How does "the end user does not have control over his own computer"? The chip only does what the OS tells it to do, and the user uses whatever OS he wants.
does the owner of the chip (i.e. the owner of the computer in which the chip resides) have full access to all keys embedded within the chip?
From reading the PDF, the answer is sorta. You can ask the chip to generate a new key pair, and then you can later enable/disable/delete that key pair whenever you want. But the private keys don't ever leave the chip.
I think you could record quite a few channels to a single disk if you use good enough scheduling. Let's conservatively assume that each stream is 8Mbps (1MB/s), the disk can write 20MB/s, and every seek costs 20ms. If you have a 1-second buffer for each channel, then writing that buffer takes 20+(1000*(1/20)) = 70ms. Thus you can write 14 streams to one disk.
Yes, it is accurate, but it's still bad. TCPA can be used by content providers to discriminate against users who run unapproved software (and you can imagine that in the future all approved software will be full of DRM). If 99% of computers are running Windows+TCPA or Windows+Pd and they refuse to communicate with anything else, it could make life very hard for the minority.
One big thing is that it provides for "memory curtains", which is a space of memory that a properly signed program can run it that is completely tamper-proof by any other software on the computer (even the operating system). It is enforced by the digital restrictions management hardware.
IIRC, Palladium has this feature but TCPA does not.
Yes, Steve Jobs said smartphones like the P800 are the future of PDAs and that other companies are doing a great job on them, so Apple probably won't make their own.
So what? RDRAM is still overpriced, no matter how you attach it. Not that it matters much in this case, since high-end RISC customers are used to getting screwed on RAM prices.
Doing something different just to be different is not innovation. Doing something different because it's better is innovation. In this case, I don't think there's much benefit to single-address-space systems.
I knew AS/400 had a tagged RAM architecture, but I was under the impression that it was for marking crypto data. Could a scheme like the above be implemented on AS/400, and if so, where can I find some reference info about the MMU?
My impression is that OS/400 has a single address space and a single level store. The hardware is mostly undocumented AFAIK.
Here's what I'm thinking. Ditch pages and memory contexts entirely. Instead, divy up a 64-bit virtual address space among individual processes, say 48 bits apiece. If a process wants to access memory outside it's 48-bits, it would need to access it through special pointers (which, thanks to a tagged-RAM architecture) could written to by the OS (allowing the OS to define its own protection and sharing mechanisms).
Look, it's AS/400!
Doing something "just for the hell of it being new" doesn't seem too popular among CPU manufacturers who are trying to actually make money.
"Downloading a novel from the net is not something I'd ever likely do myself, but mainly because reading novels on the screen of a PDA is something I might get into only if I were incarcerated, with no alternative.... You could have sex relatively comfortably on a platform of books, but not on a platform of PDA's. Hardcover books. Paperbacks might start sliding around. Though I'd still prefer paperbacks to a pile of PDA's." -- William Gibson
Cache coherence in hardware is not necessary to run a single OS image. Software can synchronize and flush the caches, but of course that's much slower.
But the stock Linux kernel requires hardware cache coherence, so the Altix must have it.
I don't know whether "moblogging" will take off or not, but I'm sure telcos will make no money from it, because blogging does not require any help from the network. Blogging is like wi-fi: it's a product, not a service, so people aren't going to pay service fees for it.
Yeah it's cache-coherent; otherwise it couldn't run a single OS image. It uses a similar interconnect as the Origin 3000, which has been described in the literature.
I don't think Steve mentioned this during the keynote, but Apple also released their own version (?) of XFree86 for OS X. I wonder how this compares to the version from Fink.
Slashdot Mirror
Apple sells a DVI-to-ADC adaptor.
Yesterday you were saying Palladium and TCPA are basically the same thing, and bashing them both. With these skills, your karma will go far.
How does "the end user does not have control over his own computer"? The chip only does what the OS tells it to do, and the user uses whatever OS he wants.
does the owner of the chip (i.e. the owner of the computer in which the chip resides) have full access to all keys embedded within the chip?
From reading the PDF, the answer is sorta. You can ask the chip to generate a new key pair, and then you can later enable/disable/delete that key pair whenever you want. But the private keys don't ever leave the chip.
