View Full Version : Intel Recall on 6-series, Cougar Point SATA, Sandy Bridge Chipset Recalled
OUCH!
http://www.engadget.com/2011/01/31/intel-finds-sandy-bridge-chipset-design-flaw-shipments-stopped/
SANTA CLARA, Calif.--(BUSINESS WIRE)-- 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.
I expect a run on these:
http://www.newegg.com/Product/Productcompare.aspx?Submit=ENE&N=40000410%20600022631%20600022673&IsNodeId=1&Description=sata%20card%20pci&bop=And&ShowDeactivatedMark=False&CompareItemList=410|16-115-072^16-115-072-TS,16-104-015^16-104-015-TS,16-132-030^16-132-030-TS
corduroygt
31-Jan-2011, 20:03
So Intel will send me a new DH67BL motherboard? It was a pain in the ass to mount my scythe big shuriken heatsink there, now I have to do it again?
So Intel will send me a new DH67BL motherboard? It was a pain in the ass to mount my scythe big shuriken heatsink there, now I have to do it again?
That's why I went and got a PCI-E SATA3 card...I hate mounting my Scythe Mugen 2.
But, yes, Intel will get mobo manufacturers to replace or repair motherboards.
I guess thats one of the good things about intel they arnt frightened to recall products, they dont do the 6 months of denial thing first
They anticipate lost revenue of $300 million and a cost to recall of $700 million. I don't know their current GM% but they're creeping up to a billion dollar mistake. Ouch.
If the error is down to 1 person wouldnt like to be him right now ;)
Steve Austin is jealous. He only cost $6 million.
http://techreport.com/discussions.x/20329
Ohhhhhhhhhhhhhhhhhhhh shit.
its the "degrade over time" bit i dont understand i thought chips didnt wear out no moving parts and all
I wonder if they are really serious about repairing existing boards. To actually remove and replace the chipset chips...
I wonder if they are really serious about repairing existing boards. To actually remove and replace the chipset chips...
who knows at this point...cost of removing and reinstalling the northbridge (plus really angry customer who has no rig for 2 week+) vs. whole new mobo?
My guess is that they'll get channel stock back to OEMs where the OEMs will replace the chipset with new ones and then they'll start replacing customer units with these reworked boards and use customer returns to make more reworked boards.
Expect genuinely "new" boards (no rework) around late April.
I guess thats one of the good things about intel they arnt frightened to recall products, they dont do the 6 months of denial thing first
Plus they give their products cool names like Cougar Point.
which reminds me of my 8th grade teacher
its the "degrade over time" bit i dont understand i thought chips didnt wear out no moving parts and all
The "degrade over time" bit pretty much means self-destruction of a portion of the chipset, much like nVidia's issues with GPUs (bumpgate) but due to different reasons.
From Anandtech's writeup (http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug):
The problem in the chipset was traced back to a transistor in the 3Gbps PLL clocking tree. The aforementioned transistor has a very thin gate oxide, which allows you to turn it on with a very low voltage. Unfortunately in this case Intel biased the transistor with too high of a voltage, resulting in higher than expected leakage current. Depending on the physical characteristics of the transistor the leakage current here can increase over time which can ultimately result in this failure on the 3Gbps ports. The fact that the 3Gbps and 6Gbps circuits have their own independent clocking trees is what ensures that this problem is limited to only ports 2 - 5 off the controller.
You can coax the problem out earlier by testing the PCH at increased voltage and temperature levels. By increasing one or both of these values you can simulate load over time and that’s how the problem was initially discovered. Intel believes that any current issues users have with SATA performance/compatibility/reliability are likely unrelated to the hardware bug.
The bad news for some of us is that the Marvell SATA3 benches about 30% lower than the Intel SATA2 circuit on Asus mobos with Intel SSDs...and the Intel SATA3, coupled with a SATA2 Intel SSD is ALSO SLOWER than the Intel SATA2 circuit. Ugh.
Intel believes that any current issues users have with SATA performance/compatibility/reliability are likely unrelated to the hardware bug.
so there are other problems with it as well ?
joker454
01-Feb-2011, 03:06
So if I'm reading this right:
Depending on the physical characteristics of the transistor the leakage current here can increase over time which can ultimately result in this failure on the 3Gbps ports. The fact that the 3Gbps and 6Gbps circuits have their own independent clocking trees is what ensures that this problem is limited to only ports 2 - 5 off the controller.
...means this only affects the sata2 ports, and not the sata3 ports? Would be cool as I'm not using the sata2 ports anyways.
This is one reason why Intel crushing VIAs and Nvidia's chipset businesses is bad, even for Intel themselves.
If they hadn't done that, they could at least have continued to sell processors while replacement southbridges pass through manufacturing and out into distribution and retail channels. Now sales will if not halt, so at least slow down substantially, potentially for months depending on how long they waited after discovering this errata before they announced its existence and issued the recall order...
digitalwanderer
01-Feb-2011, 03:55
Ooops. :oops:
corduroygt
01-Feb-2011, 04:20
My motherboard is made by Intel, I got it from Newegg, who will I be contacting?
I only have 2 HDD's and they're both connected to sata6 ports although they're sata3 drives. I don't foresee getting more hard drives that aren't usb (no more room) but I still don't want to be stuck with a faulty chipset.
I would contact Newegg, as they seem to have excellent customer service and at least be sure to get your RMA in early so as to limit the time you'd have to wait.
corduroygt
01-Feb-2011, 04:28
If the error is down to 1 person wouldnt like to be him right now ;)
Maybe he can meet up with the guy that was in charge of encrypting PS3 firmware up to version 3.55 :)
I agree with Grall (http://forum.beyond3d.com/member.php?u=515).
