Intel Chipset Design Error!!!

Discussion in 'Motherboards' started by bezant, Jan 31, 2011.

  1. alxnet2003

    alxnet2003 Limp Gawd

    Messages:
    271
    Joined:
    Oct 19, 2009
    I'm sure all the motherboard manufacturers will create new revision numbers for their specific models (e.g. going from rev 1.0 to rev 2.0) when these new SB boards are released. Perhaps Intel can keep a list of all affected motherboard model and revision numbers on their website. Seems like this would be easier since the rev numbers are all plainly printed on the motherboards.
     
  2. Taze

    Taze n00b

    Messages:
    49
    Joined:
    Nov 2, 2010
    I went into Fry's where I got both CPU and Mobo to ask if they would have a program in place for the recall. They said you can either return both CPU and Mobo and exchange for a different setup or get a refund, or even wait utill the new boards come out and exchange at that time. That's at least at the one near me. Didn't know if this was put out there or not. They also had pulled all P67 boards from their shelves.
     
  3. lightp2

    lightp2 Gawd

    Messages:
    672
    Joined:
    Oct 28, 2009
    Based on latest news, looks like shipment is resuming for certain vendors. Since no concrete info is available at this point, potential for ever expanding current chipset-based products to flow into market is increasingly.

    If possible, we suggest a software approach for detection

    1. In Linux, programmers have been known to be able to detect special issues with various SATA implementations. In general this ability applies for all FOSS. I believe Microsoft is able to do so as well. Many of them scan for detail chipset registers/board information to deduce hardware devices they are dealing with.

    2. Either it is detected at the Linux kernel-level and alerts the users that one of the affected-3G ports is in use. OR

    3. A Windows or Linux low-level utility that scans the system configuration and alerts you about potential issues.

    In this way, users do not need to waste time trying to understand where could be potential issues. All they have to do is run this utility or Linux kernel detection and the SATA status will become immediately clear without any complicated research or guessing.
     
    Last edited: Feb 7, 2011
  4. Iching

    Iching [H]ard|Gawd

    Messages:
    1,805
    Joined:
    Nov 29, 2008
    Same goes for MC. Do you think we return the bundle after 30 have passed? I had to rma almost a brand new i5-760 since ram would not work in dual channel with it. I might end up returning SB because of the flaw and keep i5-760.
     
  5. lightp2

    lightp2 Gawd

    Messages:
    672
    Joined:
    Oct 28, 2009
    One of the frequent counter-comments regarding the SB-chipset issue is that notebooks are less affected as they can use 6Gb ports.

    Type A. majority notebooks have maximum 2 SATA devices = one disk, one DVD drive.
    Type B. 3 SATA devices = one disk, one DVD drive, one eSATA.

    Type B is more complicated. Some more advanced latest notebooks have USB/eSATA combination port.

    However, even for Type A the situation is not so clear. WHY?

    1. Notebook is closed system. Some are designed especially difficult to open unless done by servicemen.

    2. General users are not technical expert. They hear this logical explanation using 2x6Gb ports, but they have no way to actually physically see clearly whether this is true or otherwise. Sometimes it is easier to open desktop casing to verify the SATA cable connection. Likely not so easy/convenient with laptops.

    3. If this is so easy, almost all the major notebook vendors will come out and say their laptops models are using only 6Gb ports, simple declaration, problem solved. However, so far we only hear they planning to address the problem.

    4. And yes, a lot of things are soldered on or pre-integrated on laptop motherboards. Even if you are able to open laptop casing, you may not see SATA cables at all. They could have directly integrated internal connectors waiting for SATA drive. How are you going to tell this is not affected ports? Anyhow I don't think normal users able to go deep on this. We hear suggestion chipset has 2x6Gb ports. Conventional wisdom says they ought to use them. Conventional wisdom is not enough when dealing with closed systems with a lot of unknown. Without technical reference, are users able to look at the model/even the laptop boards and say with 100% certainty these SATA ports come out from 6Gb? that some laptop makers may instead choose to route to 3Gb ports thinking maximum compatibilities with existing software/hardware? or maybe some use 3Gb and 6Gb ports connection to slightly differentiate their product performance segments?

