bit-tech.net

Sandy Bridge chipset flaw halts board shipments

Sandy Bridge chipset flaw halts board shipments

The design issue affects SATA performance on Intel's P67 and H67 motherboard chipsets.

If you've managed to resist the urge to plunge into the speedy world of Sandy Bridge so far, then you may have made the right decision. Intel has just revealed that it's discovered a 'design issue' with its new 6-series motherboard chipsets, and has stopped shipments of all the offending chips.

In a statement, Intel says that the issue can potentially affect the 'performance or functionality of SATA-linked devices such as hard disk drives and DVD-drives.' Both Intel's P67 and H67 chipsets support up to six SATA ports; two 6Gbps ports and four 3Gbps ports.

According to the folks over at Anandtech, who spoke to Intel yesterday, the flaw only affects the 3Gbps ports, and Intel's statement simply says that the 'SATA ports within the chipsets may degrade over time.'

How long are we talking about here? We asked a spokesperson from motherboard manufacturer MSI, who told us that it looks as though 'the performance of some of the Intel SATA ports on Intel 6-series products can degrade over a period of three years.'

Intel says it's already developed a silicon fix for the chipset issue, which will be in the hands of motherboard manufacturers by the end of February. According to Intel, the Sandy Bridge motherboard business should be back in full flow in April.

That's a long time to wait, though, and while it's good to see that Intel has released a fix, this issue raises all sorts of questions for anyone who's already purchased a Sandy Bridge system. Intel promises that it 'will work with its OEM partners to accept the return of the affected chipsets,' and also says it 'plans to support modifications or replacements needed on motherboards or systems.'

However, as Intel is only the chipset manufacturer, it looks as though any decisions about recalling boards that have already been sold will be up to the motherboard manufacturers themselves. A spokesperson from a motherboard maker that wished to remain anonymous told us that 'there is likely to be a recall situation, and from my thinking that would be down to the specific vendor to put that in place. The likelihood is that Intel will bear some responsibility in that, or we hope so anyway.'

Thankfully, MSI appears to be tackling the situation diligently, and we hope other motherboard manufacturers will follow suit. 'MSI takes the quality of its products very seriously,' commented the company, 'so at this moment we are investigating product batches that may be affected and, meanwhile, have stopped all shipments to our distributors and resellers. Additionally, we've asked resellers to hold sales of the Intel 6-series-based MSI products to customers until the issue can be resolved.'

In terms of product recall, the company says that 'end-users who have already purchased MSI Intel 6-Series motherboards are advised to wait for our update highlighting the actual products affected. We are working closely with Intel to help identify this as soon as possible. For MSI notebook and all-in-one product lines, we can confirm that there are currently no products in the UK channel which are affected by this issue.'

The situation appears to have caught everyone in the motherboard chipset business by surprise, though. Intel UK has refused to comment until it gets more information from the US, as have other motherboard makers, and we may well have to wait a couple of days before we find out exactly what's going on.

Have you bought a Sandy Bridge system, and are you worried about this design issue? As always please share your thoughts in the forums.

88 Comments

Discuss in the forums Reply
Greenie 1st February 2011, 12:08 Quote
The problem only appears to affect the 3Gb/s SATA ports, as explained over at Anandtech: http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug
memeroot 1st February 2011, 12:14 Quote
bugger
The boy 4rm oz 1st February 2011, 12:17 Quote
I'm glad I read this and the article over at anandtech. I was going to head out later this month to upgrade. I will now wait for a birthday upgrade in April :D.
Hustler 1st February 2011, 12:18 Quote
Oh well, i suppose the AMD Bulldozer team will be smiling around the water cooler this morning, they have just been given a 3 month window to finalise their product....
Metwinge 1st February 2011, 12:19 Quote
I bought a 2600k with a fatal1ty P67 board from overclockers.co.uk. Gave them a call today to find out if there was a recall and they have said that they arnt recalling the boards unless the companies notify them too.

Like Greenie states its only affecting the 3Gb sata ports, so im really not to bothered but will happily replace it with a fixed board
Spraduke 1st February 2011, 12:24 Quote
At first I was a bit worried about it, but after reading around it seems it only affects sustained use of SATA 3gb/s ports. As i use the Sata 6 ports for my 2 HDDs and my 'only used for windows installation' CD drive on the 3gb/s I think I will live.

Saying that if the RMA process looks to be quick once the dust settles I may swap it out (get the latest rev of the UD4 at least)
Spreadie 1st February 2011, 12:27 Quote
*Sleeper agent to AMD mission control*
"Mission accomplished"
:D
arcticstoat 1st February 2011, 12:31 Quote
Quote:
Originally Posted by Greenie
The problem only appears to affect the 3Gb/s SATA ports, as explained over at Anandtech: http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug

Thanks - I've put in a link to the Anandtech article as well now.
Bad_cancer 1st February 2011, 12:33 Quote
Quote:
Originally Posted by Spreadie
*Sleeper agent to AMD mission control*
"Mission accomplished"
:D

:D

+1
Kasius 1st February 2011, 12:34 Quote
Huzarr, patience provales..
faugusztin 1st February 2011, 12:35 Quote
The real question is - did Intel fire the guy who decided to use that transistor at that spot ? :D
Floyd 1st February 2011, 12:36 Quote
Not worried. Ill just wait and hope there is a recall, this has been the most stable platform ive been on!
FelixTech 1st February 2011, 12:37 Quote
Quote:
Originally Posted by Spreadie
*Sleeper agent to AMD mission control*
"Mission accomplished"
:D

