How to identify Athlon CPU? A7N8X-E Dlx having trouble with 3200+ CPU. CPU fraud?

Discussion in 'Asus' started by Erik Harris, Aug 14, 2004.

  1. Erik Harris

    Erik Harris Guest

    I bought an XP 3200+, and the sticker on the chip looks right, gives me all
    the right data, but any sticker can be stuck onto any CPU. I'm having trouble
    getting this thing running at the proper speed in my new A7N8X-E Deluxe
    motherboard, and I'm not sure which is bad, the CPU or the MoBo (but I suspect
    the CPU). I bought the OEM chip from what I've always considered to be a
    reputable seller (mWave.com, I've purchased a few things from them over the
    last 3 years or so, so they're not a fly-by-night place), but I'm suspecting
    that they might have sold me a rebadged chip (their supplier might be at
    fault, or they might be less reputable than I thought).

    Quite often, the Asus Voice POST feature tells me "system failed due to CPU
    overclocking" and won't boot up, especially if I try running at 200x11. I can
    pretty consistently run at 166x11, and the chip is recognized as a 2500+.

    Given that stickers can be easily defrauded, how do I know what CPU I've got?
    I've tried a number of CPUID programs (both of the programs from AMD's website
    plus cpu-z, from cpuid.com). They all seem to "tell me what I want to hear."
    If I'm running it as an Athlon 2500+, it tells me that it's an Athlon 2500+.
    If I run it as an Athlon 1900+ (133x11), it tells me that it's an Athlon
    1900+. Even the EAX value changes depending on what speed I'm running at.

    Also, on one of my reboots, I tried setting the multiplier to something other
    than 11. It seemed to work, but I need to investigate further to see if it
    actually runs at a different multiplier. Aren't all Athlon XP 3200+ chips
    locked at 11? The lock doesn't change if I lower the bus speed, does it?
    i.e. if I'm running at 166MHz instead of 200MHz, should I be able to increase
    the multiplier? I'd have assumed that the answer was "no."

    To sum up - is there any way to know whether or not I'm a victim of CPU fraud?
    I've still got time before mWave's 30-day warranty on the chip runs out, and
    if this chip does turn out to be bad, I want to know ASAP. :-(

    Thank you very much in advance for any help you can offer.
     
    Erik Harris, Aug 14, 2004
    #1
    1. Advertisements

  2. Erik Harris

    Ed Guest

    If you have a unlocked CPU you should be able to raise and lower the
    multiplier in the bios no matter what the FSB is set too, but there are
    limits as how far you can go up and down, tho anything from 9x to 12.5x
    shouldn't be a problem. If it's a locked CPU it should just stay at 11x
    no matter what you set it to.

    Yes the XP3200+ uses 11x (2200MHz - 200x11.0)

    Do you have PC3200 ram and have it set to 100% so it runs in sync of the
    FSB? Not sure if that has anything to do with your problem but these
    boards generally don't like running in async mode.

    Ed
     
    Ed, Aug 14, 2004
    #2
    1. Advertisements

  3. Erik Harris

    Almeyda Guest

    sounds that theres a chance that u've been taken for a ride... well either
    that or the Ram is overcloking because its rated at 333mhz(166fsb) and ur
    running at 400mhz(200fsb) and thats perhaps why ur getting the warning.

    im running a 2500+ @ 3200 speed on same mobo and i dont get any warning but
    then again my ram supports the speed im currently overclockin.

    older 2500 bartons were able to be unlocked by mobos and ran fine
    overclocked to 11x200, its possible that they sold you a 2500 knowing that
    it quite simply can be clocked to 3200 (without changing mutliplier).

    but that seems unlikely... though u need to find out. also the way to check
    is on the chip itself it will give you the details to assure you if its 3200
    or not.

