|
|
Nvidia chipsets are defective too
BumpGate Only 630, 7050, 7100 and 7150 changed for no reason
By Charlie Demerjian: Monday, 22 September 2008, 4:50 PM
LOOKS LIKE THOSE wacky folk at Nvidia are at it again, changing perfectly good chipsets for no reason. If you recall, they changed the G86 parts and the G92/G92b as well, for no reason.
Luckily, the firm seems to be spending money like there is no tomorrow, it has nothing to do with the fact that they have a few warehouses full of defective chips, they seem to want to keep their validation engineers fully employed this time of year for no apparent reason. This time, the chipsets being changed are being changed 'just because', and the bump material has nothing to do with crapping out. Really. This, this and this are vague guesses.
So, with that, we bring you word of the next PCN, this time for chipsets, or 'motherboard GPUs' if you are terminally stupid enough to believe the marketing hype.
Because this one is no different from the other two PCNs linked above, we will skip the detailed description and just bring you the pertinent details. Should you want the formatting, it is still here.
This little ditty is called "MCP73 Products Change Bump Material from High Pb Solder to Eutectic Solder"", and is confidential, NDA'd, and has a June 2008 date. The PCN # for the pedantic is PCN0353A. Nothing new there.
Page 2 has a more exact date of June 13, 2008, with a PCN submit date of June 13, 2008 as well. Planned Implementation Date is July 25, 2008, and Proposed First Ship Date for change is August 17, 2008. This means that the parts should be hitting the shelves about now, so if you want to buy a low performance MCP, you can buy in moderate safety now. It is still a Class 1 change.
The list of the affected parts is: NF-630I-A2, NF-7050-610I-A2, NF-7050-620I-A2, NF-7050-630I-A2, NF-7100-630I-A2 and NF-7150-630I-A2. It looks like this is an admission that there is absolutely nothing wrong with the 630, 7050, 7100 and 7150, they just are changing them just because. No need to panic, really.
Page 3 has the same diagram as the G92 PCN, and the same reasoning, to enhance package robustness. We guess they can't use the term failure, defect, or crib death in a PCN. Odd really. In any case, the qual data is available now, with samples going out on July 18, 2008. Eutectic bump products will be marked with a trailing R on the lot #, but no box marking for underfill this time.
The document ends up on a humorous note. There are two different dates, a generic June, and June 13, both 2008, earlier in the PCN. On the page 4, they have 7/11/08 listed for Revision A, the initial release. Come on guys, a little proofreading here and there wouldn't kill you. It would kill us, but you guys have the pretence of professionalism. µ |
|
IP卡
狗仔卡
发表于 2008-9-23 00:27
QQ好友和群
收藏
分享
好贴
烂贴
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
显身卡
楼主