"Make sure you collect the 'protection' from ARM" :P
Devolve 1st February 2011, 12:48 Quote
I was actually going to buy a sandy bridge upgrade today. Good job bit-tech is my homepage. Oh well, I may have enough for a full system overhaul by then.
BradShort 1st February 2011, 12:51 Quote
oh bollards!! Already had issues with the P67 chipset and sent a MB back to a reseller (who are trying to wriggle out of it). I usually wait to upgrade but i jumped on board the early adopter intercity to hell this time :)
skipperoo 1st February 2011, 13:11 Quote
Well that explains why my order from Scan hasn't arrived today. Looks like I got lucky and ordered it just late enough for them to hold it back till this situation is sorted.
Hustler 1st February 2011, 13:13 Quote
Quote:
Originally Posted by BradShort
I usually wait to upgrade but i jumped on board the early adopter intercity to hell this time :)

Which is why i'll be waiting for other people to beta test the new Bulldozer CPU's for 3mths after it launches in Q2 this year....
frontline 1st February 2011, 13:13 Quote
Phew, i actually had a motherboard, cpu and RAM in a basket ready to puchase a couple of times but managed to resist the temptation to see how the pricing panned out in the next few months.

Not great news for Intel, but at least it has been spotted early.
wuyanxu 1st February 2011, 13:16 Quote
Quote:
Originally Posted by faugusztin
The real question is - did Intel fire the guy who decided to use that transistor at that spot ? :D
the question should be, did Intel fire the manager who made design decision to re-use that cluster of transistors from older Intel chipsets.

the poor engineer was probably just doing what he's been told. i would blame poorly documented cell library rather than the engineer, because in IC design, all you are seeing are blocks, not how the transistor is made.


(but then, im just a poor engineer doing what i've been told. so from the bottom looking up, i find holding a single workhorse engineer responsible is wrong)
glendronach 1st February 2011, 13:16 Quote
Wait for a recall??
Now it is out in the public domain that the chipsets are faulty I would advise everyone in the UK who has one of the affected motherboards to return it to the supplier using the relevant consumer law (Sale of Goods Act) and request a full refund.
[USRF]Obiwan 1st February 2011, 13:20 Quote
instead of 300 mil dollar to reclaim the boards. Why not give away pcie sata cards. Much cheaper solution
wuyanxu 1st February 2011, 13:22 Quote
Quote:
Originally Posted by [USRF]Obiwan
instead of 300 mil dollar to reclaim the boards. Why not give away pcie sata cards. Much cheaper solution
that's Apple's way of half-arsed fixing thoughs.

also, don't forget rubber bumpers to cover up the SATA ports in question.
Metwinge 1st February 2011, 13:23 Quote
Quote:
Originally Posted by glendronach
Wait for a recall??
Now it is out in the public domain that the chipsets are faulty I would advise everyone in the UK who has one of the affected motherboards to return it to the supplier using the relevant consumer law (Sale of Goods Act) and request a full refund.

Ofc i could but i think the problem is minor as on my blu ray drive is using the 3gb sata and i need a computer so im happy to wait.
maximus09 1st February 2011, 13:28 Quote
*phew* and I was going to upgrade this month! Will wait now until it is all resolved. This will definatly effect people who are "horders" and use more than 2 HDs, like me, so not everyone will need a replacement. Although it is a bit surprising that this has happened and the fauly not only slipped passed Intel but also the MB manufacturers, hhhmmm so much for quality control!!
GravitySmacked 1st February 2011, 13:30 Quote
Not great news for Intel, somewhat craps on the launch of some fantastic chips.
t5kcannon 1st February 2011, 13:30 Quote
Quote:
Originally Posted by faugusztin
The real question is - did Intel fire the guy who decided to use that transistor at that spot ? :D

haha id not like to be in his shoes xd

SB seems by far the best system out there. With a fix, I'm most likely to upgrade to it. From what Ive read, it's important to realize exactly what the issue is. Anandtech summed it up well:

Interestingly enough the problem doesn’t affect ports 0 & 1 on the 6-series chipset. Remember that Intel has two 6Gbps ports and four 3Gbps ports on P67/H67, only the latter four are impacted by this problem.
Xir 1st February 2011, 13:51 Quote
If they put a too-thin gate on a relatively unimportant transistor like this one, you do have to wonder how the rest of the gates will hold up.
Then again gate thickness has been on the cutting edge for decades, and so far it worked out fine.
Bungletron 1st February 2011, 14:09 Quote
This is huge! Bold move going public, even before reading the anandtech article I guessed this must have something to do with high leakage from the symtoms, problem is a flaw like this was always just a matter of time as dies get smaller and transistors need to be more complex.
red4our 1st February 2011, 14:58 Quote
Uhhm, Spanish e-tailers are still happy to sell you a board. They either haven't picked up on this or haven't been advised to stop selling these boards for the time being by the manufacturers. Perhaps mañana! This could get a bit messy.
Hakuren 1st February 2011, 15:08 Quote
ROFL at Intel. X58 all the way until X68... in 4thQ of 2011.
digitaldunc 1st February 2011, 15:12 Quote
FUUUUUUUUU

I just got a full SB system plumbed in yesterday with dual loops.