    5. And users still have to inquire on those 3SATA devices type of laptops.

    A lot of clarification pending from the vendors.
     
    Last edited: Feb 8, 2011
  6. maverick786us

    maverick786us [H]ard|Gawd

    Messages:
    1,923
    Joined:
    Aug 24, 2006
    If I am not wrong, SATA 6GB is the latest trend and has backward compatibility with Sata 3GB and earlier.

    So as an alternative why don't they simply replace rest of the 4 ports with SATA 6GB? That will be an alternative solution
     
  7. alxnet2003

    alxnet2003 Limp Gawd

    Messages:
    271
    Joined:
    Oct 19, 2009
    It's because the Intel SB chipset only supports 2 SATA 3 ports. In order to replace the four SATA 2 ports with SATA 3, they'd need to add another chip to the motherboard (as some other boards have done by adding Marvell's SATA controller). The manufacturers would probably rather wait for the fixed chipset than deal with the cost and complexity of adding more chips to the board.
     
  8. lightp2

    lightp2 Gawd

    Messages:
    672
    Joined:
    Oct 28, 2009
    I really don't want to keep pushing on the same points. After I read further comments on some websites, I want to relay my experience again here.

    1. There are all kinds of counter-reasons and suggestions, which my previous posts already touch on so I am not repeating here.

    2. However, I am going to tell you actual things I have gone through, thus forming the basis on my current views on this event.

    3a. Many years ago, at previous workplace standardize on one particular brand of Intel-based desktops.

    3b. A lot of desktops purchased. At one particular year, we discovered desktops were randomly hanging left and right a few weeks after initial rollout. We thought it was malware/virus or program errors and continue to re-image the desktops again and again.

    3c. One day I came across something written on the Internet about capacitor problem. Apparently it had been going on for sometimes. I used the info to inspect those problem machines. We called support and vendor would not admit and just say continue to try and test other options. Finally the principal was called in and noted the issue. Yet only commit to replace those machines with actual severe failing problems, even then they do not have enough replacement parts and need to wait. What triggered me writing about this was that the vendor was probably aware of the issue (already written on the web) but we were continuously asked to try different diagnostic steps and not being informed about the actual issue.

    3d. And then you have bumpgate, which I try hard not to write again due to potentially bunch of fans complaints for raising old issue. Apparently many vendors were fully aware back then. Again, just like my previous experience, initially we were told all kinds of reasons, diagnostic steps and counter-measures. Only when the issue became too severe did everybody finally own up about the actual issue.

    3e. Thus in a way I am glad Intel declare the issue up-front as soon as possible. But now we are in the second part, the execution part. Intel signals it intends to recall affected parts.

    after going through such issue twice, my response is now in suspense because I am not really sure whether I will go through another round of diagnostic steps, counter-measures. and "helpful suggestions" to do many things when the involved firms are aware of the actual issue itself.

    The test is very simple. You ask all the supposed expert opinions to substantiate their opinion with their own money buying the affected chipset-motherboards. If the problem is only SATA, good and everybody is happy. And if H6x series notebooks are all SATA-6Gb connected, even better, just announce publicly and be done with it. My apology here due to past experience I am perhaps overly sensitive to nVidia-style early-response-so-many-reasons-but-finally-admit-something-is-definitely-wrong-with-bump-n-overheating.

    What I am afraid to see is if those affected laptops are pushed to unsuspecting users that will only show symptoms several years later when warranty is already out.

    Here I commend Intel for doing the right thing by declaring up-front about the issue and putting out recall. However, upon hearing further news about how various pros/vendors are trying to think about alternative schemes how to re-use affected parts, all my memory is coming back...

    Edit : I really am ok if this is 100% guranteed to be only SATA-3Gb issue and with discount. Due to past experience, I certainly hope I am not getting another surprise later. SB looks really attractive and I am considering a small system build.
     