Pay back for not supporting third party chipsets for their cpus.
aaronspink
01-Feb-2011, 06:02
its the "degrade over time" bit i dont understand i thought chips didnt wear out no moving parts and all
All chips degrade over time, it is just that normally there is enough margin that no one ever notices. In fact EVERYTHING degrades over time. It is after all the nature of the universe.
aaronspink
01-Feb-2011, 06:04
This is one reason why Intel crushing VIAs and Nvidia's chipset businesses is bad, even for Intel themselves.
If they hadn't done that, they could at least have continued to sell processors while replacement southbridges pass through manufacturing and out into distribution and retail channels. Now sales will if not halt, so at least slow down substantially, potentially for months depending on how long they waited after discovering this errata before they announced its existence and issued the recall order...
you mean the Nvidia chipset which just straight ate your data from day one?
Funny you should say that... I have a NForce chipset in my old core2quad box, it's never eaten anything, nor caused any problems whatsoever really.
It doesn't like me mixing different revision of corsair memory sticks (same timings, but different voltages as it turned out after reading the specs more carefully...), but that's more like my fault I should think. Right now I'm stuck with 2GB instead of 4 after two of the original corsair sticks died on me, but I'm having a bit of a hard time seeing how I can blame Nvidia for that.
Sour grapes...? ;)
Silent_Buddha
01-Feb-2011, 17:16
Hmmmm, I wonder when the OEMs will be able to start processing and shipping replacement boards. Looks like another month at the very earliest.
The Anandtech article mentioned one fix is to scale back the voltage going to the transistor. I wonder if this is something that could be done through a BIOS update.
Regards,
SB
DarthShader
01-Feb-2011, 19:59
The Anandtech article mentioned one fix is to scale back the voltage going to the transistor. I wonder if this is something that could be done through a BIOS update.
Sure, if that single transistor has it's own controled voltage line. I doubt the rest of the chip would like the lower voltage otherwise. ;)
PS. Charlie wrote a lenghty peace on this issue. Suggested to have a tinfoil hat nearby while reading. But some of the points raised are interesting though.
aaronspink
01-Feb-2011, 20:21
Funny you should say that... I have a NForce chipset in my old core2quad box, it's never eaten anything, nor caused any problems whatsoever really.
It doesn't like me mixing different revision of corsair memory sticks (same timings, but different voltages as it turned out after reading the specs more carefully...), but that's more like my fault I should think. Right now I'm stuck with 2GB instead of 4 after two of the original corsair sticks died on me, but I'm having a bit of a hard time seeing how I can blame Nvidia for that.
Sour grapes...? ;)
No, I owned one. The nvidia chipsets were pretty notorious for data corruption issues that spanned numerous generations (NF2/NF4/NF790/etc).
I'd say that nForce4 was the worst show for NV. The network bug was really ugly and they essentially ignored it. If you installed the NVIDIA Network Manager app you would usually get TCP corruption. They never took the app out of their driver pack even though it was trouble for a lot of people.... The chipset worked fine as long as you stayed away from that firewall app.
My GF8200 mobo has been really great though (these are NF7-based). And I liked my NF2 board fine as well (clearly better than VIA stuff). I haven't used the other nForce incarnations.
nForce2's only quirk that I remember was if you installed the NV IDE driver you might run into ATAPI probs (CDROMs). But that can be said for Intel and VIA's IDE drivers too. Intel Application Accelerator anyone?
Blazkowicz
01-Feb-2011, 21:44
my nforce 5, lowest end variation is fine but the networking seems to be faulty in a way I don't wish to understand. it's okay as that board has a lot of PCI slots, I used a decade old free 3COM nic and now a PCIe 1x gigabit one from TP-link that costs under 10€.
I previously had a good experience with VIA KT266A and KT333, both with the USB2 southbridge, I didn't have any usb or sound blaster live issue and no instability whatsover but I did stupidly kill them on my own fault.
Silent_Buddha
02-Feb-2011, 03:47
Well, my respect for Newegg has gone up a notch. I just got an e-mail from them notifying me that there is an issue with the Intel 6 series chipset. And as such they are extending the return window to 90 days or until replacements are available whichever is greater. In other words no expiration on the return window until the MB can be replaced.
They'll also notify me as soon as replacement motherboards are available. Fantastic customer service. It'd be nice if when the time comes they'll offer to send out the replacement first so that affected customers won't have significant downtime. But that might be too much to ask.
Regards,
SB
Well, my respect for Newegg has gone up a notch. I just got an e-mail from them notifying me that there is an issue with the Intel 6 series chipset. And as such they are extending the return window to 90 days or until replacements are available whichever is greater. In other words no expiration on the return window until the MB can be replaced.
They'll also notify me as soon as replacement motherboards are available. Fantastic customer service. It'd be nice if when the time comes they'll offer to send out the replacement first so that affected customers won't have significant downtime. But that might be too much to ask.
Regards,
SB
I received the same email.
Unless something has changed, Newegg has always allowed cross-shipment of replacements with a credit card number. Essentially you "buy" your replacement until they get the old one back. This is what I will push for with them when the replacement boards are in.
Silent_Buddha
02-Feb-2011, 16:24
Oh that's fantastic. :) It'll be nice not to have to worry about downtime while the boards ship back and forth.
Seeing how Intel and everyone is dealing with this issue is such a nice relief compared to how Nvidia dealt with bumpgate.
