Intel quietly adds DRM to new chips

[somebodeez]

I have a question -
Intel's new DRM chips - would the DRM stuff pretty much be dormant/inoperable/powerless without Windows Vista installed?

[ROF]

It doesn't matter which OS is installed. The DRM is hardware based as it should have been all along. when windows vista (if it's anything like the beta release) doesn't like something you are attempting to install you will get a similiar warning that windows XP provides when you attempt to install non-XP drivers. You still have the option to click install anyways but you do so at your own risk.

[Cornucopia]

It doesn't matter which OS is installed. ....You still have the option to click install anyways but you do so at your own risk.

The point is you probably WON'T have that option anymore, whether for software/programs/drivers or for content. So it does matter which OS is installed.

Scott

[ROF]

there was no issues installing software not intended for Vista Beta. It just warns you same as winXP. I fail to see the problem. at least it warns you before trashing your system.

[somebodeez]

Basically what I'm asking is suppose I were to purchase one of these new chips but I wanted to run Xp on it, not Vista. I have absolutely no interest in Vista whatsoever so in my case, take Vista totally out of the picture.

Would all of my programs and content still work as freely as they do now?

[Cornucopia]

All's I know is that you're working with Beta, and a number of things will change before it's FinalRelease. That could be one of them.

Having read Microsoft's "DarkNet" and other "Trusted Computing" papers, and having witnessed Bill himself talk on this and similar issues to university computer guys, I understand what their plan is... That's why I'm worried.

Scott

[ROF]

Basically what I'm asking is suppose I were to purchase one of these new chips but I wanted to run Xp on it, not Vista. I have absolutely no interest in Vista whatsoever so in my case, take Vista totally out of the picture.

Would all of my programs and content still work as freely as they do now?

Yes, as long as you do not violate digital rights management with said programs. In some cases, I'd say, you might even be able to continue violating software based DRM.

[Wile_E]

I just hope future software that comes out doesn't say "Vista only, will not work on WinXP". Because Vista is just a glamorized XP in disguise. Sure, there are other code changes, but there is nothing currently wrong with WinXP for gaming and other applications.

[Morse2]

Interesting thoughts on "will I have to use Vista on a DRM'd mobo". I have one word that sums it up - "antitrust". It would IMHO (for whatever that's worth) be a clear violation of (admittedly commonly violated already) US antitrust laws on the books since the time of Teddy Roosevelt for 2 horizontal monopolies like Intel (AMD is a niche player - Intel is by far the dominant player, no matter what I and other "AMD fans" would like) and Microsoft to freeze out everyone else (various Linux distros, etc) by forming a monstrous vertical monopoly.

Of course the net effect could be to further separate the declining US market from the increasingly important world market - Europe and Asia would be wise to dump both Microsoft and Intel products en masse. Since chip manufacturing has hit an "end point" in terms of development (don't look for a doubling of computing power any time soon!), Chinese made "generic" chips will soon be in a good position to compete with Intel.

Now, as far as the practical ramifications, I wouldn't lose any sleep over it - just buy current gen chips and mobos (or even last generation!) and start building parallel computers. You'll have "failover" redundancy built right in, a lower entry cost, and higher performance than any of the "software limited" computers that everyone else is stuck with. By stocking a few spares, you'll have 100% reliability that few "bleeding edge" machines can match.

Just some quick thoughts!
All the best,
Morse

[vitualis]

Your quick thoughts are highly flawed.

I would say that it is almost a given that if a DRM hardware model is successful insofar of being adopted by the media industries, the same DRM or compatible equivalent will appear on the AMD platform.

Regards.

[vitualis]

Since chip manufacturing has hit an "end point" in terms of development (don't look for a doubling of computing power any time soon!), Chinese made "generic" chips will soon be in a good position to compete with Intel.

Where did you get that weird idea?

There is no endpoint that has been reached. Sure, Netburst has probably reached near its thermodynamic barrier, but the Yonah based chips have a long way to go. The AMD side is progressive well too. I think Moore's Law has quite a few more years left in it. Progress to the 65 nm process as well as multicore architecture will significantly increase computing power.