Oddly my 1TB drive was giving off a noise which I thought was the click of death -- connected it to a 3Gbps port and it stopped and appears to be OK, though I've not had chance to do any more in depth investigation. SMART info appears to be OK, though I've read that this isn't necessarily a definitive indicator.

I think I can live using 6Gbps only -- I'll maybe try the drive in another 6Gbps port (I believe the P8P67 has more than one 6Gbps controller)

Schaedenfraude, anyone?
BradShort 1st February 2011, 15:28 Quote
I think i will register my board with Gigabyte an await their response. It has a years warranty anyway. If i return it, i have a 2500k that i cannot use . .
flipman 1st February 2011, 15:45 Quote
well in 3 years time i will have the newest board by then i upgrade every year so i dont see myself returning my board back to gigabyte i love the performance of sandy bridge over my pre built that was a core 2 quad system that had problems of its own :)
ADJB 1st February 2011, 15:52 Quote
Interesting review of an i7 SB whitebook over at The Register (http://www.reghardware.com/2011/02/01/review_laptop_sandybridge_whitebook/) which appears to show another, possibly related, oddity regarding the SATA ports.
Cyberpower-UK 1st February 2011, 15:55 Quote
Undervolt the PCH to 0.75-0.95, not a guaranteed fix but is stable and should .
Use the SATA III rather than the Intel SATA II ports.
Set up a back up and you should be OK.
Cyberpower-UK 1st February 2011, 15:56 Quote
*should last longer*
DC74 1st February 2011, 16:23 Quote
Oh well the month or so i'll have to wait to rebuild my system should see an extra £150 for my rebuild pot, guess i'll be going i7 instead of the i5 then.
Raptorkid 1st February 2011, 16:34 Quote
-Double post-
Raptorkid 1st February 2011, 16:37 Quote
Goddamnit - the one time I'm an early adopter of anything even vagely silicon related, and this happens. I wonder if that's why my motherboard wouldn't recognise my HDD at first...
Quote:
Originally Posted by digitaldunc
FUUUUUUUUU

I just got a full SB system plumbed in yesterday with dual loops.

Oddly my 1TB drive was giving off a noise which I thought was the click of death -- connected it to a 3Gbps port and it stopped and appears to be OK, though I've not had chance to do any more in depth investigation. SMART info appears to be OK, though I've read that this isn't necessarily a definitive indicator.


I noticed that when I first built my SB system, although mine was on a 6Gbps port.
Madness_3d 1st February 2011, 17:20 Quote
Asus are telling us Retailers, carry on getting your customers to use their boards at least until after chinese new year when they'll know the timeframe their looking at for rolling out these new boards.
Waynio 1st February 2011, 17:51 Quote
Quote:
Originally Posted by Spreadie
*Sleeper agent to AMD mission control*
"Mission accomplished"
:D

LOL :D.

But ouch I got the MSI P67A-GD65 & use all of the sata ports :(.

But this quote put me at ease :D.
Quote:
Thankfully, MSI appears to be tackling the situation diligently, and we hope other motherboard manufacturers will follow suit. 'MSI takes the quality of its products very seriously,' commented the company, 'so at this moment we are investigating product batches that may be affected and, meanwhile, have stopped all shipments to our distributors and resellers. Additionally, we've asked resellers to hold sales of the Intel 6-series-based MSI products to customers until the issue can be resolved.'

In terms of product recall, the company says that 'end-users who have already purchased MSI Intel 6-Series motherboards are advised to wait for our update highlighting the actual products affected. We are working closely with Intel to help identify this as soon as possible. For MSI notebook and all-in-one product lines, we can confirm that there are currently no products in the UK channel which are affected by this issue.'
schmidtbag 1st February 2011, 17:58 Quote
maybe if intel wasn't such an asshole toward nvidia not everyone who bought a SB mobo would have this problem. seriously, this is why ibm wanted 2 companies to manufacture x86 from the beginning. that and to raise competition.
chrisb2e9 1st February 2011, 18:21 Quote
I wonder if Intel will hold off on making cpu's until april then? to avoid flooding the market.
Aracos 1st February 2011, 18:26 Quote
Quote:
Originally Posted by schmidtbag
maybe if intel wasn't such an asshole toward nvidia not everyone who bought a SB mobo would have this problem. seriously, this is why ibm wanted 2 companies to manufacture x86 from the beginning. that and to raise competition.

Wuh wut? Dunno if you've heard about a little company called Advanced Micro Devices? They do x86 too...
frontline 1st February 2011, 18:42 Quote
Could this be related to the last minute inclusion of USB3 into the chipset and the possible diversion of resources away from more thorough testing of other chipset/motherboard components? http://www.bit-tech.net/news/hardware/2010/09/07/intel-usb3-will-be-in-sandy-bridge-chipsets/1
tad2008 1st February 2011, 18:53 Quote
Another good reason why Intel should stick to CPU's and well away from Chipsets and Graphics!
dudwasup 1st February 2011, 18:55 Quote
wow im glad i read this later this week i was going to pick up one
schmidtbag 1st February 2011, 19:01 Quote
Quote:
Originally Posted by storm20200
Quote:
Originally Posted by schmidtbag
maybe if intel wasn't such an asshole toward nvidia not everyone who bought a SB mobo would have this problem. seriously, this is why ibm wanted 2 companies to manufacture x86 from the beginning. that and to raise competition.