    Last edited: Feb 8, 2011
  9. Impulse

    Impulse [H]ardForum Junkie

    Messages:
    10,232
    Joined:
    Apr 5, 2001
    Doesn't CPU-Z identify the chipset stepping for your mobo? I'm sure if CPU-Z doesn't do it something else should be able to... Intel already announced they're changing the stepping (B3 or something, I forget), so as far as enthusiasts are concerned I don't think this will be a big issue moving forward. Most mobo manufacturers will probably change revisions too (which they usually silk-screen unto the mobos). The only way you'll run into issues is if you happen to buy backstock a few months from now from a less-than-reputable seller who also doesn't have a decent return policy.

    Now the average consumer's another story, but by and large Sandy Bridge has yet to arrive into the hands of the average consumer... And most of them buy laptops that will never use more than two SATA ports, so it's probably far less of an issue for them in the long run. I know the potential is always there for manufacturers or retailers to screw this and give Intel a prolongued black eye over the whole ordeal, but it seems pretty unlikely. I still wouldn't buy a laptop until the revised chipsets are out tho, unless I'm damn sure the manufacturer did their due diligence and is only using the SATA6 ports.
     
    Last edited: Feb 8, 2011
  10. Impulse

    Impulse [H]ardForum Junkie

    Messages:
    10,232
    Joined:
    Apr 5, 2001
    Unless someone finds something proving different I think it's pretty safe to say the issue was strictly related to the SATA3 ports, Intel spent a while trying to pin the issue down before going public and they wouldn't have done so if they weren't sure... Further, I think the fact that they resumed shipments for laptops that don't use said ports would put many conspiracy theorists and the paranoid sort at ease. I don't think they'd have done that if they weren't doubly sure that was the only outstanding issue w/the chipsets.

    If there were other problems or there are problems w/the SATA6 ports I'm sure we'll find out fairly soon, given the number of enthusiasts now using those ports exclusively for their OS and data drives, myself included.... System isn't built yet but it should be by this weekend, and both my SSD and my 640GB drive are going on the unaffected ports. I'll probably be shopping for a USB 3.0 enclosure since I won't be able to use my existing e-SATA bracket + enclosure for my external backup drive, that's the only inconvenience for me (and I was likely moving it to USB 3.0 anyway since e-SATA's probably gonna fade away).

    Stuff like this happens people, early adopter has come to be synonymous with guinea pig, we should know this by now. The important fact isn't whether it happens at all, it's how companies respond to these sorta issues... I think Intel has a good track record in that department. For instance, when a buggy firmware update bricked some of their SSDs last year they quickly pulled it and made no bones about replacing the bricked drives.
     
    Last edited: Feb 8, 2011
  11. Astrodave

    Astrodave [H]ard|Gawd

    Messages:
    1,062
    Joined:
    Apr 20, 2005
    Maybe asus and some of the others can use this time to figure out wtf is wrong with a lot of their boards from a memory and cold boot standpoint in the next few months. FFS, just reading these forums alone for 10 minutes was proof enough to me that SB had massive problems before the recall was even a known entity. About 50% of people using a SB system were complaining about major problems from what I could surmise. I never recall EVER seeing that with any of Intel's other platform releases. SB was simply a victim of itself and was released too early before the bugs were worked out. Asus boards are a veritable nightmare on forums for issues and the MSI board that I had my eye on got so many complaints on newegg that it sent me running for the hills. I base all of my upgrades on forum user feedback and it was quite evident with SB that it was not a stable platform. Now I know a lot of you guys will fly to the rescue for SB but I am simply telling you what my own research indicated. Word of mouth is the most powerful indicator that something is rotten in Denmark...
     
  12. Synomenon

    Synomenon [H]ardness Supreme

    Messages:
    7,589
    Joined:
    Feb 10, 2007
    Had problems with my Gigabyte GA-H67MA-UD2H sometimes freezing during post then going into a reboot loop.

    The latest BIOS update fixed that and haven't had any problems since. I think most of the problems right now are from user error or immature BIOS and / or drivers.
     