Regards,
SB
Intel is a professional company. They understand honesty towards their customers is more profitable - particulary in the long term - than trying to hide your screw-ups a la bumpgate and such. Ya gotta give 'em props for that.
They hemmed and hawed a bit as I seem to recall back around the time of the original faulty 1133MHz P3 (an errata which Tom Pabst milked for years for having stumbled over by accident... :lol:), but since then they always seem to have been quite honest and up-front with their mistakes. That builds confidence in me at least, although I do wish they would have updated firmwares for their first-gen SSDs to support trim. We sure paid through the nose for those units, so it would not have been too much to ask. :razz:
digitalwanderer
02-Feb-2011, 20:30
I got some PR from Gigabyte about this, and they seem to be handling it very well siding with supporting their customers. Hope to see the trend continue. :)
Blazkowicz
03-Feb-2011, 01:46
Intel is a professional company. They understand honesty towards their customers is more profitable - particulary in the long term - than trying to hide your screw-ups a la bumpgate and such. Ya gotta give 'em props for that.
but I'd say they can afford it without thinking about twice :razz:
Intel is a professional company. They understand honesty towards their customers is more profitable - particulary in the long term - than trying to hide your screw-ups a la bumpgate and such. Ya gotta give 'em props for that.
Only after they went through the try and hide screwups phase with stuff like the Pentium Overdrive compatibility debacle and FDIV. They're really good about these kinds of things now but they weren't born this way, and of course they can afford it more easily than others.
I think just about all industries are more proactive about happenings like this these days.
I dont share your faith swaaye
weve had bumgate, around the same time there was a problem with some brand of hard drive and the company spent months in deny mode, we had apple with the iphone, first there was no problem, then it was users holding it wrong, then it was fine just the signal display was incorrect, then the sort of owned up
I heard non-K Sandys can do limited overclocking of 4 bins.....how does it work..? ...on top of Turbo Boost...a.k.a...the 4 extra bins are free ..or it works in place of Turbo Boost?
Should i change to a Sandy powered PC...with a time bomb now...and hope Intel will "compensate" early adopters...with goodies? How did Intel deal with past hardware bombs?
Silent_Buddha
03-Feb-2011, 16:36
Each bin can individually be boosted up to 4 times. So, a non-K 2500 would be limited to 3.7 (4 cores loaded) - 4.1 (1 core loaded) ghz. Almost all reports I've seen for 2500k has it able to hit 4.4 ghz (4 cores loaded) on air at stock voltage. Whether that's worth an extra 20 USD or so is up to the end user. :)
Regards,
SB
I dont share your faith swaaye
weve had bumgate, around the same time there was a problem with some brand of hard drive and the company spent months in deny mode, we had apple with the iphone, first there was no problem, then it was users holding it wrong, then it was fine just the signal display was incorrect, then the sort of owned up
That was Seagate.
nutball
03-Feb-2011, 19:08
That was Seagate.
Funny, I thought it was WD and the crowd-sourced hysteria over head parks on their Green drives (which turned out to be a heap of crap anyway, which was why the company was denying there was a problem in the first place).
Funny, I thought it was WD and the crowd-sourced hysteria over head parks on their Green drives (which turned out to be a heap of crap anyway, which was why the company was denying there was a problem in the first place).
http://forum.beyond3d.com/showthread.php?t=51644
Took me like ten minutes to track the thread down. I was looking in the wrong forums.:oops:
Silent_Buddha
03-Feb-2011, 20:20
Well, there's also been the issue with silent data loss on Samsungs recent F4 series of HDDs when S.M.A.R.T. data is requested and provided to the system during data transfers. They've already released a firmware update for it.
The WD problem was real, but only if you used it in a system with frequent read/writes combined with enough time for the heads to park. Mostly observed when used as replacements in DVR's or in some RAID setups or 24/7 media/data servers. WD now releases a 24/7 Green drive in addition to the Green drives with more aggressive head parking.
Intel's second gen SSD's also had a problem that I can't remember now (didn't result in data loss, just lower than expected performance), but they were pretty good about getting it taken care of quickly with a firmware update.
Could go on and on about mass storage device problems over the past few years. But most of them are situations where the majority of users would never run into it to trigger it. And most have been taken care of rather quickly.
Regards,
SB
It seems to me that today with the ease of end user intercommunication by the Internet that companies have learned to not screw around too much if there's a widespread flaw because their reputation will be killed very quickly by word of mouth. Lawsuits come along shorthly thereafter too. Bumpgate had its lawsuit and those affected had the opportunity to claim.
Well, there's also been the issue with silent data loss on Samsungs recent F4 series of HDDs when S.M.A.R.T. data is requested and provided to the system during data transfers. They've already released a firmware update for it.
The WD problem was real, but only if you used it in a system with frequent read/writes combined with enough time for the heads to park. Mostly observed when used as replacements in DVR's or in some RAID setups or 24/7 media/data servers. WD now releases a 24/7 Green drive in addition to the Green drives with more aggressive head parking.
Intel's second gen SSD's also had a problem that I can't remember now (didn't result in data loss, just lower than expected performance), but they were pretty good about getting it taken care of quickly with a firmware update.
Could go on and on about mass storage device problems over the past few years. But most of them are situations where the majority of users would never run into it to trigger it. And most have been taken care of rather quickly.
Regards,
SB
...Fuck I got me one of those F4 drives.
Well, there's also been the issue with silent data loss on Samsungs recent F4 series of HDDs when S.M.A.R.T. data is requested and provided to the system during data transfers. They've already released a firmware update for it.
