I’ll give Google a sliver of credit though with Open Titan being a much better security processor idea than Microsoft’s
I don’t want a freakin’ “security processor”.
3 Likes
You don’t want Chip-to- 
security?
That makes a Micro$oft very sad!
It uses 22 communication protocol.
Good.
1 Like
I think a security processor is a great idea…with certain caveats
The design and implementation needs to be open sourced so as to allow for faster response to security issues and more eyes to go “This aint right” and improve it over time.
Black box proprietary “security” has consistently proven to be a lie as “security through obscurity” has been shown to be a fallacy on many occasions and tends to be more harmful than helpful.
2 Likes
In cases of Intel ME, AMD PSP and Pluton it’s not even a lie, but it’s purpose is to be a backdoor, not a security - so it works perfectly good for it’s actual purpose. 
While thats entirely possible for that to be the intent, I have yet to see anyone prove this claim in practice. I have no trust for these hardware vendors, but Id also like this claim to actually be proven true or false by someone. This is consistently parroted around and i feel mildly suspicious of them myself but to state it as fact till proven is a bit silly to me.
So…First it was TPM 2.0 and now that’s not good enough? Reading this makes me even happier that I just upgraded my motherboard and processor so I won’t have to worry about this crap for a while longer.
If AMD wants to add security to it’s processors i have no issue with that. I just don’t want Microsoft or anyone else involved in it.
1 Like
agreed, i would prefer them implement something like the open titan security processor using a custom RISC-V chip or something
1 Like
I would rather see it be an open source linux … thing.
are you implying a security processor using a linux kernel?
Doesnt seem like a good plan to me, while linux has plenty of security potential i dont feel its up to the task of being the linchpin of platform security and a ground up open sourced design would be far superior to anything that could be hacked together with linux.
Let’s just say so far 100% of similar nature things security / politically / medically related i had audacity to mention on this forum, which gets called out for not having any proof yet - had received very definitive proof some time later, so it’s likely a matter of time.
For Intel ME i believe proof enough would be some of those NSA whistleblowers reporting that all of their motherboards ordered aboard doesn’t have Intel ME right out from the factories, unlike same boards going to general public…I can think of only one idea why. 
But as you’ve correctly mentioned, even in complete absence of clear-cut proof yet - since none of those systems are open-source - we don’t know sh*t about what they do, and that’s enough proof and measurement of how dangerous it is, given it’s hardware which is much harder to get rid of / control rather than software…
2 Likes
No don’t read anything into it. I’m just saying something developed by AMD themselves or by help with opensource linux community in mind. Who exactly develops it is out of my wheel house.
oh ok i got you
Thats what i was suggesting was a community/group effort not something AMD does on their own
that would not go well imo lol
Sadly that we are not even close to the stage to make DIY hardware any compatible.
I saw only this dude that was able to make equivalent of the intel’s 4004, but it’s a level of the 70’s.
Perhaps ~20nm’s could be affordable to manufacture for a small-mid range business in the next 5/10 years. Using multi-chip approach, 3d transistors and other modern architecture optimisation combined with RISC-V as main architecture and SoC approach with many highly-specialised modules it’s may still somehow compete with a bleeding edge technology.
Even now tiny CPLD/FPGA startups are capable of showing that they’re even capable outperfom big tech chips, just by having more superior and specialised design. Though most of them are still not independent from TSMC or any other big chip maker.
2 Likes
I don’t have a problem with AMD helping to develop it i mean it is their processor designs. So they would have to be involved IMHO.
Being the Agnostic type Im not the type to take a lack of proof/knowledge of something as proof in itself. I wont deny the potential, but i wont claim its fact either till i have more evidence to support said claim.
1 Like
AMD was dreaming about DRM SoC integrated into their hardware and with constant access to the internet for a long time. I remember news from the 2015 that’ve been saying that they already had such stuff in development and ready to use.
Not to say that they same as Intel, Nvidia, ARM and others already had such SoCs with access to the internet, but which were serving mainly for the process managment instead of specifically DRM.
I’d say so far RISC-V looks very promising…Also you can buy this server motherboard approved / created by FSF which works with some very specific CPUs (older than Pluton of course), it costs a fortune 