    --
    Almeyda

    AMD2500XP @3200XP-11X200-2.21Ghz
    A7N8X-E Deluxe > Corsair TWINX1024 XMS3200LL DualChannel
    HIS Excalibur 9800XT 256mb @463/399 (Omegas 2.5.51 & ATI tool 0.20 No
    Artifacts)
    1x80GB SATA Seagate 8mg cache > 2x40GB ATA Seagate
    SB AudigyES > ANTEC SuperLanboy 350 Case (350W SmartBlue PS)
     
    Almeyda, Aug 14, 2004
    #3
  4. Erik Harris

    Ron Todd Guest


    Well..... Semi deja vu....

    I also bought a -E with a OEM 3200+ form MWave, with the represented
    to run swell Kingston DDR400 Value Mem. AIR, best I could get the
    thing to run was 185 x 11, front & back. Paid the bucks and bought
    the Corsair matched pair. The best I can mine to run reliably at is
    195 x 11. Got to love them Mwave 3200+ OEMs. I gave up on the
    correspondence with the Mwave folks around May as they kept insisting
    it was impossible not to get 200x11 with the Kingston and the AMD
    3200+ OEM cpu and it was all my fault......

    Best of luck.
     
    Ron Todd, Aug 15, 2004
    #4
  5. Erik Harris

    Paul Guest

    An interesting problem. The chips all start out the same, until the
    bridges program the "grade" of the chip. Fraud can be perpetrated by
    meddling with the bridges. In the case of changing a 2500+ to a
    3200+, one is 166x11 and the other is 200x11. The only difference
    might be the FSB_sense[1:0] pins. See PDF page 87 of this document:

    http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/26237.PDF

    "FSB_Sense[1:0]

    Pins FSB_Sense[1:0] pins are either open circuit (logic level of 1)
    or are pulled to ground (logic level of 0) on the processor package
    with a 1k ohm resistor. In conjunction with a circuit on the
    motherboard, these pins may be used to automatically detect the
    front-side bus (FSB) setting of this processor. Proper detection of
    the FSB setting requires the implementation of a pull-up resistor on
    the motherboard...

    Table 26. Front-Side Bus Sense Truth Table
    FSB_Sense[1] FSB_Sense[0] Bus Frequency
    1 0 RESERVED
    1 1 133 MHz
    0 1 166 MHz
    0 0 200 MHz"

    FSB_sense[1] is pin AG31
    FSB_Sense[0] is pin AH30
    VSS (GND) is pin AH32

    Using an ohmmeter, flip the chip over and measure from a sense
    pin to the nearest VSS (GND) pin. If both sense pins read 1K ohms,
    you haven't been defrauded. If FSB_Sense[0] is open circuit, you
    have a 2500+ with a new label stuck on it.

    That is the best test I can come up with.

    To help you visually, go here:

    http://www.ocinside.de/go_e.html?/html/workshop/pinmod/amd_pinmod.html

    Select "PCB View" "AMD Barton" and set the FSB to 200MHz. PCB
    view is like viewing the bottom of the motherboard, and it is the
    same view you see when looking at the bottom of the CPU chip.
    The pattern you see on the screen would correspond to what those
    two 1K resistors are attempting to do - bridge AG31 to AH32 and
    bridge AH30 to AH32. Use that pattern to help guide where to measure
    with the ohmmeter.

    Use antistatic precautions when working on the CPU.

    HTH,
    Paul
     
    Paul, Aug 15, 2004
    #5
  6. Erik Harris

    Paul Guest

    It would be nice, instead of going to this much trouble,
    if you could rely on the BIOS to read the FSB_sense pins.
    But I'm not sure the "Auto" setting in the BIOS always
    does the right thing. If might read the 00 value from
    the pins, conclude it doesn't know what that means, and
    then force the pins to 01 instead. If the BIOS played
    fair, and didn't meddle with the pin value, then you could
    rely on the BIOS Auto setting, coming up at 1833MHz, as
    proof of a fraud.

