Intel Patents Anti-Overclocking Technology

VCAGuy writes "It appears that Intel has pantented a crystal-locking technology to lock processors to the processor's clock speed. The Inquirer has a story about it, and you can read the patent description from the USPTO. Let's hope AMD doesn't try to copy this..."

It will be cracked

[PyrotekNX]

everything released as of yet has been cracked

yay, overclocking locks...

[digipak]

Just another way to ruin the life of the geek. Go Intel, make your chips even less appealing. /me pokes his Athlon XP

only a matter of time

[petronivs]

It's only a matter of time before the overclockers find a way around this. Intel will likely have some kind of undocumented override in place to make it easier, even.

AMD Won't...

[C0LDFusion]

...it's not in their best interests. The people that they get much of their profits from are overclocking enthusiasts, or at least people who consider the ability to overclock to be a plus. AMD most likely won't follow Intel in this, just like it most likely won't hold back 64-bit.

It's just another reminder that AMD+Linux=Good!

so?

[dAzED1]

I am *totally* with the anti-pantent bloat movement. But...what's the complaint on this one? That the technology is being used, or that its being patented? If its that its being used...wah. If its that its being patented - can someone explain why it isn't a valid patent?

Sure, crystals have been used to lock frequencies forever...but processes are what are generally patented, and the process of locking a processor speed with a crystal (versus locking a signal frequency, or whatever)...is it not new? Can someone explain prior art? Or is this just a case of complaining about any old patent that gets approved at all?

Wrong?

[spanky1]

While AMD processors might "crack" when you install the heat sink incorrectly, who has cracked the Intel multiplier lock introduced so long ago? Nobody.

A.K.A. "Suicide"

[Michael_Burton]

Patenting the technology isn't the same as bringing it to the marketplace, and maybe it's intended for some other purpose, like guaranteeing the reference frequency for some time-sensitive circuitry or radio-transmitter chips or something like that.

But if they're trying to tie the hands of hardware hackers, then Intel is shooting themselves in the foot, and AMD has just got a big win on a forfeit.

Re:AMD

[jimm]

Yes they can if they license the patent.

Re:What's the point?

[spanky1]

Instead of someone buying a 2.4GHz processor, for example, they'll get a 1.8 and overclock it to 2.4 (or whatever the exact numbers are). Basically Intel is wanting to ensure people buy the more expensive processor instead of overclocking a cheaper one. But what percentage of people actually overclock? 0.1%?

This is a GOOD PATENT!!!

[swordgeek]

OK, let's just get something out of the way. This is a good patent. It patents a specific method of achieving a technological end. It is directed, nonobvious, and something which would hurt their VALID intellectual property ownings to have given away to their competitors.

This is exactly the point of the patent office--to protect innovative technology. Intel has nothing to be ashamed of for patenting this, dammit.

Now if you don't LIKE the technology they've patented, then don't BUY it! If they put this on future CPUs, don't support them if you don't want. But DON'T WHINGE ABOUT THE PATENT BEING JUNK! It's not.

Re:underclockers left out

[tomhudson]

... how about underclocking to keep them cool?

Re:AMD Won't...

[bartman]

The funny thing is, that if AMD even wanted to stop people from overclocking using this patented technology, they would have to pay royalties to Intel. So as a result AMD will probably not follow that route and the consummer wins!

AMD+Linux=Good in deed.

Re:AMD Won't...

[Drakonian]

Much of their profits? What would you honestly estimate the percentage of comptuer users who overclock their CPU? I'd guess well below 0.1%.

It would probably only upset a few of their customers who aren't upgrading anyway because they are overclocking.

It's About Fighting Resellers

[zealot]

I've seen a lot of comments here asking why Intel would do such a thing, why they're trying to prevent overclocking even though it voids the warrantee.

They really aren't concerned so much with enthusiasts... the percentage of people who over clock in the total PC market is very small (they just speak loudly online).

The problem they have is with resellers (ie whitebox shops) taking a slow processor (say a P4 2.0 GHz), overclocking it, and selling it in a system as, say, a P4 2.8 GHz and marking up the price as such. To clarify, these resellers do not tell their customers the system has a P4 2.0 overclocked to 2.8 GHz and that the warrantee is voided, they say it has a P4 2.8 GHz part in it, and pocket the extra cash. So Intel loses money on sales of its higher end parts, and customers aren't getting what they paid for: they end up with an overclocked part that may or may not be completely stable.

Lazy Thinking - Major Cause of Blanket Statements

[JohnDenver]

First, I'm not making claims this is uncrackable, but you have no problem making claims that not only is it crackable, but it will be cracked, because you're under the delusion that everything has been cracked.