They did a really low dirty thing when they released the firmware fix. They kept the same firmware version numbers. Now no one can tell if they have the fixed firmware version or not other than to do a firmware flash. When it's done, you still have no real confirmed way of knowing the flash was successful other than trying to duplicate the issue.
I attribute the not changing the firmware version number directly to being a dirty, evil and underhanded company for trying to sweep the issue under the rug before anyone notices. Shame on you Samsung!
Blazkowicz
03-Feb-2011, 22:12
well you might be able to dump the drive's firmware and make a diff against the newer one. but indeed that's crap!
It seems to me that today with the ease of end user intercommunication by the Internet that companies have learned to not screw around too much if there's a widespread flaw because their reputation will be killed very quickly by word of mouth.
Again apart from intel what companies are you referring too, they all try to sweep the issue under the carpet and only admit fault after a huge amount of pressure
Silent_Buddha
04-Feb-2011, 01:10
They did a really low dirty thing when they released the firmware fix. They kept the same firmware version numbers. Now no one can tell if they have the fixed firmware version or not other than to do a firmware flash. When it's done, you still have no real confirmed way of knowing the flash was successful other than trying to duplicate the issue.
I attribute the not changing the firmware version number directly to being a dirty, evil and underhanded company for trying to sweep the issue under the rug before anyone notices. Shame on you Samsung!
Yup I don't like that they did that which is the only reason I haven't bought one yet. I figure I'll let all the drives with the pre-fix firmware clear the channel before picking one up to try out as I start to plan an upgrade to my WHS v1 machine. I believe only Samsung and WD have 667 GB platters and WHS v1 doesn't like the advanced format of EARS line of WD drives and I don't want to do the whole jumpering thing.
Still really tempted by the Samsung drive right now as they are going for 80 USD on Newegg right now.
Regards,
SB
The first priority of any publicly traded company is to maximize the benefit to their shareholders. This, in the US, is almost universally summed up as "maximize share price." All actions taken are weighted first and foremost against the impact on share price. If sweeping it under a rug will serve that purpose then all companies, Intel included, will do so. To think otherwise is naive.
No, I owned one. The nvidia chipsets were pretty notorious for data corruption issues that spanned numerous generations (NF2/NF4/NF790/etc).
Not really. The only problems I remember were with their drivers. If you used the windows ones you were clear, and the scale of the problem is just hearsay to begin with. I know my Nf2 was orders of magnitude more stable than my intel chipsets I got later.
Mize!! Wisdom again. I try to tell people that over and over, but no one seems to believe it.
NF2 Ultra is still nvidia best chipset (many thanks to Xbox Project?)....Soundstorm :love:... but it did died on me...my only chipset death...lol...my PC just failed to boot all the sudden.
I dont see the point of Nvidia chipset today...when the most performance tuning they can get out is the i/o controllers?
I have a dual Opteron server at the office running an nV chipset with no issues since 2005. I also had good luck with nV chipsets in my Athlon overclocking days, though I remember you had to be careful not to install certain nV drivers...can't remember which, but I know there was common wisdom to install the provided drivers one at a time skipping one (SATA?) because it would lead to data corruption.
I have a dual Opteron server at the office running an nV chipset with no issues since 2005. I also had good luck with nV chipsets in my Athlon overclocking days, though I remember you had to be careful not to install certain nV drivers...can't remember which, but I know there was common wisdom to install the provided drivers one at a time skipping one (SATA?) because it would lead to data corruption.
The storage driver said something like "are you sure you want to install this part there is a risk of data loss" and I said yes at first without data loss, but later said no since I did not see the advantage of using their drivers anyway.
So I'm going to guess that the primary reason Intel bothered with this recall is to protect partner relationships? I'm not sure there was much risk of end users figuring it out when a small number of early boards start having SATA die after several years...
The storage driver said something like "are you sure you want to install this part there is a risk of data loss" and I said yes at first without data loss, but later said no since I did not see the advantage of using their drivers anyway.
Yup it was the storage driver. It didn't agree with some drives.
Again apart from intel what companies are you referring too, they all try to sweep the issue under the carpet and only admit fault after a huge amount of pressure
I think that might be human nature. :wink:
But really AMD and Intel seem to be on top of their serious errata. Another one that comes to mind is how Toyota seemed quick to admit that their cars had defects in some controls.
That NVIDIA debacle might be the worst cover up in recent memory. My sister went through about 4 7900M GS replacements and Dell ended up just giving her a new notebook model. I also have a friend with a 7900M GTX that died from artifacting. That seems to me like a high failure rate. No experience with GF8M failures tho.
Actually ATI's GDDR5 issues have been annoying and they never admitted anything. Occasionally their cards and notebook products using GDDR5, from 4870 to Juniper AFAIK, occasionally had memory corruption. It may be related to GDDR5 manufacturer, video BIOS, and something in the drivers but it's not clear. There are some huge threads on some forums discussing this.
So I'm going to guess that the primary reason Intel bothered with this recall is to protect partner relationships? I'm not sure there was much risk of end users figuring it out when a small number of early boards start having SATA die after several years...
I have to guess the problem is much worse than they've said. From what I've read it must be a very thin metal trace that is going to die prematurely from electromigration or some such, but the 5-15% in 3 years, as you've noted, wouldn't explain the recall.
aaronspink
04-Feb-2011, 20:54
Funny, I thought it was WD and the crowd-sourced hysteria over head parks on their Green drives (which turned out to be a heap of crap anyway, which was why the company was denying there was a problem in the first place).