    The fraud I postulate above, is simple label tampering. A
    more complicated fraud, including the label change, would be
    modifying one bridge, to change a 2500+ 01 code to a 00 code.
    A visual check of the surface of the processor might uncover
    that kind of fraud, but the people who do this kind of stuff
    are probably pretty well set up to do it, as with the amount
    of money they would be making, they could afford some pretty
    fancy equipment.

    Are there any scratch marks near the label ?

    Paul
     
    Paul, Aug 15, 2004
    #6
  7. Erik Harris

    Erik Harris Guest

    I've actually got PC4000 RAM, and yes, it's running in sync.

    Turns out that I was defrauded, either by mWave.com, or by mWave.com's
    supplier. I got help from the nice folks on the AMD support forum. All the
    details are here: <http://forums.amd.com/index.php?showtopic=21632>

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 15, 2004
    #7
  8. Erik Harris

    Erik Harris Guest

    The chip itself doesn't give those details. A _sticker_ on the chip gives
    you those details, and stickers are very easily faked, as mine was. The
    sticker told me that the chip was a 3200+. BIOS recognized it as a 3200+,
    and automatically tried to run it at 200x11. But it was a 2500+, as someone
    on the AMD Support Forums (<http://forums.amd.com/index.php?showtopic=21632>)
    determined by seeing a closed L12 bridge in a photo I posted of my chip. For
    what it's worth, I bought this fraudulent chip from MultiWave (mwave.com).

    If not for that person's help, I'd have kept going to try to find the
    problem, and eventually narrowed it down to a "faulty" CPU, when it was just
    a CPU that couldn't be overclocked to 3200+, even though most of the 2500+
    chips could be. I suspect that trying to run it that fast may have damaged
    the chip, because I can't get it to be reliable under any speed. Even at
    133x11, I need to reset my computer 2-4 times to get it to boot up. Once it
    boots up, it seems stable, but getting it running is a challenge.

    I'll have a replacement chip, from a different vendor (NewEgg), on Tuesday,
    or Wednesday if I'm less lucky (paid for overnight shipping, but NewEgg
    doesn't guarantee that they'll actually send it out promptly, unless I pay
    them an extra $3, in which case they'll "do their best" to ship it out
    promptly). As soon as I have the new chip, this garbage 2500+ is going back
    to mWave, totally at their expense (that took a lot of effort, too).

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 15, 2004
    #8
  9. Erik Harris

    Erik Harris Guest

    Interesting... Thanks for the note; it's "nice" to see that someone else was
    defrauded by mWave. It lends credence to the idea that they're not terribly
    honest.

    Do you still have the CPU? Do you have pictures of it, or can you look at
    it? Take a look at the thread at
    <http://forums.amd.com/index.php?showtopic=21632>, where someone kindly
    pointed out that my OEM 3200+ is a rebadged 2500+, as evidenced by the fact
    that the first L12 bridge is cut. Is this the case on your CPU?

    Simply, I want to know if I should report mWave to the Better Business
    Bureau. If multiple people are being screwed by them when buying CPU's, it
    suggests that I should.

    For what it's worth, I got them to refund every cent and pay to ship the CPU
    back to them, but it took a great deal of effort and about a 45-minute phone
    call.

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 15, 2004
    #9
  10. Erik Harris

    Paul Guest

    Maybe better to report them to AMD. See if AMD wants the processor as
    evidence. I'm sure the AMD lawyers have nothing better to do than
    chase weasels like that. They won't be able to dodge AMD with the
    "my supplier did it" excuse. If they want product, they can buy it
    straight from AMD, not from some guy in an alley.