Wuh wut? Dunno if you've heard about a little company called Advanced Micro Devices? They do x86 too...

yes, amd makes their own chipsets, but for a while nvidia made chipsets for intel (while intel also made their own) and then intel basically told them to stop doing it, and when you have 1 company do all the work, something like this happens.
Initialised 1st February 2011, 19:08 Quote
I heard that ASUS will shortly announce full blown product recall.
mecblade 1st February 2011, 19:15 Quote
Ideal Situation:

Intel CEO: I've got an idea!
Intel Employee: What is it?
Intel CEO: Let's 'leak' it that our chipsets 3Gbs ports degrade.
Intel Employee: ...... Sir, you do realise April is still not among us yet? It's like another 2 months away.
Intel CEO: Meh, let's do it.
Intel Employee: ....... Ok then...... *starts typing*

On a more serious note, i feel sorry mainly for those who watercooled their SB rigs, if they wish to RMA their motherboard, they have to drain it all and take their blocks off. Still, by the looks of it if it's usable and wont degrade performance on the HDD, then it's ok to leave it temporarily.
penryn 2 hertz 1st February 2011, 19:16 Quote
you dont get the sort of problems from amd allways seems to be intel you see thats what you get for rushing tut tut ...
Sloth 1st February 2011, 19:19 Quote
Quote:
Originally Posted by schmidtbag
yes, amd makes their own chipsets, but for a while nvidia made chipsets for intel (while intel also made their own) and then intel basically told them to stop doing it, and when you have 1 company do all the work, something like this happens.
If you take X number of products with Y% chance of mishaps then wait Z time you'll find something. X is a known value (record of everything Intel's manufactured), Z is a known value (Intel's inception up to present time), Y can be derived using the two former values. The values may change but it applies to everyone, regardless of whether it's just one company or multiple. It's fairly safe to say that Intel would have still made this design error even if there were 3rd party chipset manufacturers for 1155.

Where you're getting your analysis mixed up is looking only at chipset providers for one of the two major x86 manufacturers. Yes, if Nvidia was able to produce an 1155 chipset then users would have an alternative for the 1155 socket, not for all of x86. AMD provides an alternative, as storm has already pointed out.
penryn 2 hertz 1st February 2011, 19:19 Quote
Quote:
Originally Posted by tad2008
Another good reason why Intel should stick to CPU's and well away from Chipsets and Graphics!

could not agree more ...
mecblade 1st February 2011, 19:22 Quote
Quote:
Originally Posted by penryn 2 hertz
you dont get the sort of problems from amd always seems to be intel you see thats what you get for rushing tut tut ...

yeah...AMD has NEVER had to recall ANYTHING.....yeah.....

I recall (not intended to coincide) the old Opteron Single Core processors on the E4 stepping revision were recalled in 2006.

http://en.wikipedia.org/wiki/Opteron#Opteron_recall
Sloth 1st February 2011, 19:26 Quote
Quote:
Originally Posted by mecblade
yeah...AMD has NEVER had to recall ANYTHING.....yeah.....

I recall (not intended to coincide) the old Opteron Single Core processors on the E4 stepping revision were recalled in 2006.

http://en.wikipedia.org/wiki/Opteron#Opteron_recall
See my post above, it's only math! Funny how people's memories change depending on the brand in question. :D
jcb121 1st February 2011, 19:34 Quote
I don't feel so bad buying a 1156 system a week or so 1155 came out now :)
schmidtbag 1st February 2011, 19:50 Quote
Quote:
Originally Posted by Sloth
Quote:
Originally Posted by schmidtbag
yes, amd makes their own chipsets, but for a while nvidia made chipsets for intel (while intel also made their own) and then intel basically told them to stop doing it, and when you have 1 company do all the work, something like this happens.
If you take X number of products with Y% chance of mishaps then wait Z time you'll find something. X is a known value (record of everything Intel's manufactured), Z is a known value (Intel's inception up to present time), Y can be derived using the two former values. The values may change but it applies to everyone, regardless of whether it's just one company or multiple. It's fairly safe to say that Intel would have still made this design error even if there were 3rd party chipset manufacturers for 1155.

Where you're getting your analysis mixed up is looking only at chipset providers for one of the two major x86 manufacturers. Yes, if Nvidia was able to produce an 1155 chipset then users would have an alternative for the 1155 socket, not for all of x86. AMD provides an alternative, as storm has already pointed out.

all of you are COMPLETELY missing the point i'm trying to make - i'm not saying that intel is the only chipset maker. i'm not saying that if nvidia made intel chipsets that intel is less likely to make a mistake. amd is completely irrelevant to this discussion since they didn't screw up their chipset and they didn't tell anybody to stop making chipsets for them. what i've TRYING to say is if intel didn't tell nvidia to stop making chipsets, then people would have an alternative chipset to turn to in this particular situation. HOWEVER, since intel is the only chipset manufacturer for SB, people with these CPUs get this problem, but if nvidia was still allowed to make chipsets, then they wouldn't be affected. i don't see how this is such a difficult concept to grasp.

altho intel and amd (and via) create x86 CPUs, their designs in their processors and chipsets are so drastically different that even if you made a pin compatible amd cpu to fit in an intel board or vise versa, it won't work. thats why you can emulate an efi and you still can't get mac to install on almost any amd machine. like i said before, i deliberately didn't bring up amd because it isn't relevant to this situation.
centwing 1st February 2011, 19:53 Quote
Oh damn, I knew there was something fishy when it says *no return* on the invoice when I bought msi p67 mobo last week. I was lured by the seeming great discount offer, sigh...
mremery 1st February 2011, 19:57 Quote
I was going to buy a P67 board in the very near future (after reading the next CPC of course) as I bought a 2600k last night but now I don't want to get one, put my plumbing in and have the feck!n HDD's not work. then have to RMA it.