Um, there IS a problem with the head parks on WD. Anyone with any brains sets the park timeout to its highest value as a first step after power on. The default value WD ships with is far too short and results in lots of bad interactions with background software that use the drives (linux, various windows utilities, etc).
aaronspink
04-Feb-2011, 20:56
Yup I don't like that they did that which is the only reason I haven't bought one yet. I figure I'll let all the drives with the pre-fix firmware clear the channel before picking one up to try out as I start to plan an upgrade to my WHS v1 machine. I believe only Samsung and WD have 667 GB platters and WHS v1 doesn't like the advanced format of EARS line of WD drives and I don't want to do the whole jumpering thing.
Still really tempted by the Samsung drive right now as they are going for 80 USD on Newegg right now.
Regards,
SB
IIRC F4s are AdvFmt drives as well.
I think that might be human nature. :wink:
But really AMD and Intel seem to be on top of their serious errata. Another one that comes to mind is how Toyota seemed quick to admit that their cars had defects in some controls.
That NVIDIA debacle might be the worst cover up in recent memory. My sister went through about 4 7900M GS replacements and Dell ended up just giving her a new notebook model. I also have a friend with a 7900M GTX that died from artifacting. That seems to me like a high failure rate. No experience with GF8M failures tho.
Actually ATI's GDDR5 issues have been annoying and they never admitted anything. Occasionally their cards and notebook products using GDDR5, from 4870 to Juniper AFAIK, occasionally had memory corruption. It may be related to GDDR5 manufacturer, video BIOS, and something in the drivers but it's not clear. There are some huge threads on some forums discussing this.
Swaaye you are completely wrong about the non desktop related issue. I have no idea where you got that idea. I don't want to derail this thread with it, but please do not perpetuate such misinformation.
Anecdottal evidence means nothing. I have never had an Nvidia product fail ever. I have had 3 AMD cards fail. I have had intel chipsets fail, a fujitsu and toshiba HDD. So what does that mean? Nothing. Other people have different luck. It is rare for us (random folks) to actually know the failure rates with any sort of accuracy. And you know there is bias towards products being broken as people who are upset their video card broke are more likely to post than those that are happy with a working one.
edit: And obviously there were issues with the nvidia bumpgate stuff I am not saying there wasn't. At the same time my laptop did not break that was one of those which may have been effected meaning HP and nvidia did not want to deal with it. I was lucky I guess.
Blazkowicz
04-Feb-2011, 21:40
the seagate 7200.11 was quite notorious, I did lose community data to it - first time I had it backed up to my personal hard drive before doing the firmware update, it worked but still crashed a while later.
next time we had a 7200.11 backup drive (!), updated and still running but a backup script woes made us lose data by overwriting the backups - I regretted not bringing my own drive again.
so here's the drive from hell - I would trust a 7200.12 though, as I know well every vendor fucks up at some point.
Swaaye you are completely wrong about the non desktop related issue. I have no idea where you got that idea. I don't want to derail this thread with it, but please do not perpetuate such misinformation.
About the GDDR5 issue? All I know for sure is that there is most definitely some sort of issue there and it happens with some notebooks and some desktop cards, and that it has been happening with various GDDR5-equipped cards since 4870. I've been following this for most of this year.
Primarily related to this ongoing thread (this problem has been around since late 2009, essentially all of the 5870M/G73's lifetime)
http://forum.notebookreview.com/asus-gaming-notebook-forum/515309-how-fix-your-gsod-blues.html
Which got me searching for more info and I found various articles and more giant threads about it.
If you mean the NVIDIA stuff, well I just assumed that the failures I know of for the GF7M chips were related to bumpgate, especially considering that the class action suit includes them. Screen corruption deaths in every case. I haven't had any desktop NVIDIA boards die. Actually none of these have been my personal hardware so I'm not on a personal revenge mission here lol.
the seagate 7200.11 was quite notorious
I steered clear of those because of how many returns were mentioned in the Newegg reviews. A friend of mine has a trio of them and one did die recently but like Sxotty says, "anecdotal!"
I've been avoiding the WD 2TB drives with the bad rap too. The 4kb sectors WD20EARS one I think it is.
About the GDDR5 issue? All I know for sure is that there is most definitely some sort of issue there and it happens with some notebooks and some desktop cards, and that it has been happening with various GDDR5-equipped cards since 4870. I've been following this for most of this year.
I meant the non-computer issue Toyota... (though it may or may not have computer related parts).
I haven't had any desktop NVIDIA boards die. Actually none of these have been my personal hardware so I'm not on a personal revenge mission here lol.
I steered clear of those because of how many returns were mentioned in the Newegg reviews. A friend of mine has a trio of them and one did die recently but like Sxotty says, "anecdotal!"
I've been avoiding the WD 2TB drives with the bad rap too. The 4kb sectors WD20EARS one I think it is.
I was only raising the anecdotal bit b/c I try not to let it affect my buying decisions generally. Obviously if I had 3 ATI (wow I wrote AMD first on accident apparently the re-branding is working) cards fail I kept buying them. The last card I did get an Nvidia one partially b/c I was tired of my bad luck.
I do wish there was some real data on reliability. Reviews seem too quick now. In the past there were quirks of motherboards pointed out, now I rarely see that in reviews. And the motherboards I have gotten don't seem enough better than the past to warrant this change.
The seagate drives were definitely bad and seagate tacitly admitted it. Nvidia did have an issue in bumpgate. Intel does have an issue now. I bet there have been other cases that never came to light though, and others where people pretended there was a problem when there wasn't really. If there was a database that tracked failures and returns and customers were privy to that info it would be great, but we don't have that kind of information at our fingertips.