  13. brom42

    brom42 2[H]4U

    Messages:
    3,980
    Joined:
    Mar 1, 2004
    I agree. I have the GA-H67MA-UD3H board in my main box and had a couple of issues until I updated the BIOS. It's had zero problems and been rock solid since.
     
  14. YeuEmMaiMai

    YeuEmMaiMai [H]ardForum Junkie

    Messages:
    14,211
    Joined:
    Jun 11, 2004
    According to the latest news, Intel will have shipments out sooner than expected. See it didn't take as long as all the naysayers said it would.....
     
  15. Disposed

    Disposed [H]ardness Supreme

    Messages:
    5,163
    Joined:
    Mar 2, 2010
    What news is that. Last thing I read was that intel was waiting on silicon to even sttart making replacements.

    Got a link?
     
  16. limitedaccess

    limitedaccess [H]ardness Supreme

    Messages:
    7,479
    Joined:
    May 10, 2010
    Don't be too excited, the date of initial shipment was revised from 3 weeks (on jan 31) to Mid February. Which really means the same thing, but spins well for your stock holders.
     
  17. YeuEmMaiMai

    YeuEmMaiMai [H]ardForum Junkie

    Messages:
    14,211
    Joined:
    Jun 11, 2004
    techreport stated on Monday that Series 6 chhipsets resumed shipment and that Intel "now expects to begin shipping the new parts in mid February." this is revised from end of Feb.......
     
  18. Disposed

    Disposed [H]ardness Supreme

    Messages:
    5,163
    Joined:
    Mar 2, 2010
    Considering the difference between mid feb and end of feb is a few freaking days i would have to agree that its pretty wording to keep share holders happy but means absolutely nothing in the real world.
     
  19. YeuEmMaiMai

    YeuEmMaiMai [H]ardForum Junkie

    Messages:
    14,211
    Joined:
    Jun 11, 2004
    wait what? the difference between feb 14 and feb 28 is 11 working days if you use it as mid feb if you even go with 18th that is still 7 working days and unlike most of the western world asia pretty much works 6-7 days a week so it is definately more that just a "few" days
     
  20. Disposed

    Disposed [H]ardness Supreme

    Messages:
    5,163
    Joined:
    Mar 2, 2010
    When someone says "mid month" that can pretty much mean anything. Its rather foolish to try and interpret wording like that into exact dates.

    They can say mid month and ship feb. 24 and not be wrong.

    So again its wordplay for share holders.

    When they stop using vague wording and nail down a date you can argue but until then it can really mean anything and you reading between the lines and interpreting it how you want it to be looks pretty silly.
     
  21. YeuEmMaiMai

    YeuEmMaiMai [H]ardForum Junkie

    Messages:
    14,211
    Joined:
    Jun 11, 2004
    ^ same could be said about someone who is trying to turn the situation into doom and gloom when in reality Intel is on track to fix this sooner than they expected.....

    btw, no one in their right mind would consider the 24th of Feb mid month as that clearly falls into LATE feb........nice try though......
     
  22. Disposed

    Disposed [H]ardness Supreme

    Messages:
    5,163
    Joined:
    Mar 2, 2010
    I am trying to do nothing of the sort, just because you dont want to hear it dont make it so. Im actually quite happy with my SB setup.

    Welcome to the business world where vague statements are stretched all the time. Do you live under a rock or something? This is common practice.
     
  23. Ryokurin

    Ryokurin [H]ardForum Junkie

    Messages:
    10,560
    Joined:
    Aug 14, 2001
    It was speculation due to the news happening during Lunar new year, and the migration that happens during that time and can last for up to a month. It's pretty unlikely that electronics plants will be idle that long however. Either way, they likely won't have 8 million boards ready by the end of the month, so the fact that they are starting to fab replacements does not really say much.
     
  24. Forceman

    Forceman [H]ardForum Junkie

    Messages:
    9,243
    Joined:
    Aug 7, 2005
    It is still going to be weeks until that turns into shipping motherboards though.
     
