Network News

X My Profile
View More Activity
Posted at 11:32 AM ET, 01/31/2011

Intel finds chipset problem, says it will cost $700 million to fix

By Hayley Tsukayama

Intel issued a statement that it has found a critical flaw in the design of its Sandy Bridge chipset and has implemented a silicon fix.

The company said it will cost $300 million in revenue for the first quarter of 2011 to stop production of its current chips and manufacture the new ones. Overall, the company estimates it will cost a total of $700 million "to repair and replace affected materials and systems in the market."

Part of the chipset is prone to "degrade over time," which can affect the performance of hard-disk drives and DVD drives. The faulty systems started shipping on Jan. 9 and will affect customers who bought Second Generation Core i5 and i7 systems.

Intel said it expects to have updated chipsets out to customers in late February and be back at full volume by the end of April.

Intel shares were halted Monday morning, and the stock dipped 1.5 percent when trading resumed. Rival company Advanced Micro Devices' (AMD) stock rose about 5.1 percent.

By Hayley Tsukayama  | January 31, 2011; 11:32 AM ET
Categories:  Computers, Gadgets, Tablets  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   Del.icio.us   StumbleUpon   Technorati   Google Buzz   Previous: White iPhone has flash problems, says Steve Wozniak
Next: Facebook Deals makes its European debut

Comments

Actual Chipset is called: Couger Point & Problem is Sata2 Sub System WEAROUT. Reccomendation is to change all Sata2 over to Sata3 Subsystem, for now & await return instructions..

However, As Sata2 is Well Know Performer, Problem might be fundelmently Sandy Bridge cpgpu. First of Kind. Perhaps Sata2 just is too pokey for Hogh Speed Gpu storming in background. So far, Faster Sata3 holds, While Sata2 Slowly Disintergrates, from High Speed Wear.

Billion dollar loss & ?Intel didn't test Sandy Bridge in such simply way as compatibility with Couger Point Chipset. Bah HumBug. Once again Public Been Taken for Fool.

Signed:PHYSICIAN THOMAS STEWART von DRASHEK M.D.

Posted by: ThomasStewart1 | January 31, 2011 5:27 PM | Report abuse

Actual Chipset is called: Couger Point & Problem is Sata2 Sub System WEAROUT. Reccomendation is to change all Sata2 over to Sata3 Subsystem, for now & await return instructions..

However, As Sata2 is Well Know Performer, Problem might be fundelmently Sandy Bridge cpgpu. First of Kind. Perhaps Sata2 just is too pokey for Hogh Speed Gpu storming in background. So far, Faster Sata3 holds, While Sata2 Slowly Disintergrates, from High Speed Wear.

Billion dollar loss & ?Intel didn't test Sandy Bridge in such simply way as compatibility with Couger Point Chipset. Bah HumBug. Once again Public Been Taken for Fool.

Signed:PHYSICIAN THOMAS STEWART von DRASHEK M.D.

Posted by: ThomasStewart1 | January 31, 2011 5:28 PM | Report abuse

Actual Chipset is called: Couger Point & Problem is Sata2 Sub System WEAROUT. Reccomendation is to change all Sata2 over to Sata3 Subsystem, for now & await return instructions..

However, As Sata2 is Well Know Performer, Problem might be fundelmently Sandy Bridge cpgpu. First of Kind. Perhaps Sata2 just is too pokey for High Speed internal to cpu, Gpu storming in background. So far, Faster Sata3 holds, While Sata2 Slowly Disintergrates, from High Speed Wear.

Billion dollar loss & ?Intel didn't test Sandy Bridge in such simply way as compatibility with Couger Point Chipset. Bah HumBug. Once again Public Been Taken for Fool.

Signed:PHYSICIAN THOMAS STEWART von DRASHEK M.D.

Posted by: ThomasStewart1 | January 31, 2011 5:30 PM | Report abuse

sorry about excess posts, WP Server went Beserk & Stated: NOT Due to Word Remove & posted again.

Shows new software has glitch in comments here.

Signed:TS, If possible to remove first two, OK.

Posted by: ThomasStewart1 | January 31, 2011 5:35 PM | Report abuse

Wow, it never even occurred to them to properly test the device before shipping.

Intel ain't what it used to be!

Posted by: thomasmc1957 | February 1, 2011 12:34 AM | Report abuse

Failing to anticipate the necessary test is quite different from failing to test. When you're the market leader, you're inevitably going to encounter issues that no one has seen before.