I bet there have been other cases that never came to light though, and others where people pretended there was a problem when there wasn't really. If there was a database that tracked failures and returns and customers were privy to that info it would be great, but we don't have that kind of information at our fingertips.
No doubt. Well we know that most of VIA's Super 7 and Athlon chipsets were probably worthy of recall. ;) Data corruption built in southbridge feature? Check! Defective AGP over 3 generations? Check! :D
Regardless of the fact that many of the online product reviews seem to be written by folks with questionable mental acuity, they do seem to give an indicator of quality if you see a lot of returns or DOAs. Those Seagate drives have hundreds of failure reviews on Newegg and it averages to an overall mediocre rating, for example. That's about the only source I can think of besides the occasional report from a group such as Google with the HDD reliability study.
Hard drives are a particularly difficult product to judge because they have always had questionable dependability. Even if a line is inherently flawed, it's not necessarily worse than another product! You simply can not trust any hard disk to not die on you at some point.
Silent_Buddha
05-Feb-2011, 20:49
Aye HDD's are particularly hard to determine the source of problems.
For example, many issues with HDD failures is due to improper or insufficient packing combined with rough handling in transit. UPS especially have gotten really bad recently. And Newegg has really started to skimp on packing at times for their free super saver shipping.
This accounts for many of the drives that fail within a month or two of being received.
Installing 7200 RPM drives into external enclosures with insufficient cooling will also lead to early HDD death. Same applies for stacked HDD's in a computer case with insufficient airflow for high rpm drives.
I myself have 2 of the launch 1.5 TB Seagate drives still operating without a hitch in my computer. Reputation and internet hysteria can sometimes inflate the problems with one maker while at the same time making a worse problem at another manufacturer appear less problematic. BTW - this isn't to say there weren't problems with the 1.5 TB Seagates. There were, but mine for some reason didn't have the affected firmware. And no updated firmware was ever released for mine. Launch unit with slightly different product ID.
And when much of the reporting is driving by brand loyalty, brand reputation, and anecdotal evidence, it can be extremely hard to determine anything. For example, I myself have never had a Maxtor drive (had 6 or them at one time) last longer than 2-4 years before dying. Yet a friend of mine swore by them since he'd never had one die on him. Can either of that be extrapolated to indicate the manufacturer is particularly good or bad at making HDDs? Nope. But it does influence individual buying habits. I stopped buying them, while he continued. And in contrast to all that, I had 5 of the infamous Deathstar series of IBM drives. Not a single one of those died. One is actually still running in an SFF computer up at the ranch for one of my cousins. :p In fact, the only brands of HDDs where I've never had one die on me? Connors, IBM, Hitachi, and Samsung. But I wouldn't go so far as to say they are any better than WD, Seagate, Fujitsu, or any others.
About the best you can do is look for any official word on recalls or problems. And then either return the drives for warranty replacement or flast the firmware if a firmware is released that addresses the issue.
Regards,
SB
Speaking on how things can be odd indeed, I still have an original 60GB IBM Deathstar drive that is still running and perfectly functioning. The only issue it had were about 12000 hours ago, most likely due to a power spike when the PC's PSU blew up.
Like anything else, there are always exceptions to the norm, and not all models fall into the same internet hysteria.
Though on hard drives, there are two types, those that have failed and those that have not failed. All hard drives will fail. It's not a question of if, but of when.
Model Family: IBM Deskstar 40GV & 75GXP series (all other firmware)
Device Model: IBM-DTLA-307060
Serial Number: YQDYQFXD358
Firmware Version: TX8OA50C
User Capacity: 61,492,838,400 bytes
=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x000b 100 100 060 Pre-fail Always - 0
2 Throughput_Performance 0x0005 100 100 050 Pre-fail Offline - 0
3 Spin_Up_Time 0x0007 253 253 024 Pre-fail Always - 136 (Average 55)
4 Start_Stop_Count 0x0012 100 100 000 Old_age Always - 2711
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 4
7 Seek_Error_Rate 0x000b 100 100 067 Pre-fail Always - 0
8 Seek_Time_Performance 0x0005 100 100 020 Pre-fail Offline - 0
9 Power_On_Hours 0x0012 093 093 000 Old_age Always - 54761
10 Spin_Retry_Count 0x0013 100 100 060 Pre-fail Always - 0
12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 1712
192 Power-Off_Retract_Count 0x0032 098 098 050 Old_age Always - 2711
193 Load_Cycle_Count 0x0012 098 098 050 Old_age Always - 2711
194 Temperature_Celsius 0x0002 171 171 000 Old_age Always - 33
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 4
197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0008 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x000a 200 200 000 Old_age Always - 3
Speaking of the Seagate .11 debacle, I wonder how much that magnificient cockup contributed to them losing the largest manufacturer position to WD for the first time in well over a decade. Maybe two decades or even more. They've obviously been biggest for far too long though by the time the .11 drives launched.
I seem to recall that maxtors actually were worse. Google had that study that showed cold not heat killed hdd remember it?
My HDD that died are numerous, but never a WD. The 18GB drive they released right before the deathstars is also still running in a PC I gave away. That HDD is definitely on borrowed time. Sorry we are getting off topic.
Is there any repository of failure info? I do agree with Swaaye that within bounds you can get an idea by reviews. If an item has a review average of 4 with over 100 reviews and a specific model has 100 reviews and is a 2 then I would worry. It is really hard to get much more info than that though.
I seem to recall that maxtors actually were worse.