(just checked OCUK)
OCUK don't seem to see it as much of an issue, if you search for a P67 board (P8P67 pro in this case) they have written this:

- Potential Sandybridge issue:
- In some cases, motherboards featuring the 6 Series Chipset (P67 & H67) that have Sata devices connected to ports 2-5 could potentially degrade over time.
- Motherboards with devices connected to Ports 0 & 1 remain unaffected or using the 6Gb/s ports.
- Please note if your board is affected by the above issues, your statutory rights remain to RMA the board.
- Tests have found that under 5% of boards are affected.
- PLEASE NOTE THERE IS NO POTENTIAL ISSUE TO SANDYBRIDGE PROCESSORS
Guinevere 1st February 2011, 19:57 Quote
Quote:
Originally Posted by penryn 2 hertz
you dont get the sort of problems from amd allways seems to be intel you see thats what you get for rushing tut tut ...

Yet, with AMD you've still had boards with hardware / cpu / chipset compatibility issues. I had an athlon board with a god-awful opti usb chipset that was just one of it's many failings. Friends who had the same board had the same issues.

At least these days these things tend to get picked up and dealt with. Better to know there's an issue and have the law on your side to get it fixed than suffer in silence with a faulty bit of kit while the manufacture plays dumb

It used to be like this:

In the mid and late 90s, back when Sony still made CRT TVs in Wales, they manufactured a batch of faulty teletext chips. They knew they were unreliable, but the cost of the fix was too great so they kept installing them for years in all the TVs. Every single one of them had a bad chip and would randomly overlay a frame of random symbols, sometimes a few times an hour sometimes not for months.
penryn 2 hertz 1st February 2011, 20:02 Quote
Quote:
Originally Posted by mecblade
Quote:
Originally Posted by penryn 2 hertz
you dont get the sort of problems from amd always seems to be intel you see thats what you get for rushing tut tut ...

yeah...AMD has NEVER had to recall ANYTHING.....yeah.....

I recall (not intended to coincide) the old Opteron Single Core processors on the E4 stepping revision were recalled in 2006.

http://en.wikipedia.org/wiki/Opteron#Opteron_recall

see point only one thing remember intels recall of a defective line of Pentium chips ...
Sloth 1st February 2011, 20:23 Quote
Quote:
Originally Posted by schmidtbag
all of you are COMPLETELY missing the point i'm trying to make - i'm not saying that intel is the only chipset maker. i'm not saying that if nvidia made intel chipsets that intel is less likely to make a mistake. amd is completely irrelevant to this discussion since they didn't screw up their chipset and they didn't tell anybody to stop making chipsets for them. what i've TRYING to say is if intel didn't tell nvidia to stop making chipsets, then people would have an alternative chipset to turn to in this particular situation. HOWEVER, since intel is the only chipset manufacturer for SB, people with these CPUs get this problem, but if nvidia was still allowed to make chipsets, then they wouldn't be affected. i don't see how this is such a difficult concept to grasp.

altho intel and amd (and via) create x86 CPUs, their designs in their processors and chipsets are so drastically different that even if you made a pin compatible amd cpu to fit in an intel board or vise versa, it won't work. thats why you can emulate an efi and you still can't get mac to install on almost any amd machine. like i said before, i deliberately didn't bring up amd because it isn't relevant to this situation.
It's a very easy concept to grasp. What Storm and I have said is that it's a limited view. It doesn't matter whether or not a certain socket has working chipsets, you can just use a different socket. It's already covered. Your argument only applies if you must have that socket.

Consider how many other things could have failed and caused 1155 systems to be unusable. In this case it's the SATA 2 controller. What if it was the memory controller packaged with the CPU? Obviously no option of multiple suppliers there. What if the SATA 2 controller was on chip with the CPU? What if the architecture itself was flawed? What if Nvidia's chipset had a probvlem too? It doesn't particularly matter. In this case could a second chipset provider have allowed an alternative? Sure. But it really doesn't matter because there's still various other Intel and AMD sockets.
Eiffie 1st February 2011, 20:33 Quote
This is a real shame, everything about sandy bridge seems to be such a solid buy, the price vs what you get is very nice and theres a good selection of motherboards for ATX and M-ATX users from what I've seen. While I don't plan on buying one to replace my desktop PC, over the past few weeks I've been thinking about picking one up and starting to make a media center PC that can handle HD movies and do some light gaming, more specifically I wanted to try a M-ITX build but there is a lack of p67 M-ITX boards on the market so I was gonna settle for one of the h67's but now that you can't find a single sandy bridge motherboard on newegg it looks like that whole plan got put on hold. If the situation doesn't improve I might just wait for the next gen of AMD CPUs and hope for the best. I was really looking forward to trying something new and doing a mini build but sadly it looks like that wont happen for a while.
Initialised 1st February 2011, 20:37 Quote
http://www.gigabyte.com/press-center/news-page.aspx?nid=984
Quote:
Originally Posted by gigabyte
Notice: Motherboards Based on Intel® 6 Series Chipsets
GIGABYTE Acknowledges Intel’s Alert for Regarding 6 Series Chipset
2011/02/01
Taipei, Taiwan – February 1, 2011 – GIGABYTE TECHNOLOGY Co., Ltd, a leading manufacturer of motherboards, graphics cards and other computing hardware solutions, today announced that GIGABYTE has been alerted by Intel about a 6 series chipset design error. GIGABYTE is working closely with Intel® to minimize the inconvenience to customers and retail stores with regard to related issues.

