Sandy Bridge design flaw - Intel halted on NASDAQ - updated 2/8/11.

[Vette73]

This does not affect the 6Gb ports, ONLY the 3Gb ports.
So make sure your Harddrives are on the 6Gb ports and wait for the new chipset and you should be ok.

http://www.anandtech.com/show/4142/intel-discovers-bug-in-6series-chipset-begins-recall

 

[Idontcare]

What is really cool of Intel here is that they have known about this issue for some time now...long enough in fact that they had time to debug it, find the root-cause, implement a candidate fix, confirm the candidate fix actually works, and have started full-scale production of the corrected chipsets.

All while they were busy letting their OEM suppliers ship and sell known bad product to you, the customer.

At least AMD had the ethics to issue a stop-notice asap when they first uncovered the issue with the Barcelona TLB.

I remember well the early days of the Pentium FDIV bug, originally they required the customer to convince them that the FDIV bug would actually imperil your data (my father worked at an engineering firm at the time, he actually had to make the case to Intel that the FDIV bug would be a problem for any company relying on their chips to build bridges and buildings)...they only changed this initial policy after the fact when there was much outrage over the situation.

All you poor chaps who bought a known (to Intel) bad chipset this past month...I'd be super pissed if that happened to me. Intel knew they had bad product in the field and they sat on that info until they could make it seem like it was no big deal in the press because they had fixed chips working their way through the fab now.

 

[Arkaign]

Looks like these things will be a good option for people wanting to keep using their SB builds until new mobos arrive :

http://www.newegg.com/Product/Produc...82E16816115075

 

[Vette73]

^^

Yea Anand says they, intel, are just contacting OEMs today. So must be nice at gigabyte, asrock, etc... to be getting a pile of calls/e-mails and not have any idea what is going on. You think they give them a heads up first to get ready and have answers.

 

[SlitheryDee]

I just looked at newegg and MOST 1155 boards do NOT have extra SATA ports from a 3rd party contoller.

Only 15 out of 58 boards have extra 6Gb SATA ports

There's an article on the front page of this site about the issue. It looks like the intel 6gb ports are safe. Only the 3gb ports are affected. So at least we can use the native 6gb ports.

 

[Spikesoldier]

What is really cool of Intel here is that they have known about this issue for some time now...long enough in fact that they had time to debug it, find the root-cause, implement a candidate fix, confirm the candidate fix actually works, and have started full-scale production of the corrected chipsets.

All while they were busy letting their OEM suppliers ship and sell known bad product to you, the customer.

At least AMD had the ethics to issue a stop-notice asap when they first uncovered the issue with the Barcelona TLB.

I remember well the early days of the Pentium FDIV bug, originally they required the customer to convince them that the FDIV bug would actually imperil your data (my father worked at an engineering firm at the time, he actually had to make the case to Intel that the FDIV bug would be a problem for any company relying on their chips to build bridges and buildings)...they only changed this initial policy after the fact when there was much outrage over the situation.

All you poor chaps who bought a known (to Intel) bad chipset this past month...I'd be super pissed if that happened to me. Intel knew they had bad product in the field and they sat on that info until they could make it seem like it was no big deal in the press because they had fixed chips working their way through the fab now.

typical intel behavior.

early adopters screwed again. just like the G1 SSD owners

 

[Arkaign]

There's an article on the front page of this site about the issue. It looks like the intel 6gb ports are safe. Only the 3gb ports are affected. So at least we can use the native 6gb ports.

Yeah, though there are only two of those ports, and many/most people with SB custom builds will have more than 2 sata devices. Typically a raid array, plus an extra drive, plus an optical drive, sometimes more.

Anyhow, grabbing a PCIe Sata controller isn't the end of the world, Intel should offer a free 4-port controller of this sort to those that are severely inconvenienced by such a flaw.

 

[Arkaign]

typical intel behavior.

early adopters screwed again. just like the G1 SSD owners

Intel screws up about as rarely as anyone else, every major tech company has problems with things :

Apple
Microsoft
AMD
Intel
IBM
Nvidia
ATI (back when they were seperate)
etc.

This is not 'typical Intel behavior', this is just something that slipped through, similar to the Phenom bug, deathstar failure rate, etc.

 

[grdax]

More info from an Intel conference call
http://www.legitreviews.com/news/9957/

Thanks. Then it looks like we can use our SB setup until the fix is ready. We would be down for about a week at the longest while waiting for the replacement to come.

I wonder if my two 6gb SATA port is backward compatible...hmmm.

 

[Spikesoldier]

Intel screws up about as rarely as anyone else, every major tech company has problems with things :

Apple
Microsoft
AMD
Intel
IBM
Nvidia
ATI (back when they were seperate)
etc.

This is not 'typical Intel behavior', this is just something that slipped through, similar to the Phenom bug, deathstar failure rate, etc.

ok but they dont do anything for the customer and leave them hanging.

would you be ok that you got sold a piece of shit? when they knew it was a piece of shit?

