DigiTimes: AMD market share may hit record low in Q1-16

[The Stilt]

Most of the consumer board bioses are built with so outdated AGESA that the microcode isn't even affected. For example the newest bios for the newest ASUS boards (970 GAMING/Aura) is based on V1.5.0.2 AGESA and uses 06000822 version microcode. ASUS isn't even remotely the most sloppy company when it comes to updating their bioses. The newest bioses for ASRock newest AM3+ boards are built with V1.2.9.0 AGESA, which dates back to 2H/2012

 

[Phynaz]

AMD Servers: New error allows user to take over the entire server, to be fixed with microcode update



www.computerbase.de/2016-03/amd-opteron-microcode-patch-fuer-fehler-in-piledriver-cpus

Details from the kernel mailing list:

https://lkml.org/lkml/2016/2/26/876

It's pretty significant when you can force a CPU to execute data in userspace that has the NX bit set.

 

[caswow]

Wrong topic

 

[William Gaatjes]

Details from the kernel mailing list:

https://lkml.org/lkml/2016/2/26/876

It's pretty significant when you can force a CPU to execute data in userspace that has the NX bit set.

Here is an English version :

http://www.theregister.co.uk/2016/03/06/amd_microcode_6000836_fix/

Question :
I have an A10-6700. I have a piledriver core too. Is this bug happening with server cpu's only or do richland cpu's also suffer from this bug ?

 

[The Stilt]

Here is an English version :

http://www.theregister.co.uk/2016/03/06/amd_microcode_6000836_fix/

Question :
I have an A10-6700. I have a piledriver core too. Is this bug happening with server cpu's only or do richland cpu's also suffer from this bug ?

Despite both Richland and Orochi Rev. C being based on the same microarchitecture (Piledriver), the cores are not exactly the same. Piledriver APU cores should not be affected by errata #815, of which fix caused this issue.

 

[William Gaatjes]

Despite both Richland and Orochi Rev. C being based on the same microarchitecture (Piledriver), the cores are not exactly the same. Piledriver APU cores should not be affected by errata #815, of which fix caused this issue.

I see. Thank you.