I'm not going to bother making a huge list of things which haven't been cracked, instead I'll give you one: RSA Encryption

RSA isn't uncrackable. It's not designed to be uncrackable. Instead, it's designed in such a way that cracking it will take a VERY VERY long time with today's technology. (Hundreds or thousands of years, depending on the key size?)

RSA will probably be cracked on some level in the future, but it realistically it won't be cracked in this decade or two or five, which is good close enough for most applications.

Maybe this won't be technically uncrackable, but what will one have to go through to crack it? Cracking Hardware isn't like cracking Software.

What about underclocking?

[Masem]

My next computer purchase for a linux box, I plan to get a mid-range chip, then underclock it a few notches as to reduce it's operating temperature and thus extend the reliability of the chip. I don't want to spend a fortune and worry my hair out over whether my CPU is running too hot or not. While I can understand Intel's concern with resellers of their CPUs falsing advertizing faster chips but in reality selling overclocked ones, I hope Intel realizes that it's better to allow consumers to be able to over/underclock the CPU, and instead pursue legal actions against resellers that fake CPU speeds, instead of going for an overpreventative hardware solution.

Re:My processor is my processor...

[pVoid]

This has nothing to do with DRM... and everything to do with specs.

You do NOT buy a 110V hair dryer and stick it in 220V just so your hair dries faster. In the same way, overclocking isn't a design spec... it's pure and simple not safe and stable, even if your computer *looks* stable. Small instablities tend to only manifest themselves after a server has been up for a long time under lots of load... not right after a reboot... Just because you don't see them, doesn't mean they aren't there.

On a side note, neither the dryer manufacturer nor Intel will provide support for products used out of spec... but there isn't a slashcode site where people bitch about how so and so makes hair dryers that burn your hair off if you put em in 220V... (get my point?)

Re:yay, overclocking locks...

[Daetrin]

What idiot on a corporate IT team would overclock a CPU? Not many worth their paycheck, that's for sure. At least not while it has any value on the books.

You're not making any sense, they're trying to promote the anti-overclocking technology as a _selling_ point, especially to big corporations. They know that most big corps wouldn't overclock the CPU, and they're reasuring them that a third party won't secretly overclock the CPU and then sell it to them.

As the previous poster pointed out, they're marketing to the group they expect to make the most from. They know there are people who like to overclock their CPUs, but that number is fairly small compared to the number of CPUs they sell to corporations, who want assurance of quality.

Re:curses...foiled again!

[Angry White Guy]

Unfortunatel, as we hit the high speeds of processors, the performance gains are not justified. I have a Celeron 300 overclocked to 450Mhz. That's a 50% increase in speed. You can (without using liquid nitrogen) not get such an increase on modern hardware. Overclocking was good for the PC industry. I'd be hard pressed to say that it still is. It is a little side-note in the home-pc history, that's it.

Re:yay, overclocking locks...

[Darren Winsper]

What was wrong with the old method of stamping the frequency on the chip? My old K6-200 had "200MHz" (or sommat) stamped on one of the corners. You couldn't change it or remove it without making it obvious it had been tampered with.

Keeping the key with the lock

[muzzmac]

Your argument against the parent post is correct. However your example is clearly not the same as the technology that the parent refers to.

I would also put it forward that the parent had no idea what he was talking about though.

When trying to encrypt media in things like DVD's, satelite feeds etc etc etc you need to encrypt the data so that the bad guys can't interrupt it and you need to decrypt it so the legitamate users can read it.

I think this is what the parent post sorta meant. (I don't believe that really)
RSA encryption is not the same thing. If someone gave someone to you encrypted with RSA encryption and also gave you the decryption key it would be cracked. Not the encryption itself but the decryption key can then be compromised. This is the reason that most people today believe it is impossible to safely protect media from copying but still allow it's use.

The Intel thing is different again as I assume (having not RTFA) that the protection would be embedded on the chip. You would need a pretty steady hand to modify something on a CPU at the scale it is fabricated I would guess. :-P

Also, the protection is not trying to protect someone copying data so encryption technologies are not the trick. It is trying to stop you using more CPU cycles per second. I think this could probably be done in a way that is not accessible (price wise) to the average consumer. Let's face it the only reason overclocking is popular at all is because it is free. If it cost much more money you would just buy faster CPU's on day one.

Re:This is a GOOD PATENT!!!

[Anonymous Coward]

It's totally obvious. If you've ever worked in an RF lab, you have probably seen a device called a frequency counter [google.com], which is a standard piece of lab bench equipment that does exactly this.