Posted by: slar | February 1, 2011 9:05 AM | Report abuse

SATA 6Gpbs ports are not affected because they are controlled by separate parts of the PCH.

ThomasStewart1, they know the problem, it's the chipset, not the CPU. High speed wear has nothing to do with it. Per anandtech ( http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug ), "Intel biased the transistor with too high of a voltage, resulting in higher than expected leakage current." So it's a design flaw in the PCH, nothing more. The only 'wear' is from normal use, no matter what the speed because of the degradation of the transistor over time from being used at all due to the high voltage.

Nor does it have anything at all to do with SATA II being "too pokey for High Speed internal to cpu, Gpu storming in background." I'm sorry, but that doesn't even make sense. Using a slower SATA interface might bottleneck performance (though at full speed even that is doubtful with SATA II), but this has nothing to do with operating the ports at capacity. See above for the real reason.

Posted by: hokiealumnus | February 1, 2011 9:38 AM | Report abuse

Dear Alumnis: Heres another article, by Charlie. Charlie, Mike & Rob used to write together:

http://www.semiaccurate.com/2011/01/31/intels-recall-explanation-doesnt-hold-any-water/

Notice Intels Controller Explanation dosn't Hold Any Water, Is Key Senario. Charlie suggests another BumpGate, with BAD Soldering on Some. My Explanation is Units Go Out of Parameter with use on some specific assortments of Hardware. Within PCH, problem seems to be I/O HDD, yet cpgpu also has iMX & DDR3 Memory controller, Slow Sata 1/2 might just wind up parsed thru wrong gates, being slow & problem occurs with 2 or more Sata 1/2 HDD, NOT Just One. So Merely Controller, well why dosn't problem always occur, especially as just One HDD seems Clean & Any Combo HDD into Sata3 Slots is Clean of Any HDD Mix. Sata3 being Faster & able to handle entire mess into Faster cpgpu onboard controllers.
Could be way too late to rapidly change Sandy Bridge & Cougar Point Takes blame or Fixit Spot. Remember, iMX can go ~8 Giga transfers sec, while Sata often rolls on own at ~20 MB/s. Timing Mechanism just gets signal mixed up in Parse Enviorment.

Somewhere in PCH System in controller &/or cpgpu, theres lack of coordination & Machine Emits STOP Signal.

Signed: vondrashek md.

Posted by: ThomasStewart1 | February 1, 2011 11:48 AM | Report abuse

Is the same chip that the Government wanted to put to tracer back door it so it could 'monitor' any PC and the input-output?

Posted by: KBlit | February 1, 2011 11:53 AM | Report abuse

Sorry, Crowd Is Correct, Issue is ports 2-5 Sata2. Ports 0-1 are O.k. on Couger Point Sata2. Leakage & Heat when OverBurdened. NO STOP Signal, Just Degradation of Signal.

Next: Options:

If customers would like to keep their systems they can get a free SATA 3G discrete hard drive controller card added into their system to get around the chipset issue. The issue resides specifically with the hard drive controller in the chips for controllers ports 2-5, the SATA II 3Gbps ports. It does NOT affect the SATA III 6Gbps ports, or ports 0 and 1.

So Wait to buy, Return or Get Card that plugs into iMX system thru pci-e16X subsystem. Looked up Mike Musgrove & moved to Radio.Pop was Buried last month.

Well, if any cpgpu issue,Remains Undisclosed. However, Sata ports configure themselves, so NOT actual placement or specific physicial port location to plug into?, its after you use port 0,1 & Sata assigns numbers 2-5 to next Sata plugs,Mainboard does list Sata Ports by Specific number, However. While higher speed slot added on can make all work Properly. Might be cpgpu & lower speed & number of lower speed HDD in Sata2 subsystem, as more than 2 HDD or Location ofSpecific Port Input. Might Be My Birthday.Might be Robs' 20 Years Older Than Looks & looks 20 Years older than WP Pic & Rough as Heck, at CES this year.

World+Deceptions.

TS.

Posted by: ThomasStewart1 | February 1, 2011 12:57 PM | Report abuse

Thank you for the clarification TS. That semiaccurate piece (a lot of which I find to be just that) seems to have been written prior to the anand piece I linked earlier.

The discrete controller option is a good one, but shouldn't be required. With any luck, OEMs will make the replacements as painless as possible. Time will tell.

