Slashdot Mirror
Intel's Anti-Overclocking Technology Simplified
John Thorensen writes "Found a fantastic article on Intel's recent Anti-Overclocking patent at Fastsilicon.com. Worth the read, as it also explains some of the technical and ethical issues of overclocking. Good to see that some tech journalists can still write material understandable by an average person."
Personally, I'd like to be able to underclock better so it would be easier to built a really quiet PC. Although there are a few articles about it, silent PCs are an underserved area of the market.
--- Often in error; never in doubt!
I though it was my property and i could do whatever i wanted with it. Soon we'll be hearing: licenced not owned?
Let's start overclocking, enhancing, and reverse engineering EVERYTHING to protest these laws.
Preventing overclocking is just corportate bs. Remember the liminal messaging of Brave New World, "I'm tired of old things. I want new things. If it's broken, don't fix it. Throw old things away."
In all honesty, people probably break as many chips as they enhance and overclocking helps profits for chip makers. Anyways, you can use this code, compared against the time/date clock to determine if a chip is overclocked. Software/electronic patents are a bunch of bullcrap for things like this because it's so damn simple to recreate the effect.
You can't judge a book by the way it wears its hair.
just my 2 cents.
If you're looking here for something insightful or thought provoking, you're probably looking in the wrong place.
These anti-OC measures are not meant to bug the home tweaker. They are to prevent retailers from OC'ing CPUs and selling them as if they were higher rated CPUs. Apart from the obvious unfairness towards the unsuspecting average customer, it also irritates Intel because of OC'ed CPUs tend have a shorter MTBF. Also Intel has to deal with all these burned out CPUs as they are returned. This ofcourse can have no positive effect on their reputation. Hence these anti OC precautions. I think it's a good thing.
You know it makes sense, a little reminder from jointm1k.
force licencers of the newer computer busses to include some sort of anti-overclock mechinism directly on the bridge chips on the MB itself (since the article says the clock speed isn't controlled by the processor, I'm assuming that's where it's done)
Sure, it's facist, but it seems cheaper and a bit simpler.
-Ryan
AUWYHSTOT (Acronyms are Useless When You Have to Spell Them Out Too)
The article mentions that this would protect against a vendor who sells overclocked CPUs as if they were originally made to run at that speed. But I don't think this is the only reason. They surely want to stop people from overclocking so that they buy a faster CPU instead.
I think if the only point of this was preventing vendor overclock it could be done much easier: Make the CPU tell the motherboard what frequency it was supposed to run at. Then when you start the computer the BIOS would perform a simple check and show a message like "Intel Pentium 4 at 3.5 GHz (OVERCLOCKED! Should be 3 GHz)".
Nobody who intentionally overclocks his/her system would care much about having this banner, it's even a way of bragging about how much you overclocked your CPU. But it should be effective for avoiding vendor overclock.
Actyually, having had expeirence with this, it's not "fraud".....and I can explain why. This dealer advertised and sold a machine of a specific speed, they didn't sell a system DESGINED for that speed. I know someone who got taken by buying some overclocked machines, when this small buisness owner attempted to sue the computer dealer, the judge threw it out, sayign there was no fraud. If he had said it was a system DESIGNED for that speed it would have been fraud.
sell 2 chips, the OEM locked version and the stand alone overclockable version. overclockers happy, OEM consumers not getting ripped off
You are correct. If you buy a chip with this on it, you can do whatever you want. If such a chip exsisted (there are no consumer chips with this technology) it's part of the spec of the chip.
I'm pissed because my AMD Athlon can't factor large prime numbers in O(1) time. I own the damn thing, it should do what I want, even if it was never designed to do it.
Um, maybe this is a dumb question.....
But of the overclockers out there, those of you that have built the ultimate gaming machines, etc....
Aren't you using AMD?
I admit, every PC I own has an intel processor.... and I haven't overclocked a PC in, oh, 10 years or so - the last time I "built" my own machine (I got tired of doing it, I just buy them off the shelf now)
I was kind of under the impression that most people who are building their own machines these days, and intend to overclock, use AMD processors anyway.
Is that not the case? It's a genuine question, out of curiosity, how many of you are actually overclocking Intel vs AMD?
Ok, maybe I just don't get it.
If they can generate this "comparison pulse" inside the chip without relying on the main board's clock signal, why can't they just use that to run the chip? Why bother with using the external source and doing a whole comparison operation?
c) therefore likely to wipe out your data with no recourse for you.
So, if a non-overclocked Intel chip malfunctions and wipes out the data on my machine, I can file a complaint against Intel and get compensation?
Riiiight.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Eventually, with all these profit-guarentee laws, its going to be illegal to purchase a competitor's items.
Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
It wasn't my lawer it was that of a small buisness that I knew the owner of. I don't know if there was an appeal, but the judge (who I would assume was un-connected with the dealer) said that it didn't violate the letter of the law. The guy sold exaclty what he said, a computer RUNNING at a given speed. It's not on the level, and not right, but it's not illegal. Judges are bound by the law, not personal feelings.
If you own a piece of hardware, sitting in your hand, then you own it, not some subset of its functionality deemed 'acceptable' by its manufacturer.
And the manufacturer has every right to design a product that will only operate in that subset of acceptable functionality.
Intel does not give one nuclear frog fuck about the life of their processors.
I think that the nuclear frog fuck should become the standard unit of measurement of disinterest.
Erm, parasitic capacitance is inherent in silicon, it's not produced by bad etching. The reason people talk about parasitic cap so much these days is that it has come to dominate the delay equation for logic paths. This equation basically says (Sum of gate delays + Sum of wire delays + Required Setup Time + Clock Skew = Fastest cycle time).