I would think that all of the big manufacturers (most which are now gone, eaten up by others) were roughly equal. Claims that one or the other was much worse than the rest have in my experience only been based on anecdotal evidence, and the specific manufacturer always varied from one person to the next . Conner, Seagate, Maxtor, Samsung, Quantum... All of those and more have I seen named as The Worst Brand Ever over the years... :lol:
Barring for certain dud models, freight damage or bad batches of drives, it is my belief it's pretty much a non-issue which brand you pick. Of course if there's statistically accurate figures available somewhere across a lot of model series and over a long period of time, then a pattern could emerge, sure.
Google had that study that showed cold not heat killed hdd remember it?
I wouldn't be surprised if cold was worse than heat (within reason of course), due to the need for good lubrication of the mechanical parts. Cold lube typically does not work very well, and metal-to-metal contact is bad for any mechanical device and particulary a high-precision piece of equipment like a HDD.
Silent_Buddha
06-Feb-2011, 06:33
Storagereview.com used to have a well populated user submitted reliability database, but it's been extremely lacking the past few years due to the site almost dying roughly 4-5 years ago. It's had a resurgence, but the drive reviews aren't as good or extensive as they used to be, IMO. And the reliability database of their's doesn't get nearly as many user entries as it did a long long time ago.
But back when it was actively used with many user submissions, it ended up with almost all drives being relatively equal in reliability, IIRC.
In their forums they also have (or had) some IT guys from Fortune 500 companies who went through 10's of thousands of HDDs a year. And for them, HDD failures were also fairly equal among the various vendors.
Regards,
SB
I thought the reason seagate was able to buy them was Maxtor was doing a bunch of silly stuff and made a few crud drives. I cannot remember specifically what the deal was since it was a long time ago. I think they were rebranding some bad drive from someone else or something like that maybe.
I had a pair of 30GB 75GXPs for about 5 years. I think one eventually died but I can't remember what I did with the other. The drives became rather whiny late in life because of the usual ball bearing wear and I really can't stand that sound anymore.
Thread thoroughly derailed but I can't think of anything more to talk about on the main subject. :D
Sinistar
07-Feb-2011, 02:02
Every Maxtor drive I have owned ended up sounding like a 1541 trying to get past the disk copy protection. Seagate drives have never failed me.
Storagereview.com used to have a well populated user submitted reliability database, but it's been extremely lacking the past few years due to the site almost dying roughly 4-5 years ago.
I think they lost a lot of credibility when they lost this database in a hard drive crash.
You'd think people running a site called Storagereview would have proper backup.
Cheers
Meh, hard drives are luck of the draw IMHO.
Between home and office I've had virtually every brand fail on me at some point and I've had "terrible" drive (Deathstars) that lasted forever.
BTW did someone already point out the irony that intel killed all their chipset competition and now has no chipsets to sell their processors? I find it humorous.
It is entertaining, yes. ;)
Actually I'm fairly amazed that this kind of flaw got past their quality control.... Heads must be rolling over there.
It is entertaining, yes. ;)
Actually I'm fairly amazed that this kind of flaw got past their quality control.... Heads must be rolling over there.
It seems to have been a late "addition" to the chipsets, so it's possible most of the QC testing was done before then.
Dr Evil
10-Feb-2011, 10:01
BTW did someone already point out the irony that intel killed all their chipset competition and now has no chipsets to sell their processors? I find it humorous.
The situation doesn't look as severe as atleast I first thought. They are now continuing to ship faulty chipsets to OEMs, who commits not to use the problematic SATA-ports in their computers, also it seems Intel is going to start shipping fixed chipsets next week.
http://newsroom.intel.com/community/intel_newsroom/blog/2011/02/07/intel-provides-update-on-support-chip-design-issue?cid=rss-258152-c1-264281
Intel is resuming shipments of the Intel® 6 Series Chipset.
Only computer makers who have committed to shipping the Intel® 6 Series Chipset in PC system configurations that are not impacted by the design issue will be receiving these shipments.
In parallel, Intel has started manufacturing on a new version of this support chip. Intel now expects to begin shipping the new parts in mid February.
I wonder how far that green light stretches. Does it allow OEM systems to ship that come equipped and only specified to supporting only 1-2 SATA drives without having to butcher the additional SATA connectors? If so, it could save those OEMs and by-proxy Intel some money from having to rework anything.
I foresee the faulty chipsets being resold into the grey market with less scrupulous vendors using them as if they were flawless. This will cause failures later on just like the bad capacitor issue. Someone remember to revisit this issue in a couple of years.
on a sort of related note if you want to run on a sandy bridge platform with a non sandybridge cpu
asrock have made a board that uses the p67 chipset but with a lga 1156 socket (instead of the normal 1155)
p67 transformer
http://www.asrock.com/mb/overview.asp?Model=P67%20Transformer
on a sort of related note if you want to run on a sandy bridge platform with a non sandybridge cpu
asrock have made a board that uses the p67 chipset but with a lga 1156 socket (instead of the normal 1155)
p67 transformer
http://www.asrock.com/mb/overview.asp?Model=P67%20Transformer
Isn't that kind of backwards from what one would want today?
Asrock should have made and 1156 chipset board that accepts 1155 cpus :)
Dr Evil
11-Feb-2011, 15:54
Isn't that kind of backwards from what one would want today?
Asrock should have made and 1156 chipset board that accepts 1155 cpus :)
Yeah and this Asrock marketing guy does not make too much sense, even if he tries hard :)
(3m07s and onward he tries to provide some reasoning)
i-XmiyI64TY
Asrock boards are always fun to read about. Nothing is normal over there.
Core 2 boards using 865PE so they have DDR1 support and AGP.
Socket 939 boards with AGP and PCIe and AM2 riser upgrade cards
It all seems unnatural but at the same time it's magical!