It seems to me that these days, applying something obvious in a novel context is increasingly being considered the same as actually inventing something novel, at least in the patent application process.

I'm pretty sure this is not always going to be a good thing.

--FP

Re:so?

[greenrom]

You shouldn't be able to patent processes that are obvious given current prior art. If you look at the diagram in the article, what's being proposed isn't anything particularly unique or new.

According to the diagram, it looks like they use an input clock to drive a counter. Then, after a set number of cycles of the internal crystal oscillator, you look at the value of the counter. If it's above a certain number, you know the input clock is too fast (somebody is overclocking it).

This is EXACTLY how a frequency counter works. Only frequency counters do some extra math so they can display the frequency in Hz or MHz, or whatever is appropriate. This is a simpler case because you're only concerned with crossing a set threshold.

So really, what you have is a patent for a design that has been around as long as crystals and flip-flops existed. The only thing that's really new here is that they're using it to prevent people from overclocking their processors. In my opinion, you shouldn't be able to get a patent for that. But what do I know? I didn't think Amazon should have been able to patent a one-click checkout even if they were the first ones to do it.

Re:curses...foiled again!

[luzrek]

I am using it in a media center, but with SMB instead of NFS, running Mandrake 9.1beta (looks like I'll have to get the "stable" version now) and a Haupage 401b WinTV card. So far I haven't had any problems with performance from it (in fact I have been suprised at how well it is performing). Most of the suprising performances coming out if it seem to be related to the "off-chip" processes like sound, video, and TV, although boot-up times are amazingly short. However, it does not re-build a kernel very quickly. The biggest pain so far has been compiling the ALSA sound drivers. However, I would not recommend using an EDEN processor in a video game machine. Remember that the early TiVo's used a 33MHz processor.

On the other hand for 1000$ you could build a 6 Eden beowolf cluster since they have built in ethernet and the bios for the EPIA-M's support boot from network. (1 harddrive + 6 EPIA-M (Eden 600MHz) + 1 Case + 1 Hard-drive + 6 128MB DIMMs) and still draw less power than one high end AMD or Intel chip.

Re:yay, overclocking locks...

[Natalie's Hot Grits]

It has been rumored that after Intel gets its yield up, they no longer have to rate each CPU for MHZ, and just randomly mark each one however they want depending on what the market asks for. If they need more 2.8ghz this month, they will mark more of them 2.8ghz. Since they are all tested high enough, they don't have to worry about which goes where. Essentially, the same exact CPU is being sold in different packages, with different prices, and as different products.

When Bausch and Lomb did this with yearly, monthly, and weekly contact lenses, they were convicted of fraud. Now you can't find the so called "yearly" contact lenses they used to make, and they don't cost 150$ a pop. Now the eye doctor will tell you to buy the weekly's and clean them every day and keep them until they tear, if you are interested in saving money.

Re:Whats Next ...

[Shadow99_1]

I think they worry more about potential (& for companies like Intel potential is profit) lost profit.

It's very simple from Intel's POV: High Mhz CPU's cost more, thus making the company more money per CPU. Intel has been hurting profit-wise due to lowering their astronomical prices on CPU's/ Intel needs those astronical prices on CPU's to support it's bloated self. Overclocking lets people buy cheaper CPU's & OC them to the level of higher CPU's cutting into Intel's profits. So bloacking OC = Increase in profits.

Intel knows that they have the marketing clot to avoid any real damage to their rep from this. Heck they've managed to make most of the world think they need a "Intel Inside" CPU for years... Worse yet most people (aka most non-technical people) don't realize their are other companies that make CPU's due to Intel advertising. So they loose a few people to AMD as they were just into for the OCing, oh well they gain larger profits from making those people who stay with Intel pay more...

It's not like thsi is the frist time they've tried to limit overclocking...

Re:so?

[greenrom]

Well, for one thing, the processor isn't the only thing that needs a clock. Remember, the data bus needs to be synchronized to the processor clock too. I suppose you could put an oscillator and clock driver in the same package as the processor and have it drive the whole system, but think what that would do. Want to build a laptop that runs at a slower clock when idle to save power? Too bad. The clock is fixed in the processor. Want to make a motherboard with 2 processors? Guess you'll need a bunch of extra hardware to handle the differences in clocks since there's no way to use a common clock for both processors.

There's a bunch of compelling reasons not to embed the oscillator in the processor package. However, the only compelling reason I can see for putting it inside the processor package is to frustrate all the evil overclockers.