Posted by: hokiealumnus | February 1, 2011 4:13 PM | Report abuse

Found Annaund Shrimpi article Informative, Annaund is Belgium.HiCookie on Fb is another Belgium Worth Reading.

Seems Musing above are bit late. However, Mfgs often downplay problems into simplier terms. turns out ports 2-5 are entire Sata 1/2 subsystem, so NO sata 1/2 HDD will ever work Perfectly & STOP IS final outcome.

When Intel describes Problem as One Transistor Gate, More likely One Entire subsystem. Good Point is refurbished Sandy Bridge Units Should Abound this Spring. As Good as New. Often 1/2 Price.

Sata1/2 Voltage problem extends into SSD useage, as SSD Runs at 1.8 Volts & Sata 1/2 system runs 3.3 Volts to 5 Volts. Big Problem year ago & covered Up. SSD that Rated to near 100 MB/s Where Running at 50 KB/S Due to Stuttering from Voltage OverLoad. So BeWare, SSD Still has problems going into Sata 1/2, Yet Flys Like Jet into Pci-e system.Same With Cougar Point, in Entire Sata 1/2 Useage. BTW Those Pci-e Boards Can be Expensive, as they Work. User needs Just Plug in pci-e board for Sata1/2, much,much less.

Overal Believe Best Is To Chop Out old Subsystems that are slow, ide, Sata1/2 being part & Floppy. End of Problem, New Dawn Be Breaking in High Speed Computing.

For Now, Wait for Refurbs'. Annuand mentions Sandy Bridge Stepping B.0 is Problem,now, So B.1 be number to look for as Well, or Higher B.2, B....Until Entire System Works Properly. Seldom C Stepping, as far as know, maybe when completed.

Thanks for putting Up With Rave, Heres' Tip O' Day:
Regularly CLEAN Your Computer Insides., Dust Collects on cpu Fan & Causes Overheating, Which Stops Media Stream & crashes Computer. Be Gentle, Vacumn cleaner makes Electrons in chips go flakely, so use (Camel?) Hair brush on Vacumn. Then look into cpu fan, often dust collects on heatsink like grey paper, scrape that completely off with small tool. Dust Removal should be Several times year. If can see dust coming out Power supply Fan, Get Into case & clean entire mechanism several times Over, Throughtly, is Good Scanario.

vondrashek md

Posted by: ThomasStewart1 | February 1, 2011 8:37 PM | Report abuse

Hi Rob & Friends. Two Hours Ago Intel Announced that NO Corrections, Nor Refunds will be issued for Cougar Point Chipset, Here:

An Intel spokesperson, meanwhile, said that the company did not have plans to formally recall affected products, although its customers, PC OEMs, were free to act as they wished.

Intel has only Once recalled product, in 1993, due to Arthemetic Unit problem, which makes display, gibberish.

More to UpDate from Intel, tomorrow, 2Feb'11.

Named Intel Spokespersons' nam e is Curious, salvator. Sounds Pavlovian.

Monyanah, My Friends. Oh, to further explan subsystem involved, ide first came with CD rom & HDD, in early 1990s', However, after millions of machines where assembled & Sold, Intel found that ide/cd rom had NO cd rom player on/off function, so software for cd rom player 1.01 was released, made quite headline splash then. In same year,Vechile: dodge caravan made Air Conditioning Unit with NO On/Off Switch that worked, took 18 Months to Correct. So Impossible Is Possible, maybe software can correct Cougar Point, yet sounds GRIM. Since merely 1 Transistor gate Is Problem & that is Counter for bits coming thru, Been Suggested, that Specific Transistor Can be cut off, simple thing, by targeting defective metal of one transistor with overload, till breaks. Lets wait & see. Maybe forget Sandy Bridge/Cougar Point till Summer. 8 Million Intel Machines are Affected by Above Engineering Glitch.

:) TS

Posted by: ThomasStewart1 | February 1, 2011 10:42 PM | Report abuse

Post a Comment

We encourage users to analyze, comment on and even challenge washingtonpost.com's articles, blogs, reviews and multimedia features.

User reviews and comments that include profanity or personal attacks or other inappropriate comments or material will be removed from the site. Additionally, entries that are unsigned or contain "signatures" by someone other than the actual author will be removed. Finally, we will take steps to block users who violate any of our posting standards, terms of use or privacy policies or any other policies governing this site. Please review the full rules governing commentaries and discussions.




characters remaining

 
 
RSS Feed
Subscribe to The Post

© 2011 The Washington Post Company