Intel Chipset Design Error!!!

does the bug affect all ports or just the SATA 2 ports?

because i have not had any issues with my 2 SSDs on the SATA3 ports
 
Eh, all P8P67 ATX size boards including the vanilla version have it. It seems some of the mATX have 3 6G ports. So I'd move over to the Marvel for precaution.

Well, I'm not 100% sure which mobos have what...but the one on anandtech site in the article about this topic doesn't appear to:

_DSC7088sm.jpg
 
according to FiXT from HWC, only the SATA 3Gb/s are affected and not the SATA 6Gb/s. Thank god!
http://www.hardware canucks.com/forum/press-releases-tech-news/40373-design-flaw-intel-6-series-chipset-halts-manufacturing-recalls-imminent-2.html#post480342
 
Well, I'm not 100% sure which mobos have what...but the one on anandtech site in the article about this topic doesn't appear to:

_DSC7088sm.jpg

P8P67 = Asus boards. That's a stock Intel board. Only the GB UD7 has an extra 2 SATA ports. All Asus ATX boards seem to have the 2 ports using the Marvel as well. MSI GD65 has it but I'm not sure if any other MSI boards do.
 
My MSI P67A-GD65 has 4 SATA III ports. I don't plan on using the SATA II ports so I probably won't go through the RMA process.
 
It's odd that over at anandtech they call this a "statistical error" instead of a design error, even though it's quite obvious that the change of metal layer(design change) does completely solve the problem. If it was a statistical problem that IIRC cannot be completely solved, changing the design would not have made a difference.
 
OMG what a bunch of weenies whining about a statistcal fault lol.....thought y'all were hardcore......lol
 
Christ so much dramatic crap in this thread. :rolleyes:

Believe it or not, it's not the end of the world. Your SB motherboard is not going to kill your dog, rape your mother and then take a shit on your bed...
 
OMG what a bunch of weenies whining about a statistcal fault lol.....thought y'all were hardcore......lol

i think they dont want to end up like Microsoft did with the red ring of death. I think alot of corporations learned from that mistake.
 
My MSI P67A-GD65 has 4 SATA III ports. I don't plan on using the SATA II ports so I probably won't go through the RMA process.

I can understand not wanting to going through the RMA...but I'd like to have the option of using All of the ports on my mobo...this is basically a mobo swap...a definite pain in the butt, though. But I was planning to over to a new case later anyway.
 
i think they dont want to end up like Microsoft did with the red ring of death. I think alot of corporations learned from that mistake.

they caught it early in production, they halted shipments, they setup a fund to take care of any potential problems, what else do people want?

crap happens and intel is dealing with it.........if a board has a problem, just rma it and wait for a new one, i am sure a few days without a computer won't kill anyone.
 
My plans for afterwork today were to go pickup an Asus Sabertooth P67, 2600K and two kits of Corsair Vengance for 16GB Memory. Not happening anymore, will be saving my monies. May its a sign to wait for X68 later this year.
 
I'm not particularly worried, it seems pretty clear that while performance might degrade on the SATA II's, it won't actually harm/corrupt your HDD or data. I've got 4 SATA III ports so i should be ok. I did notice though when I built the system that my OS HDD wasn't recognized by the SATA III controlled right off the bat.

Will I RMA later if given the option? Probably.
 
It's odd that over at anandtech they call this a "statistical error" instead of a design error, even though it's quite obvious that the change of metal layer(design change) does completely solve the problem. If it was a statistical problem that IIRC cannot be completely solved, changing the design would not have made a difference.

How do we know it completely solves the problem ? Sure, they *think* that is the cause, but until they get back working silicon, *and* test it, nobody is sure.

Early last week Intel duplicated and confirmed the failure in house.
So they knew about this for a few weeks (months?) and then about a week ago they confirmed it.

The SATA bug exists in hardware and there’s no way to provide a driver or firmware update that can fix it. The fix requires a metal layer change, which will result in a new hardware stepping (resulting in the ~3 week delay before replacement hardware is ready).