    Paul
     
    Paul, Aug 15, 2004
    #10
  11. Erik Harris

    Erik Harris Guest

    My BIOS recognized it as 200x11 by default (but wouldn't actually boot
    successfully).
    I've got pictures up at http://www.eharrishome.com/CPUPic1.jpg and
    http://www.eharrishome.com/CPUPic2.jpg

    Someone on the AMD forums (http://forums.amd.com/index.php?showtopic=21632)
    saw these pictures and pointed out that the first L12 bridge is cut, as a
    2500+ was. Dunno about the other bridge sets and how they match up with a
    2500+ or 3200+, but that alone tells me that this CPU was not manufactured as
    a 3200+. However, not only was the sticker replaced with a 3200+ sticker,
    but the chip was somehow modified (perhaps in the way you detailed in your
    message) to make the BIOS automatically attempt to boot up at 200x11 when it
    identifies the chip on its own.

    Sneaky. MultiWave (or their supplier) got caught in their fraud when they
    let this chip go out modified and apparently didn't test it to find out that
    it won't actually run at 200x11... Actually, I can't get it to boot reliably
    at ANY speed. Just so I have a working PC until I get my replacement CPU
    (from NewEgg, hopefully on Tuesday), I've got it clocked down to 133x11, and
    even at this speed, I need to hit the reset button 2-4 times before my system
    will POST.

    MultiWave insists that they're a victim in this, too, but given that someone
    else here has also gotten an OEM 3200+ CPU from mWave that won't boot at that
    speed, I've gone ahead and written up a review of the company for BizRate,
    FatWallet's reseller ratings, ResellerRatings.com, and Pricegrabber.com..
    Are there any other vendor rating sites out thre that get lots of usage? I'm
    not aware of any others, but if any are anywhere near as popular as the four
    I've already posted my rating on, I'd like to "share the wealth" and expose
    this issue.

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 16, 2004
    #11
  12. Erik Harris

    Ron Todd Guest

    I did not say I was defrauded by Mwave. I only say it didn't work the
    way I understood it was supposed to, I did not find their responses
    satisfactory, and there position is it was my fault. I still disagree
    with this but I am not making and issue of it.
    Yes, I still have it, and it is still running at 195. I pulled the
    heat sink and the sticker looked proper.
    I don't know, I would have to pull the thing apart to check. I don't
    have the time to do it at the present. Mine is running cool at the
    moment, room 82F cpu 125F per ASUS probe. That doesn't sound much
    like an over clocked 2500 does it?

    I am curious, what did Mwave say when you confirmed the bridge cut
    identified it as a 2500+ instead of a 3200+?
     
    Ron Todd, Aug 16, 2004
    #12
  13. Erik Harris

    Ron Todd Guest

    FWIW, mine auto reads to 200x11 and 200fsb, but with the Kingston
    Value RAM, it wouldn't boot. I could stable operation at about
    180x11. Switching to the Corsair matched pair, it would boot and run
    for about a hour (average) before doing various things such has
    spontaneous reboots. With the Corsair it has run nicely at 195 since
    June.
     
    Ron Todd, Aug 16, 2004
    #13
  14. Erik Harris

    Erik Harris Guest

    I tried. I couldn't find anything on AMD's website for reporting fraud. I
    sent an email to their tech support folks - the web-email interface bombed
    out on me with an error, but I got a confirmation message via email, so I
    assume they got my message. They haven't responded.

    If you know who to contact at AMD regarding fraud, let me know. My ISP's
    news server is having trouble with delays (my messages don't appear until
    many hours, or even a day after I post them), so if you do know who to
    contact, please let me know via email ASAP, not just via the group. I say
    ASAP only because I plan on sending this chip back to MultiWave on Wednesday,
    since they won't refund my money until they get the chip. If going through
    AMD is a viable option, though, I'll send it to them instead of mWave and
    just contest mWave's charge on my credit card.