Regards.

[Morse2]

>>>...your quick thoughts are highly flawed...Where did you get that weird idea?
<<<

Peace, friend, they frequently are but that's part of being human!

>>>...I would say that it is almost a given that if a DRM hardware model is successful insofar of being adopted by the media industries, the same DRM or compatible equivalent will appear on the AMD platform...<<<

Right, and if both AMD and Intel adopt DRM that freezes out anyone but Microsoft's OS and programs (which was the concern I was addressing), that is definitely an antitrust violation of the greatest magnitude. Whether or not anyone in the US would care these days is another matter. I for one am old fashioned enough to wish for a new Roosevelt to "bust those trusts".

AFA processor speeds, it's pretty much a "given" in the industry that we will see AT BEST a tripling of processor speeds in the next 15 years. Compared to the 20 fold increase we've had over the last decade, or the 2000-fold increase of the last 2-1/2 decades, yes I think it's fair to say that today's computers will not be grossly outdone by any COTS systems of the next decade. Buy an AMD64 and a copy of MS x64 or any of the 64 bit Linux distros and you've got most of what will be called "advances" of the next several years, early.

Parallel processing? Great - once programming support is available. But how many threads will be the "industry standard" for parallel programming support? 2? 3? 4? Perhaps 8? (I can already imagine a marketing campaign for a "V8" chip ). FWIW, I'm typing this on a computer that's part of a binary cluster (soon to go to a trinary cluster), so I very much AM a believer in parallel processing. It's easier said than done, as I'm discovering though.....(and it REALLY consumes energy - try 450 watts on idle and a lot more than that when I've got both nodes fired up at capacity).

Historically speaking, the real speed increases of the last 2-1/2 decades primarily came about due to refinements in the lithographic techniques used to fab chips that permitted smaller structures with a greater areal density of transistors. Intel has apparently licked the 65nm fab (thus eliminating the "space heater" feature of their current line), though AMD is rumoured to have "issues" there (not surprising, given their delays in achieving high yields with the 90nm process combined with their SOI system).

Unfortunately, below the 65nm scale there are some heady problems that do not appear to lend themselves to easy solution - one that recently got press was the lack of a suitable high dielectric constant material for use in on-die capacitors. The current ones are not proving suitable to the small fabs and the ones under development have either been grossly expensive or problematic with chip cleaning solvents used between layers of the fab. Perhaps a decade of development work will produce an appropriate material - or perhaps not. Developing novel materials seldom lends itself to a time-table (one reason why we'll not see an economically viable hydrogen economy any time soon, but that's another story).

So, even though we're still a bit short of the absolute quantum limits of scale for pushing electrons around, we're approaching those absolute limits asymptotically (i.e. slower as we approach the absolute limit).

Sorry if this old windbag is rather boring (and weird! ).

Pax vobiscum and All the best!
Morse

[DereX888]

well, obviously no one will be able to produce something like chips with 1nm technology, so in general I agree with you. Moore's Law has to stop at some point, and increased CPU speeds will halt for a while soon, be it 30nm or perhaps 10-15nm; but some new technology, different than the current ones - perhaps use of molecules or something in this design will allow for further development of smaller and faster chips, who knows what future will bring.
However Moore's Law is flawed because it will be impossible to speed-up further any processing machines once they reach molecular structures and speeds close to speed of light, obviously, since it will be impossible to go smaller and faster beyond use of electrons and protons to act as transistors at some point in future
Unles physics as we know it will again cease to be right and new ground-breaking laws will appear (i.e. speeds faster than light, quarks etc).

[somebodeez]

I got lost with

(snip)
...due to refinements in the lithographic techniques used to fab chips that permitted smaller structures with a greater areal density of transistors.

(snip)
...the lack of a suitable high dielectric constant material for use in on-die capacitors.

My head hurts

[jagabo]

Unles physics as we know it will again cease to be right and new ground-breaking laws will appear (i.e. speeds faster than light, quarks etc).