Silent_Buddha
11-Feb-2011, 22:16
Heh, I had one of the Socket 939 boards with both AGP and PCIE. Was great at the time as my budget was really tight. Allowed me to upgrade base components and reuse my existing high end card instead of replacing everything all at once. Thing ran on the ULi chipset before Nvidia bought them up. I was pretty impressed with the stability and quality of the ULi chipset. Shame they are gone.
Regards,
SB
http://www.dailytech.com/AMD+to+Intel+Your+Chipset+Woes+are+Helping+Us/article20870.htm
Not surprisingly, AMD is profiting greatly from Intel's gaffe according to Fox Business News. "We have some customers and retailers who have come to us specifically as a result of Intel's chip problem," stated AMD exec Leslie Sobon. "Some retailers have had to take things off their shelves, so they call us to ask what they could get from our OEMs that's similar. And OEMs are asking us for product, as well."
Way to rub it in AMD.... anyone remember the Barcelona [edit: was it Agena and not Barcelone that had the bug?] TLB bug?
Until Bulldozer ships and beats Sandy Bridge AMD is just talking trash for no good reason IMHO.
3dilettante
14-Feb-2011, 16:09
[URL]Way to rub it in AMD.... anyone remember the Barcelona [edit: was it Agena and not Barcelone that had the bug?] TLB bug?
Agena was the desktop SKU of Barcelona.
Both were affected by the bug until the B3 stepping.
Until Bulldozer ships and beats Sandy Bridge AMD is just talking trash for no good reason IMHO.
Agreed 100%. I don't think that was a good move for AMD.
And what exactly did AMD gain? Probably less than 1% in marketshare whereas if AMD slip up with Bulldozer they have a hell of a lot more to lose.
Blazkowicz
14-Feb-2011, 17:59
what about they release their AM3+ platform at least, so we don't have to buy an outdated mobo.
I'm not a fan of the disparition of competition for AMD chipsets either (can't tell a difference between a 770 and a 870 chipset, lack of really cheap full ATX boards)
at least there's no discrimination on advanced features (both 64bit and VT on ontario/zacate, IOMMU on all 9xx chipsets)
Intel doesn't want you to build a server with dual core Atom, 64bit, VT, 4-8 GB ram, it wants you to buy a core i3 and overpriced motherboard instead.
http://techreport.com/discussions.x/20420
More poking fun... don't want to hot link the pic up there but it's definitely worth 30 seconds of your time.
AMD, you better deliver on Fusion (which I thought was dead as marketing term, make your minds up AMD.)
Somebody elsewhere brought up a good point.
SB isn't exactly a significant amount of Intel product volume so it's probably not even remotely as damaging as it seems. It's a PR problem though.
Even without SB they still have offerings superior to everything AMD sells....
Albuquerque
15-Feb-2011, 22:01
I guess that's really the sad thing though, isn't it? The upper echelon of Intel's 'old' non-SB equipment is still capable of curb-stomping anything that AMD sells. I mean, talking trash is great, but it's more like Toyota talking trash to Ferrari. "Yeah your F430 catches in fire hahahahaha -- people are totally coming in droves to buy Camrys now, and you should TOTALLY see our new Sienna SUV!"
Wait, what?
Talk shit when you have something to back it up, AMD :)
Except Toyota owns Ferrari :)
If AMD sold that much more than intel they would be having caviar parties and dancing through the night not freaking out.
Silent_Buddha
16-Feb-2011, 20:30
Somebody elsewhere brought up a good point.
SB isn't exactly a significant amount of Intel product volume so it's probably not even remotely as damaging as it seems. It's a PR problem though.
Even without SB they still have offerings superior to everything AMD sells....
I should hope not, if there were a lot of ME running around, this would be a really odd world. ;)
Except Toyota owns Ferrari :)
Ermm, you mean Fiat, right? :)
Regards,
SB
So back on the topic sort of, will this effect the lga2011 chipsets? I did not notice that. I wonder b/c I want a hex core not a quad core if I go intel. Are those coming in August or what?
Silent_Buddha
17-Feb-2011, 23:24
I'd be surprised if Intel repeats this particular mistake with the LGA2011 chipset, especially as that's meant primarily for the business world with some ethusiast bleed off.
Regards,
SB
What I meant was: Is there carry over that would have effected the lga2011 if they had not realized this? And if so will it delay the lga2011 release? I suppose it could delay the release anyway though as I imagine they might be a bit overly careful after this disaster. (Want to spend longer testing and so forth)
Silent_Buddha
20-Feb-2011, 01:30
That's hard to say. Would more stringent testing from an enterprise class chipset have revealed the flaw? Will the enterprise class chipset re-use (without changes) the same implementation? The possibility is always there, but there's no way to say with any amount of certainty that it would happen.
Regards,
SB
Just an update:
I got my Asus Sabertooth P67 Rev. 3.0 motherboard replacement ("Advanced RMA" so I get the new one then send back the old one) from Newegg today.
Sigh. Motherboard swaps are a bitch.
At least it arrived before you went flying off to China... That would surely have been an even bigger bother, right? :razz:
At least it arrived before you went flying off to China... That would surely have been an even bigger bother, right? :razz:
:)
Now I have to decide whether to do the swap before or after though..."oh bother"
Silent_Buddha
10-Mar-2011, 17:53
Bah, I'm on the "waiting list" at Newegg. They are swapping out MB's as they arrive and in order of when the MB was ordered.
Regards,
SB
vBulletin® v3.8.6, Copyright ©2000-2013, Jelsoft Enterprises Ltd.