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 16, 2004
    #14
  15. Erik Harris

    Erik Harris Guest

    Sorry, you're right, you didn't say you were defrauded by mWave.. I said it
    based on the fact that your symptoms pretty much match mine, and match that
    of a chip overclocked beyond its ability. I should have "someone else may
    have been defrauded by mWave." Either way, the words are mine, not yours. I
    didn't mean to put words in your mouth.
    My sticker looks proper, too. The CPU's sticker was replaced with one that
    is "correct." The CPU was also modified to make my BIOS "recognize" it as a
    200x11 chip when it's not.
    Actually, it does. The majority of unlocked 2500+ chips are more than
    capable of running at 200x11. It turns out that the one they tried to pawn
    off on me isn't, though. Keep in mind that all Barton chips are the same
    when they're first made. Whether it's a 2500+ or a 3200+, it started off as
    exactly the same thing and produced exactly the same way. The difference?
    The 2500+ chip was tested successfully at a lower speed. That does NOT mean
    that it was unsuccessfully tested at a higher speed (though it may have
    been), because they need to produce X number of chips at each speed. If they
    already have their quota of 3200+ chips, the rest are going to be sold at
    lower speeds - even if they can run just as well at the higher speeds.

    You should only need to pull the heat sink off to look at the bridges, just
    like you did to look at the sticker, unless your chip has a sticker covering
    the bridges, as some of them do (mine had no sticker at all there, so the
    bridges were clearly visible).

    Another thing you can do to test this out is change the speed in BIOS - not
    just the FSB, but the multiplier. A legit 3200+ processor will NOT let you
    change the multiplier. Mine did, because it's a 2500+ that was made before
    AMD started locking their chips (week 39 or 2003, or something like that).
    Having a locked multiplier doesn't guarantee anything either, though. The
    2500+ and 3200+ both have a standard multiplier of 11, making it possible to
    make a 2500+ masquerade as a 3200+ even when locked (since the bus speed
    isn't locked, just the multiplier).
    They claimed to be a victim of this mistake a well. The manager I spoke to
    insisted that this was a production mistake on AMD's part. When I pressed
    him on the possibility of an unscrupulous supplier, he said that MultiWave
    gets all of their AMD processors "from official AMD channels." I did note
    that he didn't say that they get them directly from AMD. He didn't have
    anything in particular to say about the fact that the L12 bridge told me
    conclusively that this was produced as a 2500+ chip.

    The CSR I spoke to didn't care WHAT I said. I insisted on getting my money
    back, and not paying them a penny. She insisted that I had to pay the
    original $8 shipping. I told her I'd take my case to the BBB. I even said
    "look, if you give me too much of a hard time, I can just contest this charge
    with my credit card company. That way, you won't get a cent OR your CPU
    back." Even then, she said she couldn't refund my shipping. Her manager
    could, though, and he did, with the caveat that he wouldn't refund my money
    under the assumption that mWave was at fault. I told him that I accepted the
    possibility that they may have received a modified CPU from their supplier.
    He interrupted me saying "not mat have, did!" before I could point out that
    it doesn't matter who was at fault, because as the vendor, they are
    responsible for what they sell (at least insofar as fraud is concerned - I
    won't blame the vendor for a DOA product; that's different).

    To the CSR's credit, she was firm, but she remained polite and professional
    the whole time. I assume that she didn't have the authorization to drop the
    shipping charge, but it's also possible that she was just afraid that if she
    did, she'd get in trouble, even if she could do it.

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 16, 2004
    #15
  16. Erik Harris

    Paul Guest

    Well, I searched around, and I guess this kind of thing _never_
    happens :-( I guess AMD would like to keep their head in the
    sand. All I could find is the contact page - maybe you'll get
    some advice there.

    http://www.amd.com/us-en/Corporate/AboutAMD/0,,51_52_3592_712,00.html

    I could have sworn I read an article about remarked Athlons, but
    tonight a few quick searches aren't turning up anything of use.

    Paul
     
    Paul, Aug 17, 2004
    #16
  17. Erik Harris

    Ron Todd Guest

    Thank you for the information and clarification. I will have to check
    my cpu and find out if I have something other than what I paid for.
    The thought of writing all those letters is really unpleasant.