Take a leap into hyperspace:

http://www.newscientist.com/channel/fundamentals/mg18925331.200

[Morse2]

Hi Derex;

Yep, I probably overstated things, since there are proposed technologies like photonic and quantum computing that are nibbling at the edges. I'm skeptical as to whether or not they will ever be viable in the home environment, but then again you never know.....

For the time being though, and without modification, the original Moore's Law has already been dead for a while (doubling transistor density every year was the original definition, which is why the fabs are a little predictable in their size - it's just a math game to double the areal density on each succeeding scale). Applied today that would imply the transition from 90nm to 65nm in one year (that year has been up for a while), 65nm to 45nm in another (2007?!), 45->32, 32->23, 23->16, 16->11, 11->8, 8->5.6, etc. each with one year intervals. And at today's prices, a $10billion USD price tag for fabrication machinery attached (assuming no radical new technologies needed between transitions as we approach the absolute quantum limits). Without further modification (perhaps a doubling of density every 3-5 years?) I can't see Moore's Law being taken seriously in the future. Frankly an old physics and electronics nerd like me has trouble calling anything so mutable a "law"; nice advertising gimmick though.

As a practical matter, since any decent computer of the last couple of years can manage multimedia just fine with a little tweaking, I don't see the point in further impoverishing oneself to "keep up with the Jones' ". And if there is some "killer DRM app" (from the RIAA and MPAA's view) that renders the snazzy new computers useless for AV (without "renting" the media as they desire), then flat sales of the "wundercomputers" will give even the most hardened of business types pause for reflection about the wisdom of giving the public what they really do not want (I wonder how the execs who dreamed up divx are doing these days?). Who will fill the void if the "big two" adhere to some DRM the public finds repellent? Cheap, generic Chinese computers, in exactly the same manner as their $40USD DVD players replaced the $400USD Sony DVD changer in my home AV system (got tired of sitting through all those PUO's for ads and FBI warnings, as well as "region locked" discs that kept me from buying discs when travelling - so the $40 Cyberhome is actually the more desirable unit to me!).

All the best!
Morse

[DereX888]

Morse, I agree completely.
Since day 1 of my first dvd writer I started not only making backups of my own DVDs, but I also did it for the sheer defiance of Digital Restrictions Management existing on them (PUOs, encryption etc). Actually, come to think of it, I do it just for this reason, just to show a finger to those who want to put any restriction on me. As many other people I can easily afford buying all of the same movies again and again if I would ever need to replace broken DVD, instead of making their backups... but why would/should I?
Jackass Valenti and other DRM-advocating ******** won't see even one my penny spent twice for the same product, not now nor ever in my lifetime

[vitualis]

Right, and if both AMD and Intel adopt DRM that freezes out anyone but Microsoft's OS and programs (which was the concern I was addressing), that is definitely an antitrust violation of the greatest magnitude. Whether or not anyone in the US would care these days is another matter. I for one am old fashioned enough to wish for a new Roosevelt to "bust those trusts".

Sorry, but you are completely missing the point on hardware DRM. There is absolutely no reason why hardware DRM will be OS specific just as DVD-ROM drives with hardware authentication are no OS specific. Sure, it may be licensed making it difficult for open source and free OSes, but I'm pretty sure it will not be locked to Windows. With Apple moving to an Intel architecture, it makes such DRM much easier to implement, in fact, on the majority of consumer desktops independent of OS.

AFA processor speeds, it's pretty much a "given" in the industry that we will see AT BEST a tripling of processor speeds in the next 15 years. Compared to the 20 fold increase we've had over the last decade, or the 2000-fold increase of the last 2-1/2 decades, yes I think it's fair to say that today's computers will not be grossly outdone by any COTS systems of the next decade. Buy an AMD64 and a copy of MS x64 or any of the 64 bit Linux distros and you've got most of what will be called "advances" of the next several years, early.