A recent statement from Intelindicates that they expect full 6 series chipset volume recovery in April, 2011. GIGABYTE will produce and deliver new motherboards with the updated Intel 6 series chipset when the new chipsets become available. GIGABYTE prides itself in having the highest service quality in the industry, and as such we will resolve this issue with minimal impact on our customers.

Additional information will be added to the official motherboard websiteand the GIGABYTE Tech Daily blogas the situation develops.
woof82 1st February 2011, 23:12 Quote
Quote:
Originally Posted by Greenie
The problem only appears to affect the 3Gb/s SATA ports, as explained over at Anandtech: http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug
Quote:
Originally Posted by arcticstoat
Intel has stopped shipments of its 6-series chipsets, because of a SATA design flaw.

http://www.bit-tech.net/news/hardware/2011/02/01/intel-sandy-bridge-chipset-flaw/1

Ah, well this would explain why I can't access my hard drive.
:(
[USRF]Obiwan 2nd February 2011, 06:59 Quote
You guys can rant for hours about Intel making a mistake. Fact of the matter is, if Intel did not say anything about this design flaw. Nobody would have known and would be happily running their PC for a few years. But some people would have trouble with their Sata300 ports at some point in the future or when putting more voltage to the chip-set. (aka over-clocking)
mecblade 2nd February 2011, 07:35 Quote
Quote:
Originally Posted by penryn 2 hertz
see point only one thing remember intels recall of a defective line of Pentium chips ...

I too remember that, but the fact is you seem to be stating these problems never hit AMD. I was merely reminding you that AMD has recalled products before. ;)

Ironically, the response to the Pentium flaw which intel initially refused to refund put the media spotlight on intel and combined with a uprise in advertising, propelled it to a household name :)
V3ctor 2nd February 2011, 08:32 Quote
It's a storm in a glass of water... I have one of those chipsets and I'm not worried, when the recall begins I'm sure that my board (2 months old at the recall time) will not have any flaw... It's going to degrade over 3 years, there could be one or another that dies before that, but I think that the bug was found very quickly, and it doesn't concern me at all.
Cyberpower-UK 2nd February 2011, 10:03 Quote
[quote=GigaByte, Posted by Fugger on XS:]GIGABYTE technology received an official announcement from Intel indicated the 6 series chipest was concerned designed error, due to 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.
Based on a long time usage may considered causes the performance problem of SATA 2.0, for the originally native chips and additional SATA3.0 of 6 series, is still stably function and worry-free.

This is a situation happened in all motherboard manufacturers. On behalf of the NO.1 motherboard reputation, the GIGABYTE Technology will take its responsibility for worldwide customers and partners.
We have suspended all the shipping of 6 series B2 stepping based motherboard.
For avoid this impact, GIGABYTE will co-operate with Intel’s B3 stepping schedule to re-new our new 6 series motherboards to the market. And absorb the cost regarding to this product exchange.

Announcements:

GIGABYTE have stopped shipping out and selling all Intel 6 series B2 stepping based motherboards.
All the Intel 6 series B2 stepping based motherboards stocks are retrieve for all disti, and dealers.
Any sold Intel 6 series B2 stepping based motherboards are accept retrieve in all conditions based on B2 stepping problem.
The Intel new 6 series B3 stepping chipset supply schedule will be in April. GIGABYTE will corporate with Intel exact production schedule, to provide the B3 stepping products to our customer.
The actual cost for retrieving Intel 6 series B2 stepping, all dealers and disti must to keep the receipts indicate the descriptions for further claim approval and internal check.
GIGABYTE will be using brand-new motherboard based on Intel 6 series B3 stepping to produce new 6 series motherboards.



GIGABYTE is sincerely taking its action providing dealers and distis to make all the product exchange for this issue.
And also request all dealers stop selling any B2 stepping base motherboards from now on, and to assist customers how to choose and use the normal SATA 3.0 connector, to avoid SATA 2.0 problems it may be caused. [/url]

Basically telling customers to use SATA 3 where possible until a replacement is available. Expect ASUS, MSI et al to follow suite with similar statements later today.
D B 2nd February 2011, 12:14 Quote
well, glad I didnt have the money to upgrade yet

I was kinda surprised that my Yahoo news feed informed me of this a day or so sooner than BT had :shrug:
LatinSpirit 2nd February 2011, 15:50 Quote
Just when I have a delivery date on my 2500k (the last bit I needed to get my new build done!) I get this news :( :( :(

And as someone else said earlier, this is also my first time adopting new technology on release.

