should not let files get damaged on hdd.
Hooking it up on 6gbps ports anyway :)

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 6:13 am
by nutball
ame wrote:However I disagree with 2. If Intel knew about this PRE release they would have saved hundreds of millions of dollars on shipping alone. Read Anand's article about it, you might appreciate the fact that the cost of repair and labor is minor compared to the cost of shipping the parts back to OEMs and back to customers.
Oh I appreciate the cost implications, and I have read the article! On the other hand we're talking about products that shipped from Intel many months ago, to the motherboard makers, then to the distributors, then to the retailers... they were sitting in retailers warehouses on the date of SB launch. The faulty product was already out in the wild, it just hadn't got to the end customer yet.

Note that I'm not suggesting they knew about it months ago, merely that they may have found out about it shortly before launch.
According to Anand's report they have first learned about the problem a few days ago and have been testing to find the exact cause. I'm not saying its not possible but its highly unlikely IMO.
Well, maybe. Time will tell, as I doubt the lawyer community will be able to restrain itself from launching a class-action lawsuit alleging similar things to my random musings :)

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 6:18 am
by ame
nutball wrote: Well, maybe. Time will tell, as I doubt the lawyer community will be able to restrain itself from launching a class-action lawsuit alleging similar things to my random musings :)
No doubt :lol: :( :lol:

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 9:21 am
by CA_Steve
nutball wrote:Note that I'm not suggesting they knew about it months ago, merely that they may have found out about it shortly before launch.
Having worked in the semi industry for a pretty long time (not Intel or anyone that I comment on this site), I can tell you that nothing makes your bowels loosen more than a potential lines-down situation at your customer - let alone a recall. Yow. It's awful on so many levels.

So, Intel knowing about this problem shortly before launch and continuing to process wafers, assemble, and ship product to their customers, is about as likely as Elvis still being alive.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 9:40 am
by m0002a
nutball wrote: ...if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing!
WTF? Since when are customers who buy product at retail considered to be "beta-testers"?

A beta-test is conducted prior to general product release with a controlled list of customers/users, all of whom sign waivers that they understand the pitfalls of using the beta-test products.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 9:51 am
by dhanson865
m0002a wrote:
nutball wrote: ...if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing!
WTF? Since when are customers who buy product at retail considered to be "beta-testers"?

A beta-test is conducted prior to general product release with a controlled list of customers/users, all of whom sign waivers that they understand the pitfalls of using the beta-test products.
Products go through a Alpha, Beta, etcetera - Release schedule no matter if you do testing or not.

The implication is that the order in some cases is Alpha - Release - Beta or that it was Alpha - Beta - Release when it should have been Alpha - Beta - Gamma - Release. In shorter words the implication is that the product was released too early without enough testing and at least one revision before it's truly ready.

Some companies use waivers and NDAs to beta test a product, some just release it and hope for the best. Normally Intel isn't that company and that is what makes this so newsworthy.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 9:51 am
by nutball
m0002a wrote:
nutball wrote: ...if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing!
WTF? Since when are customers who buy product at retail considered to be "beta-testers"?

A beta-test is conducted prior to general product release with a controlled list of customers/users, all of whom sign waivers that they understand the pitfalls of using the beta-test products.
Calm down! I know exactly what a beta-tester is.

I used the term euphemistically in the context of replying to the sentiment in the post I replied to - that it's better to wait a bit for tech to mature after it comes on the market. That process of maturation involves what...? Bugs and problems being found in the field that weren't found in earlier testing?

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 11:16 am
by ame
m0002a wrote:
nutball wrote: ...if it weren't for the early adopters doing the beta-testing then the cautious late adopters would be the early adopters doing the beta-testing!
WTF? Since when are customers who buy product at retail considered to be "beta-testers"?

A beta-test is conducted prior to general product release with a controlled list of customers/users, all of whom sign waivers that they understand the pitfalls of using the beta-test products.
Yeah, he didn't mean beta testing per say, just that there is a time period until technology matures.