I deny your "given". It was pretty much a "given" that the 65 nm process was impossible 10 years ago on silicon as well due to signal degradation on the increasingly narrow (and thus, higher resistance) pathways. They are now a reality with a number of simple advances in technology.

There are still many technological fronts that can potentially massively increase computer power of CPUs. The lithographic process can continue to be reduced. Improvements in CPU design also lead to marked improvements. Multilayer / 3D CPU design is being investigated.

Even further into the future, using CPUs on a synthetic diamond wafer (yes these are already possible) will instantly allow more powerful CPUs even with no design improvements (as diamond conducts heat much better than silicon). Novel materials like nanotube transisters.

And of course, even further there is optical and quantum computing.

I think that it is short sighted to say that we have reached anywhere near the end of Moore's Law. Correction. Yes, in terms of doubling the number of transistors yes. However, computational power is not entirely a function of the transistor count.

If you do a search on at the New Scientist website for "computer chip" it returns dozens of articles for the last several years on numerous technologies which can significantly enhance chip performance.

Regards.

[Morse2]

Hi Vitualis;

I certainly didn't want to give the impression that I thought the current DRM schemes would limit OS choices! Reading through the previous posts, "stream of consciousness" style, it was just an extension of Somebodeez' concerns about running the software of his choice under Vista (which ARE valid - I have indeed run afoul of MS' "data execution prevention" system that was implemented on XP Service Pack 2 - which can be turned completely off, rather than building a lengthy database of "opted out" processes, but it's a pain that involves editing your boot.ini file). And, yes, in my opinion it would be an interesting situation if the big monopolies were to play the game that blatantly.

AFA increasing computing power, I sincerely hope you're right (since tinkering with computers would get boring if there were no real changes coming along; then again, I still build valve based amps, so I never give up on anything! ) - but I'm afraid I will have to agree to disagree with you for now. IMHO computers are at the same stage of development as automobiles or firearms were in the mid '40's, or manned rocketry in the late 1960's (i.e. small improvements are certainly still possible, but for the most part they are mature technologies that will require fundamental and very expensive breakthroughs to change significantly).

Let's both hope I'm wrong though and leave it at that, though. Friends? (gesture of offering to shake hands on it - in a virtual sense!)

All the best,
Morse

[vitualis]

Sorry, if my writing style is somewhat abrupt!

I enjoy a good discussion and I don't see why anyone needs to be worked up about discussing computer technology.

As of DEP, I have it turned off too (I have an Athlon 64). However, this is due to software incompatibility rather than any philosophical reason. Indeed, as it is supposed to help prevent buffer overrun attacks, I would very much prefer to keep it on (except that Nikon has still yet to fix their software!!).

With regards to CPUs, I personally do not think we are anywhere near reaching the "end of the road" so to speak. I agree with you that plain old silicon chips are nearing the thermodynamic barrier, insofar that we won't see more than maybe an order and a bit of magnitude improvement in performance. However, I'm banking on increasingly novel approaches.

For example (I can't find the article but I remember reading it), there are chips even now that can be reconfigured at a hardware level (essentially flashing it) so that it can do certain hardware tasks but without a specific hardware chip. Now, these are already prototypes of these chips that are no longer "once flash" but rather, can be reconfigured up to 10 times per second.

If this technology matures, it could provide a huge boost to computing. General purpose CPUs are always somewhat limited in specific tasks, which is why we have GPUs on our graphics cards. But imagine if we had a secondary reconfigurable chip that updates itself on the fly to improve performance. For example, say we were encoding MPEG-4 video, the chip reconfigures itself to provide hardware optimisations for MPEG-4 encoding --> massively improving performance for that task.

Etc.

Regards.

[Rudyard]

As someone asked earlier, what exactly will this DRM do to me in these cirumstances:

1 - backing up a commercial dvd.
2 - Using video editing software to cut and paste bits of video.
3 - Downloading mp3's
4 - ripping cd's to my hdd

These common tasks, how will they be affected exactly?

And also i am all set to buy a new PC equipped with PD dual core which up until now I was excited about. If I am not connected to the net (which this pc wont be) will it have any effect on me at all?