The question is, is a ~3 week delay enough time to retest the chip to make sure that the new change don't cause side effects in other parts of the chip.
That means, that this could be drawn out into June/July or longer depending on how well they test it this time, it would NOT be fun if they have a back-2-back recall.
 
So is the temporary solution to move my SSD and other hard drives to the SATA 6GB connections?

I hope RMA = a new replacement board and not a 'fixed' board that someone else has returned.
 
Christ so much dramatic crap in this thread. :rolleyes:

Believe it or not, it's not the end of the world. Your SB motherboard is not going to kill your dog, rape your mother and then take a shit on your bed...

QFT!

On top of the new MB/chipset introduction which always creates confusion, now you have another cause for the builder to RMA a MB.

If I were a "Gary Key" type, I'd be self medicating with heavy doses of narcotics and alcohol.

I've always found it works well....problems or not. :D
 
I *use* five SATA devices on my Sabertooth P67 Asus so the 4x SATA3 won't cut it. As a result I just picked up a Highpoint 620 PCI-E x1 SATA3 card with two more connectors. The hassle of a recall would be too much unless they do cross-ship replacements.
 
I'm not particularly worried, it seems pretty clear that while performance might degrade on the SATA II's, it won't actually harm/corrupt your HDD or data.
Actually according to Anandtech who got it from Intel, the performance degradation is caused by a high number of errors. It is these errors that cause the controllers to try resending the same data again and again, resulting in the performance degradation.
So there is a high risk of data corruption.
 
How do we know it completely solves the problem ? Sure, they *think* that is the cause, but until they get back working silicon, *and* test it, nobody is sure.


So they knew about this for a few weeks (months?) and then about a week ago they confirmed it.



The question is, is a ~3 week delay enough time to retest the chip to make sure that the new change don't cause side effects in other parts of the chip.
That means, that this could be drawn out into June/July or longer depending on how well they test it this time, it would NOT be fun if they have a back-2-back recall.

They duplicated the bug a week ago, they did not understand the cause. Duplicating the issue is the first step in troubleshooting, not the conclusion. It took them a couple days to figure out the cause and then some more time to come up with possible solutions.

Unfortunately for everyone they determined that only a replacement would suffice. I think its obvious they don't want to lose a billion dollars over this. Considering the ramifications, I think they got through the problem relatively quickly. I think the OEMs wish Intel had notified them first and given them at least 24 hours to come up with a support/rma story.
 
rumors say that the issue was not caught in-house at Intel but at an OEM...

Does that honestly matter? Intel already said the chipsets passed their standard testing procedures in-house and were deemed the chipsets worthy. The complete randomness of the issue was brought to their attention and they finally were able to reproduce the issue.

Sounds like an issue that could have easily happened to any manufacturer. I'm sure someone will B&M over Intel's testing/validation procedures and say they didn't do enough.
 
The question was posted a few pages back, but it was noted in Anandtech's article that a metal layer change was necessary in the silicon. That means that more likely than not this affects 100% of all SB chipsets and not just a batch's worth.
 
If this is just a Motherboard Issue why did Newegg Deactivate all the Sandy Bridge CPU's?
 
I can understand that but I would of still ordered my cpu but I do see your point now
 
Newegg pulled all the motherboards i had a biostar 1155 in my cart & its now gone completely off the site.
 
They arent... and the P67s are still for sale at newegg.

Try adding a mobo to your cart: You may see a message like: "...has been removed from shopping cart due to quantity limitation/insufficient stock. Some combo items might be affected by this."

Navigating the product pages for these mobos is also screwed up.
 
Christ so much dramatic crap in this thread. :rolleyes:

Believe it or not, it's not the end of the world. Your SB motherboard is not going to kill your dog, rape your mother and then take a shit on your bed...

Not the smoothest of statements, but it sums it up none the less.

I see more systems shipping with the Marvell controller being used as it is. And, most folks who think themselves serious computer users are using a 3rd party card anyways. The story isn't as sensational as the EMO kind would like. Just let them over react and get it out of their system.
 
With all SATA ports currently used on my board, the potential loss of half of them is not acceptable.

I'll be initiating an RMA from Tiger tonight. Then I'll go back to using the C2Quad until April/May.
 
Back
Top