Every modern product can have failures that were missed during "official" testings, just this year we've seen it in Toyota cars Apple phones and now Intel chip sets. We all know the meaning of the word recall and it wasnt invented by Intel.
Intel, it seems, is trying to do the right thing by announcing the issue ASAP, as opposed to the other example I mentioned - Toyota tried to cover it up until it was too late and people were killed, and Apple dismissing it as if it never happened. Apple should thank their blindly loyal customers and brilliant interface design team - the proposed plastic cover workaround was accepted by the public.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 12:55 pm
by merlin
While it's obviously egg on the face. Intel totally did right. They announced the issue in a relatively timely manner( and are basically funding a refund/repair of everyone's motherboards. Far better than the problems with apple, toyota, nvidia, bp *cough cough*, etc. There are many companies that have done far worse. The major issue is obviously anyone's upgrade plans are dropped to march or april, but it's not like there aren't plenty of computer alternatives for the desperate.

But the clear and obvious takeaway (which I feel is true for almost all new products), don't buy the first version or give it time to flush out all the possible issues. I'm glad to have generally bought my things at least 3-18 months after the initial release of most products. There's no reason to buy early unless you really love paying too much and just having pointless bling. Reliability is far far more important in my book and always will be every time I buy something.
Although I did kinda break my rule on the Radeon 6850 by buying it after only 2 months of release, but then again I only paid like half price ;)

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 1:03 pm
by nutball
Looks like Charlie isn't entirely convinced by the official version of events either (and we all know how much Charlie likes a good conspiracy!) :)

http://www.semiaccurate.com/2011/01/31/ ... any-water/

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 1:22 pm
by merlin
Semiaccurate, aka completely inaccurate crap fud. It's sad I think semiaccurate is more fud than fudzilla. We've already gotten a pretty good detailed explanation directly sourced from Intel by Anandtech that explains exactly what went wrong and that it was because of a new shipping revision of the chipset. They couldn't have caught the issue way earlier because it was only caused by the revision B itself.

Details here: http://www.anandtech.com/show/4143/the- ... t-sata-bug

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 2:09 pm
by cordis
Yeah, I'm in the semiconductor industry, and Anand's explanation sounds very plausible. Tool support isn't really strong for mixed voltage domain stuff, and even though Intel's been doing it for a while I can see how a problem could have popped up. Sounds like they're on top of it and doing the right things, always good to see.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 2:51 pm
by Tzupy
Saying that 'Intel knew before Sandy Bridge launch' is as wrong IMO as saying 'Intel didn't know ...'.
Intel is a huge corporation, and IMO a few people knew about the problem well before the SB launch.
Their problem may have been that they couldn't locate / prove the malfunction to their direct bosses.
And who would dare to postpone the Sandy Bridge launch without definitive proof?
Some tech guys? I don't think so, it's only the board of directors that can take such a decision.

Re: Sandy Bridge, Part 6: Recall

Posted: Tue Feb 01, 2011 9:04 pm
by Vibrator
I think Intel reacted in the proper way to this situation.

Re: Sandy Bridge, Part 6: Recall

Posted: Wed Feb 02, 2011 6:39 am
by m0002a
nutball wrote:Calm down! I know exactly what a beta-tester is.

I used the term euphemistically in the context of replying to the sentiment in the post I replied to - that it's better to wait a bit for tech to mature after it comes on the market. That process of maturation involves what...? Bugs and problems being found in the field that weren't found in earlier testing?
Unfortunately, especially in the software world, many companies do expect their early adopters to be beta-testers. All software has bugs (usually lots of them) and there is continuous improvement with software patches throughout the life-cycle of the software. Upgrading software at customer sites is very cheap, and many products have automatic software upgrade features installed so in many cases the software installed is fixed before the customer even encounters a problem.

I don't believe that it is the same with hardware, and especially for a company like Intel. A mistake can be extremely costly to fix and get to the customer who has already purchased the product or for OEMs have have manufactured motherboards using a chipset ($1 Billion in this particular case), not to mention the bad press that might delay customer adaptation of the product.