From that Web page: "Supports PIP(picture in picture) function. One for live and the other for captured program."
Looks like it only has one tuner.
I think you could record quite a few channels to a single disk if you use good enough scheduling. Let's conservatively assume that each stream is 8Mbps (1MB/s), the disk can write 20MB/s, and every seek costs 20ms. If you have a 1-second buffer for each channel, then writing that buffer takes 20+(1000*(1/20)) = 70ms. Thus you can write 14 streams to one disk.
Yes, it is accurate, but it's still bad. TCPA can be used by content providers to discriminate against users who run unapproved software (and you can imagine that in the future all approved software will be full of DRM). If 99% of computers are running Windows+TCPA or Windows+Pd and they refuse to communicate with anything else, it could make life very hard for the minority.
One big thing is that it provides for "memory curtains", which is a space of memory that a properly signed program can run it that is completely tamper-proof by any other software on the computer (even the operating system). It is enforced by the digital restrictions management hardware.
IIRC, Palladium has this feature but TCPA does not.
Has Apple announced it's not going to make a PDA?
Yes, Steve Jobs said smartphones like the P800 are the future of PDAs and that other companies are doing a great job on them, so Apple probably won't make their own.
But I'm sure you can run Tremor.
I'm guessing the FPGA just implements some glue logic to connect the peripheral chips (USB, TCPA, etc.) to the EBC bus on the 405LP.
Eclipse *is* a really nice application, but I don't think IBM's motives in creating it were at all community-minded.
:-)
Yeah, they're really hurting the community by replacing proprietary Swing with open-source SWT.
The fragmentation is a little disappointing, though.
No, but why does it matter?
So what? RDRAM is still overpriced, no matter how you attach it. Not that it matters much in this case, since high-end RISC customers are used to getting screwed on RAM prices.
Doing something different just to be different is not innovation. Doing something different because it's better is innovation. In this case, I don't think there's much benefit to single-address-space systems.
I knew AS/400 had a tagged RAM architecture, but I was under the impression that it was for marking crypto data. Could a scheme like the above be implemented on AS/400, and if so, where can I find some reference info about the MMU?
My impression is that OS/400 has a single address space and a single level store. The hardware is mostly undocumented AFAIK.
Here's what I'm thinking. Ditch pages and memory contexts entirely. Instead, divy up a 64-bit virtual address space among individual processes, say 48 bits apiece. If a process wants to access memory outside it's 48-bits, it would need to access it through special pointers (which, thanks to a tagged-RAM architecture) could written to by the OS (allowing the OS to define its own protection and sharing mechanisms).
Look, it's AS/400!
Doing something "just for the hell of it being new" doesn't seem too popular among CPU manufacturers who are trying to actually make money.
Threads and processes are almost the same thing in Linux, so HT benefits them both.
I agree with you except in one case: attestation. The only use for attestation is to discriminate against users who run unapproved software.
Yes, the TPM (Fritz) chips are tamperproof (to an extent) and each one has a unique private key.
To be fair, the de100c was a completely self-contained MP3 jukebox. It's still more expensive than the ZapStation, though.
"Downloading a novel from the net is not something I'd ever likely do myself, but mainly because reading novels on the screen of a PDA is something I might get into only if I were incarcerated, with no alternative. ... You could have sex relatively comfortably on a platform of books, but not on a platform of PDA's. Hardcover books. Paperbacks might start sliding around. Though I'd still prefer paperbacks to a pile of PDA's." -- William Gibson
Cache coherence in hardware is not necessary to run a single OS image. Software can synchronize and flush the caches, but of course that's much slower.
But the stock Linux kernel requires hardware cache coherence, so the Altix must have it.
I don't know whether "moblogging" will take off or not, but I'm sure telcos will make no money from it, because blogging does not require any help from the network. Blogging is like wi-fi: it's a product, not a service, so people aren't going to pay service fees for it.
Yeah it's cache-coherent; otherwise it couldn't run a single OS image. It uses a similar interconnect as the Origin 3000, which has been described in the literature.
I don't think Steve mentioned this during the keynote, but Apple also released their own version (?) of XFree86 for OS X. I wonder how this compares to the version from Fink.