 

[richierich1212]

ok but they dont do anything for the customer and leave them hanging.

would you be ok that you got sold a piece of shit? when they knew it was a piece of shit?

Seems like they did know about it because how could they have discovered degradation so quickly? SB has only been out since the 9th.

 

[Spikesoldier]

how long has it been in the hands of engineers and reviewers?

 

[richierich1212]

Guys on OCN are speculating about EVGA knowing about this issue and their delaying of releasing mobos for SB. lol

 

[Accord99]

At least AMD had the ethics to issue a stop-notice asap when they first uncovered the issue with the Barcelona TLB.

Wasn't it only for Opterons though? Phenoms were still being sold.

 

[Wreckage]

How many people will never know they have a bad part until it's too late?

Also "Cougar Town" heh, heh.

 

[khon]

What happens to those of us with a system builders license for Windows ? Won't replacing the motherboard mean that we have to get a new license ?

 

[Spikesoldier]

What happens to those of us with a system builders license for Windows ? Won't replacing the motherboard mean that we have to get a new license ?

i really think thats only an issue on oem boards that check the DMI serial number programmed into the board.

 

[drizek]

I just looked at newegg and MOST 1155 boards do NOT have extra SATA ports from a 3rd party contoller.

Only 15 out of 58 boards have extra 6Gb SATA ports

Oh, I didn't realize that P67 had native 6Gb SATA. I thought they were all using a third party controller.

They did say that their 6Gb ports are not affected though.

 

[HumblePie]

What is really cool of Intel here is that they have known about this issue for some time now...long enough in fact that they had time to debug it, find the root-cause, implement a candidate fix, confirm the candidate fix actually works, and have started full-scale production of the corrected chipsets.

All while they were busy letting their OEM suppliers ship and sell known bad product to you, the customer.

At least AMD had the ethics to issue a stop-notice asap when they first uncovered the issue with the Barcelona TLB.

I remember well the early days of the Pentium FDIV bug, originally they required the customer to convince them that the FDIV bug would actually imperil your data (my father worked at an engineering firm at the time, he actually had to make the case to Intel that the FDIV bug would be a problem for any company relying on their chips to build bridges and buildings)...they only changed this initial policy after the fact when there was much outrage over the situation.

All you poor chaps who bought a known (to Intel) bad chipset this past month...I'd be super pissed if that happened to me. Intel knew they had bad product in the field and they sat on that info until they could make it seem like it was no big deal in the press because they had fixed chips working their way through the fab now.

Uhh... according the Anandtech article, this is not the case. Anandtech specifically states that Intel did not know the exact nature of the bug until last week. They didn't receive complaints until they had already produced 100K units first. Even then, they weren't able to reproduce and track down what the problem was until last week. This is why they do not have a replacements in place and won't until the end of February at the earliest. This is why they are expecting sales to HALT and lose them $300M in lost sales in total (meaning the lifetime of this product) and then spend $700M in replacing all the recalls.

I don't know, but this seems pretty ethical to me.


Even with all that, there is known workaround for the meantime. If your board has another sata controller or at least the sata 3 controllers, use those. If not, buy a sata card.

 

[khon]

Are the any problems with swapping your boot drive from a 3Gbps port to a 6Gbps one ? Or can I simply turn off my computer, swap ports, turn it on again, and go on as if nothing ever happened ?

 

[aigomorla]

Guys on OCN are speculating about EVGA knowing about this issue and their delaying of releasing mobos for SB. lol

now i didnt want to make this public but..

there delayed (i think) because peter (shimano) and all the other good designers left EVGA.

 

[dangerman1337]

What is really cool of Intel here is that they have known about this issue for some time now...long enough in fact that they had time to debug it, find the root-cause, implement a candidate fix, confirm the candidate fix actually works, and have started full-scale production of the corrected chipsets.

This was the truly sad thing, Intel had clearly working 6 series chipsets since IDF San fransisco 2009 (the SB demonstration) and yet they didn't find this flaw.

 

[Axon]

Uhh... according the Anandtech article, this is not the case. Anandtech specifically states that Intel did not know the exact nature of the bug until last week. They didn't receive complaints until they had already produced 100K units first. Even then, they weren't able to reproduce and track down what the problem was until last week. This is why they do not have a replacements in place and won't until the end of February at the earliest. This is why they are expecting sales to HALT and lose them $300M in lost sales in total (meaning the lifetime of this product) and then spend $700M in replacing all the recalls.

I don't know, but this seems pretty ethical to me.


Even with all that, there is known workaround for the meantime. If your board has another sata controller or at least the sata 3 controllers, use those. If not, buy a sata card.

Well, Anand is just repeating the company line.

 

[Continuity28]

Are the any problems with swapping your boot drive from a 3Gbps port to a 6Gbps one ? Or can I simply turn off my computer, swap ports, turn it on again, and go on as if nothing ever happened ?

There wouldn't be any issues with that.

 

[HumblePie]

Well, Anand is just repeating the company line.

Uhh, from the circumstances and reactions revolving around this incident, seems legit to me. But you just go right on ahead and bust out your tin foil hat.