I haven't even got round to getting my P8P67 out of its anti-static bag yet......I don't know whether to finish building it or not, even if it is for a month or two so I can test the CPU and RAM, and then exchange the board.

I wonder if the returns process will be? Through Asus or through the e-tailer?
Podge4 2nd February 2011, 16:04 Quote
Found this clip on youtube. Thought it was well funny. Hitler gets informed of the SB fault.

http://www.youtube.com/watch?v=6yEQSX4biR4
skipperoo 2nd February 2011, 16:14 Quote
Just had an email saying my order has now been dispatched and apologies for the delay. Not sure if that means the faults been given the all clear to Scan by Intel?
Krayzie_B.o.n.e. 2nd February 2011, 16:32 Quote
degrades in 3 years!! by that time I will have upgraded my mobo twice. I'm gonna keep my Asus p8p67 until April then buy another one. I've tasted the sugar of 2500k and I'm not going back.

So until then I got 2 months to torture this board before I send it off on a "Warranty Issue" HA ha.
Fingers66 2nd February 2011, 16:46 Quote
Quote:
Originally Posted by Krayzie_B.o.n.e.
degrades in 3 years!! by that time I will have upgraded my mobo twice. I'm gonna keep my Asus p8p67 until April then buy another one. I've tasted the sugar of 2500k and I'm not going back.

So until then I got 2 months to torture this board before I send it off on a "Warranty Issue" HA ha.

I think this is the best course of action, especially if Asus say that they will RMA any SB rev B2 board whether it is working or not (like Gigabyte have already done).

Being without a rig until sometime in April/May would be a pain.
thelaw 2nd February 2011, 17:47 Quote
Quote:
Originally Posted by skipperoo
Just had an email saying my order has now been dispatched and apologies for the delay. Not sure if that means the faults been given the all clear to Scan by Intel?

SCANS LATEST UPDATE

Sandy Bridge Important Accouncement

We have just had Important Update information from Intel, Asus, Msi and Gigabyte 16:30Hrs Today - 01/02/11.

This is an agreement Scan has reached with Intel/Asus/Msi & Gigabyte and is not for the time being a general to all vendors. We suggest you check with your Vendor/Supplier.

All of the above parties have agreed on a intermediate solution to the Sandybridge problem.

As there is No Immediate Danger/Fault and as this fault is likely to affect a small number of boards over time and Replacement Stock will Not be available until realistically in April/May, all the parties above have Guaranteed Direct Swap Out of all P67/H67 Boards bought from Scan should you the end user wish to do so at a later date.

Based on this Swap Out Guarantee We have decided to make all P67/H67 Boards available for sale again.

The Decision to buy and use still lies upon you the end user customer.
BradShort 2nd February 2011, 18:09 Quote
Quote:
Originally Posted by thelaw
SCANS LATEST UPDATE

Sandy Bridge Important Accouncement

We have just had Important Update information from Intel, Asus, Msi and Gigabyte 16:30Hrs Today - 01/02/11.

This is an agreement Scan has reached with Intel/Asus/Msi & Gigabyte and is not for the time being a general to all vendors. We suggest you check with your Vendor/Supplier.

All of the above parties have agreed on a intermediate solution to the Sandybridge problem.

As there is No Immediate Danger/Fault and as this fault is likely to affect a small number of boards over time and Replacement Stock will Not be available until realistically in April/May, all the parties above have Guaranteed Direct Swap Out of all P67/H67 Boards bought from Scan should you the end user wish to do so at a later date.

Based on this Swap Out Guarantee We have decided to make all P67/H67 Boards available for sale again.

The Decision to buy and use still lies upon you the end user customer.

How can they continue to sell a board that has been acknowledged as not fit for purpose by the manufacturers? The interesting point is that Scan's announcement says "Direct Swap Out" not vendor . . . .

That is very surprising!!
thelaw 2nd February 2011, 18:40 Quote
Quote:
Originally Posted by BradShort
How can they continue to sell a board that has been acknowledged as not fit for purpose by the manufacturers? The interesting point is that Scan's announcement says "Direct Swap Out" not vendor . . . .

That is very surprising!!

Likely because waiting till Late March/April will kill alot of business off if they could not continue to trade in some format...its not like they are going to go back to selling 1156/1336 kits, this way computer stores/suppliers can continue to trade knowing they can swap boards out direct with the manufactures later at little/no cost and it allows them to continue to sell sandybridge.

