Hi All, Alan Cox wrote: > Most of the 'classic' K6 chips people have tested have a second bug where > a very specific bizarre execution sequence can lock the chip up until you > powercycle. Its not realistically going to happen by accident so its only > an issue for a multiuser machine with shell access and users who read > bugtraq ;) > > > to order a few more machines and I am wondering if they should be > > AMD K6-2 or Celeron-A (almost same prices for same MHz). Some K6-2's can run with a 100 MHz front-side bus and relatively large secondary cache which can prove to enhance performance for some compute bound applications. (They are a favorite of many "gamers" and OPENGL fans for their floating point 3D instruction enhancements). > The K6-2 doesnt have the problem Alan, some of the early K6-2 processors ARE affected by this bug. Please review the attached statement provided to me by AMD. Thanks, Lyle -- Lyle P. Bickley | Bickley Consulting West Inc. [EMAIL PROTECTED] | 1697 Grant Road V 650-428-0621 | Mountain View, CA 94040 F 650-428-0599 |
In essence all model 6 and model 7 processors are affected by this erratum. Only a subset of the model 8 processors are affected (AMD-K6-2). However, there is no way, through software or visual inspection, to determine if you have a stepping of the model 8 that resolves the erratum. All shipments of the model 9 (AMD-K6-3), which will be launched next year, will not be affected.