  25. jdub

    jdub n00b

    Messages:
    2
    Joined:
    Feb 8, 2011
    Don't use or ever plan on using the effected sata ports. Running two Crucial SSD's in raid O on the Intel SATA6 ports, with the marvels as my backups as needed (Asus Sabertooth P67).

    700meg read is working out well for me.

    Still, I plan on exchaning for a new one when the time comes. A non-issue for this SB early adopter...
     
  26. Skillzero

    Skillzero [H]Lite

    Messages:
    123
    Joined:
    Jun 28, 2010
    Was considering building an affordable desktop with sandybridge but its kind of hard when intel has broken it, LOL.
     
  27. Synomenon

    Synomenon [H]ardness Supreme

    Messages:
    7,589
    Joined:
    Feb 10, 2007
    Funny, mine is working very well...
     
  28. dioxholster

    dioxholster Gawd

    Messages:
    801
    Joined:
    Aug 11, 2010
    Had Intel had a competitor this wouldve been a disaster.
     
  29. Forceman

    Forceman [H]ardForum Junkie

    Messages:
    9,243
    Joined:
    Aug 7, 2005
    Classic quote. Nice chuckle from that one.
     
  30. bezant

    bezant Gawd

    Messages:
    755
    Joined:
    Oct 7, 2009
    Exactly what I've been thinking...
     
  31. InfernalMachine

    InfernalMachine Limp Gawd

    Messages:
    418
    Joined:
    Jan 8, 2004
    I'm sure their competitors are still very pleased while they kick themselves :)
     
  32. Skillzero

    Skillzero [H]Lite

    Messages:
    123
    Joined:
    Jun 28, 2010
    Intel kind if disappointed me.

    Was 2 seconds from building a desktop.. Decided portability would be better and went to go buy a laptop. Now I can't find anything "up to date" as all the SB laptops have been pulled.

    More waiting, I guess :(
     
  33. Synomenon

    Synomenon [H]ardness Supreme

    Messages:
    7,589
    Joined:
    Feb 10, 2007
  34. Spittle

    Spittle n00b

    Messages:
    62
    Joined:
    Oct 13, 2010
    Shipping to OEM partners on February 14. Most likely won't start shipping to board manufacturers till a later date.
     
  35. Night_Hawk-19

    Night_Hawk-19 Gawd

    Messages:
    739
    Joined:
    Jun 20, 2004
    least 4 months. Mind you I have X58 chipset and probabily wait for later chipsets.
     
  36. Cyclone

    Cyclone Limp Gawd

    Messages:
    503
    Joined:
    May 2, 2001
    I wonder how long until they ship to themselves. I need the Intel mATX P67 board 2 weeks ago. :(

    I keep placing orders for it at places that show and/or tell me on the phone they guarantee they have it in stock and the orders keep getting cancelled. My credit card has 6 separate charge/refunds from this over the past few days.
     
  37. Disposed

    Disposed [H]ardness Supreme

    Messages:
    5,163
    Joined:
    Mar 2, 2010
    Unless you can get lucky and find one somewhere you wont be getting your hands on one for at least a couple months.
     
  38. lightp2

    lightp2 Gawd

    Messages:
    672
    Joined:
    Oct 28, 2009
    I read the official announcement on vendors' websites. Suggestion points to April/May almost universally for actual physical motherboards with fixed P67/H67 chipsets per general availability.
     
  39. sybreeder

    sybreeder Limp Gawd

    Messages:
    193
    Joined:
    Oct 24, 2010
    The funniest thing about it is that Intel decided to sell damaged chipsets anyway. For example gigabyte renew selling boards with that issue again so will be again avaible in stores and in OEM PC.
     
  40. limitedaccess

    limitedaccess [H]ardness Supreme

    Messages:
    7,479
    Joined:
    May 10, 2010
    Intel's announced they will let partners implement the chip and sell them in systems that will not be impacted by premature SATA 3gb/s failures.

    For instance many laptops have only 2 physical ports present that will connect to 2 devices.