As technology shrinks (0.25um -> 0.18um -> 0.13um, etc), the gate delay essentially goes to 0 (not exactly, but I'll simplify). The wire delay keeps getting larger and larger. Why? Because as the geometry decreases the width (and spacing) of the wires decreases. Unfortunately the height of the wires is mostly unchanged. As the width and spacing go down the height effect starts to dominate.
Picture the diffence between two skyscrapers in a downtown city block, and two suburban estates on 2-acre plots of land. The suburbs are the older process technologies, aspect ratios are around 1:1, and very far apart. The skyscraper has 10:1 or higher aspect ratio and the spacing is far less than the height. As the previous poster points out, the capacitance depends on the surface and thickness of the layer. It also depends on the area. The two skyscrapers can "see" a lot more of each other -- this causes the parasitic cap to go up, a lot.
Bad etching can make this worse, but in a well-controlled manfacturing process this variation is on the order of +/-10-20%. Really bad problems are due to actual defects (tiny bits of dust) that cause shorts or opens in the circuits, and then the part just fails completely.
If you hotrod your car, and it turns into a pile of trash:
1 - you chose to do the hotrodding.
1a - It wasn't done without your knowledge by an unscrupulous dealer and passed off as a reliable product. Saleen sells modified Mustangs as a unique and improved product under their own name. They don't try to take a V6 Mustang and sell it as a Cobra under the Ford brand.
1b - making those modifications has voided your warranty, so the original manufacturer is now longer liable. It's much harder to tell that a chip has been overclocked, and thus that the manufacturer should not be responsible for claims against lifetime, reliability, or performance.
2 - likely no no one but you even knows about the modifications, so the manufacturer's reputation doesn't suffer.
If you take off the tinfoil hat, you might realize that Intel isn't engaging in some conspiracy against individual overclockers in their garage. Really, they're not out to get you personally.
Back when this was filed, there was a significant issue of unscrupulous "remarkers" relabelling chips as being of a higher speed grade. That potentially damages Intel's reputation if there results in anybody concluding that Intel's processors are unreliable. There are many ways of fixing this problem, of which this is one. Any technology company with their heads screwed on right will investigate many possibilities, and file patents on methods and techniques that look like they might turn out to be promising.
Before screaming "GROW UP INTEL" you should look a little more closely. There's really nothing to see here...
Sounds depressingly familiar... In the UK, until around 1984, it was illegal to obtain telephone service from anyone other than the (government owned) British Telecom (Kingston Communications in Hull) - governments don't like competition! It still is illegal not to subscribe to the state TV company, if you own a TV: you're free to subscribe to other channels as well, but you have to subscribe to the BBC as well - even if you're in a transmission blackspot and don't receive it! For that matter, until not that many years ago, there were still state monopolies or near-monopolies on everything from milk to steel - and you even had a limit on how much money you could take with you on holiday. Of course, in these days of credit cards, that kind of control would be almost impossible to maintain effectively, but back when moving money meant taking cash or travellers' checks, it was much easier.
It's always been a reflex of such governments: if "your side" is losing in a market, instead of competing, just tax, restrict or outright prohibit the competitors. In the UK, taxis, pubs (bars) and farm production are all subject to quotas and often price-fixing - no competition allowed! I'm all in favor of proper regulation - food safety, roadworthy taxis driven by non-axe-murderers and non-toxic drinks - but when the government tries to push prices up artificially, or ban competition to bow to political lobbying from taxi-drivers, farmers or bar owners, it's gone WAAAAAAAAY too far. Who can actually say, honestly, that there can be too much choice for our own good?
bullshit. If they're doing this on purpose to make it hard to break, they're not going to connect some pins you can short to turn it off, or it would defeat the purpose of it (stopping vendors from selling fake O/Ced systems). This would be implemented inside the CPU, so you'd have to crack open the plastic case, at the very least. If they can implement their timer and pulse-counter on the same silicon as the rest of the CPU, you'd have to hack the silicon in a clean-room. That's nice for people with access to a high-quality clean-room (and something to hack silicon with!), but most people would buy a brand-new Alpha workstation instead of buying the gear it would take to even _try_ to O/C such a CPU. (An on-chip implementation would need a high-frequency oscillator, but you can't make inductors in an IC. They'd probably have some sort of laser-trimmed RC oscillator if they did it all on chip.) Even if there were some off-chip components, like a quartz oscillator to provide a reference frequency (they could use a standard freq, and multiply it on chip according to the rated clock speed, which could be burned into a specially prepared area with a laser or something), you'd have to crack open the CPU, which might be mechanically very difficult to do without damaging the silicon, esp. if Intel wanted that to be the case. You could then replace the reference quartz crystal with a faster one. (As long as underclocking was allowed, you could use crystal twice as fast, and then you wouldn't have to replace it every time you wanted to try a different speed.)
:) (and where overclockability is pretty well tested for that kind of CPU), and even then only by a little bit.
Anyway, the issue here isn't whether O/C'ing is still possible, it's whether it's worth it. If you're more likely to destroy the CPU (while trying to "unlock" it, or otherwise) than you are to make it run faster, it doesn't matter what's theoretically possible.
Intel should sic the lawyer on people who sell relabeled CPUs instead of doing annoying shit like this. Buying a 3GHz CPU means you're buying a piece of silicon, and a guarantee that it will work right at 3GHz. All bets are off if you take it beyond that; The guarantee doesn't apply, but it's still your piece of silicon. Not being able to try it at higher speeds makes it less valuable. I hope, as the article suggested, that any CPUs incorporating this are noticeably cheaper than they would otherwise be. I really like stable computers, so I only overclock my older computers that need to feel a bit faster
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter@cordes ,