[CaptainVideo]

As someone asked earlier, what exactly will this DRM do to me in these cirumstances:

1 - backing up a commercial dvd.
2 - Using video editing software to cut and paste bits of video.
3 - Downloading mp3's
4 - ripping cd's to my hdd

These common tasks, how will they be affected exactly?

And also i am all set to buy a new PC equipped with PD dual core which up until now I was excited about. If I am not connected to the net (which this pc wont be) will it have any effect on me at all?

My opinion is:
1. Won't be affected
2. Won't be affected
3. Won't be affected
4. Won't be affected

You'll still be able to rip DVD's and CD's to backup or create into Xvid/DivX or MP3/Vorbis/FLAC.

However the new technologies such as Blu-Ray, HD-DVD, HDCP DVI displays will be locked down as will any online audio/video subscription services that will probably require Vista in order to use.

But ripping a normal audio CD or DVD should still work under Vista even with Intel's useless hardware DRM. Why? Because you can turn off the hardware DRM according to what I've read.

[jagabo]

As someone asked earlier, what exactly will this DRM do to me in these cirumstances:

1 - backing up a commercial dvd.
2 - Using video editing software to cut and paste bits of video.
3 - Downloading mp3's
4 - ripping cd's to my hdd

These common tasks, how will they be affected exactly?

I suspect there will be two classes of Vista systems -- trusted and untrusted.

Trusted systems will be those with the appropriate hardware (DRM enabled CPU, chipset, sound card, speakers(?), HDMI monitor with HDCP) on which only signed software has been installed. These will play HD DVD, DRM protected online content, etc. It's likely that trusted systems will be allowed to rip unprotected audio CDs. Enough people have come to expect this that Microsoft probably can't disable it. And the content is unprotected to start with.

You can be pretty sure software like DVD Decrypter and DVDFab Decrypter won't be signed by Microsoft. Once you install one of those packages your system will become untrusted. You'll be able to rip DVD's like you do now, but your system will no longer play HD DVD or any DRM protected content.

Downloading and playing MP3? Since there are legitimate sources of MP3 files you'll likely be allowed to play MP3 files on trusted systems. Will you be able to install P2P software without your system becoming untrusted? P2P software like BitTorrent and eMule may not be able to obtain signatures. The RIAA P2P software probably will be trusted -- but since they will use DRM to protect their content they aren't really file sharing, they're bandwidth sharing. Not only will you be paying for the music and video you download, you'll be donating bandwidth to distribute those files to other paying customers.

[paulw]

Jagabo.. Another good reason not to go down the Vista path..

[Rudyard]

Thanks for the replies but one thing I thought was that this DRM doesnt matter which operating system your running as its hardware based? This was mentioned in earlier posts.

If I get my new PC and just run XP should all be cool?

[Trident5]

all i can say is fu intel Hello AMD

[oldandinthe way]

This thread seems to have coupled trusted computing with DRM.

They aren't the same. AMD machines will also be included in trusted computing.

Here is a link to an intro to trusted computing if anyone cares.

http://www.wavesys.com/solutions/TrustedComputing.html

Trusted computing has been languishing for years. Intel has had support in their motherboards for several without software support. Vista may be the last chance for vendors of trusted-computing technology.

[JohnnyMalaria]

all i can say is fu intel Hello AMD

Well, before you run blindly to AMD, you'd better do a Google on AMD's 'presidio' technology...their Trusted Computing platform.

[jagabo]

This thread seems to have coupled trusted computing with DRM. They aren't the same.

But trusted computing is needed to close the last hole in DRM. As it is now you can install software that intercepts audio or video after it's been decrypted and before it gets to the audio and display drivers. Trusted computing will close that hole.

[oldandinthe way]

Vendors of trusted computing solutions are concerned that they will get swept into the DRM morass.

Only time will tell if they are.

Future plans includes subscription services for trusted status, if it carries the DRM baggage with it it will fail and they know it.