That's been common practice since roughly forever. The only difference here is that you get the option to "uncripple" it, rather than having to buy a new, uncrippled version if you want to upgrade.
BTW: that was on Slashdot earlier today. Two stories later, we get:
Science: Deleting Certain Gene Makes Mice Smarter
An anonymous reader writes "Deleting a certain gene in mice can make them smarter by unlocking a mysterious region of the brain considered to be relatively inflexible, scientists at Emory University School of Medicine have found.
I guess the next step is God-vs-Intel patent litigation.
And adds a fifth leg, a bald spot on their head, spider pinchers and venom, and only one eye. But they are smarter. ;)
The problem with deleting a gene is that these usually code for a protein. And specific proteins, as we've been finding over and over again, affect more than one trait -- often ones that have almost nothing apparent to do with each other. It will be interesting to see, here.
I think the article hints in this direction, though it also does seem to suggest that the RGS14 gene is activated only in the CA2 region... which mitigates against a more systemic effect.
By the way, the RGS14 gene is also present in humans.
B) When the PIN is retrieved off the upgrade card, it and the CPU system identifier are sent to Intel. Intel stores the PIN, and system identifier in its database. Any future attempt to use the PIN will fail unless it's for the same system identifier.
C) Intel uses its private key to encrypt the system identifier, and returns it for storage in the board's non volatile memory. Each time the system is started, the encrypted system identifier is sent to the CPU and decrypted using Intel's public key. If the resulting system identifier matches that on the CPU then the upgrade is enabled.
Assuming that the processing in the CPU is not in modifiable firmware, hacking this would involve deducing Intel's private key, which is not likely to happen. Nor is it likely to be leaked unless Intel's internal security is very lax.
I'd have thought the authorities (there's a relevant legal phrase which eludes me for the moment) would come down on Intel if it didn't make the mechanism available to other MB manufacturers.
I haven't gone into the details of HDCP, but it appears to suffer from the problem that content has to be encrypted by hardware in the possession of the user, and then decrypted by other hardware also in the possession of the user. This means that any keys have to be physically present, or derived from keys that are physically present, and thus in principle subject to compromise. It appears that the master key can be also be deduced.
By constrast, in the scheme I outlined, obtaining the private key either involves a leak from Intel, or cracking the public key encryption mechanism. In the case of the latter, I think a few unpaid-for CPU upgrades would be the least of the problems.
This was done on IBM mainframes in the early '90s. We called it "Rent-a-MIP". The process could be controlled over the phone lines through the normal maintenance interface. Customers paid for the processors they needed that month. The problem reusing this idea here is that each device has to have its own secret (key). Controlling this process is expensive. Otherwise one simply has to copy the certificate that Intel supplies to one legitimate processor.
As I proposed it, each device has the same public key on the CPU. Each CPU needs has a destinct system ID, but it already does. Intel need to track system ID against PIN number, but it's clear from the Privacy FAQ that they'll be doing that.
The expression I was actually looking for was "antitrust". It's more about abuse of market power than a restraint of trade. Buch such abuse is unlawful in many jurisdictions, including the important markets for CPUs (China possibly excepted).
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.