I've hesitated to join this discussion, as interesting as I've been finding it, but would like to respond to one particular point raised...
Ombra wrote:Don't expect too much from AMD processors. I've been running across some distasteful info on them, that suggests they have at least begun to join the DRM fanclub. Looks like AMD-V and maybe VT-x or VT-d is in the PhenomII x6. Have to research it more.
They're virtualisation extensions, and particularly useful in the server space now that a lot of companies seem to have jumped on the 'virtual server' bandwagon; I'm afraid there it's a case of "if one competitor has it, the other needs it or loses by default", and whether or not they can be used for the purposes of Rights/Restrictions Management is somewhat moot.
If you want a CPU without AMD-V, you're going to have to go back a long way - "Pacifica" tech started to be integated into CPUs with the Athlon 64. Although amusingly the socket 939 CPUs didn't support it IIRC.
VT-x is the Intel equivalent, which began to appear in the Pentium 4. Some mobos will allow dis/en-abling it in the BIOS, so you at least have the option of disabling it in hardware. VT-d is the latest version (Nahalem and newer CPUs, not all of them have it, will need to check in Intel Ark for exactly which ones) the AMD version is AMD-Vi... both allow more direct hardware I/O from a virtual machine.
While I'm sure that improving virtualisation could be used for Rights/Restrictions Management (at least as far as improving the security of the guest OS from the host OS and vice versa) none of those techs appear to be aimed solely at implementing DRM in the manner you appear to be concerned about. The things i'd be worried about are NX bit and Trusted eXecution Tech along with their AMD equivalents - while TXT gets paired with VT-d, that pairing appears to be using VT-d to assist in isolation of memory/hardware allocations - essentially it appears that TXT uses VT-d to interact like a virtual machine with hardware when dealing with 'DRM enabled' software requiring it.
However, that use of VT-d by TXT does not appear to make VT-d a 'nasty' in its own right... but it
can be when called upon by TXT.
I'm by no means an expert on this stuff though.