    Best of luck and i pray you reach a satisfactory resolution, Ron
     
    Ron Todd, Aug 17, 2004
    #17
  18. Erik Harris

    Kylesb Guest

    | In article <>, Erik Harris
    |
    | > On Sun, 15 Aug 2004 15:21:28 -0400, (Paul)
    wrote:
    | >
    | > >Maybe better to report them to AMD. See if AMD wants the
    processor as
    | > >evidence. I'm sure the AMD lawyers have nothing better to do than
    | > >chase weasels like that. They won't be able to dodge AMD with the
    | > >"my supplier did it" excuse. If they want product, they can buy
    it
    | > >straight from AMD, not from some guy in an alley.
    | >
    | > I tried. I couldn't find anything on AMD's website for reporting
    fraud. I
    | > sent an email to their tech support folks - the web-email
    interface bombed
    | > out on me with an error, but I got a confirmation message via
    email, so I
    | > assume they got my message. They haven't responded.
    | >
    | > If you know who to contact at AMD regarding fraud, let me know.
    My ISP's
    | > news server is having trouble with delays (my messages don't
    appear until
    | > many hours, or even a day after I post them), so if you do know
    who to
    | > contact, please let me know via email ASAP, not just via the
    group. I say
    | > ASAP only because I plan on sending this chip back to MultiWave on
    Wednesday,
    | > since they won't refund my money until they get the chip. If
    going through
    | > AMD is a viable option, though, I'll send it to them instead of
    mWave and
    | > just contest mWave's charge on my credit card.
    |
    | Well, I searched around, and I guess this kind of thing _never_
    | happens :-( I guess AMD would like to keep their head in the
    | sand. All I could find is the contact page - maybe you'll get
    | some advice there.
    |
    |
    http://www.amd.com/us-en/Corporate/AboutAMD/0,,51_52_3592_712,00.html
    |
    | I could have sworn I read an article about remarked Athlons, but
    | tonight a few quick searches aren't turning up anything of use.
    |
    | Paul


    I've also seen that article Paul. A lil searching turned up this
    article, which is the one I recall reading:

    http://www.xbitlabs.com/news/cpu/display/20030503024652.html
     
    Kylesb, Aug 17, 2004
    #18
  19. Erik Harris

    Erik Harris Guest

    I ended up calling their tech support, and giving him all of the info,
    including the thread on AMD's forums, my invoice number, etc. I found
    nothing about contacting their legal department, but he said that he'd
    forward the info to them.

    Fortunately, AMD didn't want my CPU. If they took it, I wouldn't be able to
    get my money back from MultiWave. Also, if they took it, they'd be legally
    obligated to give me something in return - probably a 2500+, he said, which
    wouldn't help me much.

    In MultiWave's defense, when he spoke to his manager to ask if they wanted
    the CPU or just the info, he came back saying that his manager was very
    surprised that my remarked CPU came from MultiWave, because he (like I)
    understood them to be very reputable, and indicated that it's possible that
    one of their official distributors is remarking OEM CPU's. Either way, AMD
    is going to look into it and determine who is at fault, and who their legal
    department should go after.

    In contrast to the above, MultiWave's price was pretty much too good to be
    true. If they were selling modified chips thinking that they were legit, why
    were they undercutting every other reputable dealer by $40? (That's not an
    insignificant amount on a $195 chip!) That seemed odd when I bought it, but
    I went ahead and did so anyhow, because I believed MultiWave to be reputable,
    based on their ratings on the review sites, and based on my own previous
    experiences with them.

    --
    Erik Harris n$wsr$ader@$harrishom$.com
    AIM: KngFuJoe http://www.eharrishome.com
    Chinese-Indonesian MA Club http://cimac.eharrishome.com

    The above email address is obfuscated to try to prevent SPAM.
    Replace each dollar sign with an "e" for the correct address.
     
    Erik Harris, Aug 18, 2004
    #19
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.