I am a prime example i am going to sink £950 into a new sandybridge system next month including memory and other bits like Psu's/case etc - but was not going too unless i either saw the new boards out or had some sort of assurance i could swap out the board later....if this was not on offer i would hold off till april when the new boards were out as others had also mentioned in the forums and elsewhere...so retailers would miss out on alot of trade in the meantime and not just with sandybridge kit...
lexloofah 2nd February 2011, 19:15 Quote
Can devices be plugged into 4 sata 6Gbps at the same time? Would this then avoid the whole issue?
red4our 2nd February 2011, 19:27 Quote
Quote:
Originally Posted by red4our
Uhhm, Spanish e-tailers are still happy to sell you a board. They either haven't picked up on this or haven't been advised to stop selling these boards for the time being by the manufacturers. Perhaps mañana! This could get a bit messy.

edit 2/2/11 - Yay, mañana has arrived and Spain has caught up. All retailers here have pulled their 1155 mobos
kosch 2nd February 2011, 21:08 Quote
Quote:
Originally Posted by Spreadie
*Sleeper agent to AMD mission control*
"Mission accomplished"
:D

I've been trying to start AMD sabotage rumors on other forums :p
ssj12 3rd February 2011, 07:18 Quote
wow... this site is massively behind the news... dear god I saw this on Nordic Hardware a week ago.
Also the announcement of the GTX590, which has yet to be posted here.
frontline 3rd February 2011, 08:29 Quote
Statement from Asus:
Quote:
Taipei, Taiwan (February 1, 2011) — Intel® on January 31, 2011 announced the detection of a design error in the new Sandy Bridge-based Intel® 6 Series support chip, also known as Cougar Point. The shipment of existing Sandy Bridge products has been suspended by Intel®, and the production of an updated support chip has commenced. The design error affects only the support chip, and not 2nd generation Intel® Core™ processors such as the i5 and i7.

Acting on our philosophy and promise of inspiring innovation and persistent perfection, we've created a comprehensive response to this development to safeguard the best interest of ASUS customers around the world. We will provide total warranty services to maximize computing and usability, as each ASUS product is delivered with an uncompromising quality pledge. This includes the hassle-free return and/or replacement of all affected ASUS products. We’re doing all this effective immediately – and as a first step, shipments of current ASUS Sandy Bridge-based products have been halted. The suspension covers all distribution and retail channels — ASUS only sells quality-assured products, at all times and across all product segments.

For all customers who already own any ASUS product affected by the design error identified by Intel®, simply visit http://event.asus.com/2011/SandyBridge/notice for detailed warranty and service hotline information. We’ll be happy to address any questions or concerns you may have regarding this issue, as well as provide information on future updates to Sandy Bridge technology.

ASUS notebook and desktop products are covered by a warranty service that addresses the design error, with both swap and sales return offered as options to customers.

For motherboards, we’ve confirmed with Intel® that the Cougar Point design error does not affect SATA ports 0 and 1. Additionally, ASUS has been a pioneer in including extra SATA 3Gb/s and SATA 6Gb/s ports beyond Intel® specifications on a wide range of motherboards. These ports are not affected by this design error, offering customers an option for uninterrupted computing. However, we’re also offering you the option of returning any possibly affected motherboard for a swap or sales return. All warranties for affected motherboards will be reset to give you enough time to carry out any exchanges.

ASUS has also started an initiative to contact customers regarding this issue, and we’ll cover all channel partners, distributors and registered buyers of affected ASUS products. As mentioned above, you may opt to return any affected products to ASUS for a swap or sales return.

We’ll post updates through our official website and other channels regarding any developments or fixes to Sandy Bridge products. Thank you for your understanding, support and cooperation, and we apologize for any inconvenience this may have caused you.

Sincerely,

ASUS
thelaw 3rd February 2011, 10:04 Quote
I dont get it, if Asus are stating

"For motherboards, we have confirmed with Intel® that the Cougar Point design error does not affect SATA ports 0 and 1. Additionally, ASUS has been a pioneer in including extra SATA 3Gb/s and SATA 6Gb/s ports beyond Intel® specifications on a wide range of motherboards. These ports are not affected by this design error, offering customers an option for uninterrupted computing"ve

Why would they need to offer a recall why not say, "our stuff is made differently by us so its not affected"?
Fingers66 3rd February 2011, 10:26 Quote
Quote:
Originally Posted by thelaw
I dont get it, if Asus are stating

"For motherboards, we have confirmed with Intel® that the Cougar Point design error does not affect SATA ports 0 and 1. Additionally, ASUS has been a pioneer in including extra SATA 3Gb/s and SATA 6Gb/s ports beyond Intel® specifications on a wide range of motherboards. These ports are not affected by this design error, offering customers an option for uninterrupted computing"ve

Why would they need to offer a recall why not say, "our stuff is made differently by us so its not affected"?

Because Intel told them to? When you are the biggest gorilla of them all, people do what you say. :D

Who else thinks that rev 3 of the SB boards are going to be the most tested motherboards ever sold?
digitaldunc 3rd February 2011, 18:42 Quote
Quote:
Originally Posted by thelaw
I dont get it, if Asus are stating

"For motherboards, we have confirmed with Intel® that the Cougar Point design error does not affect SATA ports 0 and 1. Additionally, ASUS has been a pioneer in including extra SATA 3Gb/s and SATA 6Gb/s ports beyond Intel® specifications on a wide range of motherboards. These ports are not affected by this design error, offering customers an option for uninterrupted computing"ve

Why would they need to offer a recall why not say, "our stuff is made differently by us so its not affected"?

Because some of the boards have additional controllers that are presumably unaffected but still have the Intel ones.
seedleon 4th February 2011, 06:10 Quote
Kudos to Asus for their response. Luckily I haven't been affected by this Sandy Bridge fiasco but it must be an absolute PR nightmare for Intel to cope with.
shadow 25th February 2011, 09:41 Quote
gigabyte announced application form to register B2 mobo for exchange into B3:
http://uk.gigabyte.com/press-center/news-page.aspx?nid=988
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums