Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Intel Offers Protection Plan For Overclockers

samzenpus posted more than 2 years ago | from the use-mostly-as-directed dept.

Intel 101

MojoKid writes "Intel today unveiled a pilot program that provides warranty protection to overclockers in the event they get a little bit overzealous with pushing the pedal to the metal. For a fee, Intel will provide a one-time replacement of certain processors that are damaged by overclocking and/or over-volting. It's completely optional and in addition to the original three-year standard warranty that already applies to Intel's retail boxed processors. Intel isn't yet ready to flat-out endorse overclocking but the Santa Clara chip maker is perfectly content to provide a 'limited remedy if issues arise as a result of an enthusiast's decision to enable overclocking,' for a modest fee, of course. The deal applies only to certain Extreme Edition and K-series (unlocked) processors currently, in Intel's Core i7 and Core i5 families."

Sorry! There are no comments related to the filter you selected.

Organized trolling campaign by GreatBunzinni (-1)

Anonymous Coward | more than 2 years ago | (#38745162)

GreatBunzinni [slashdot.org] has been posting anonymous accusations [slashdot.org] listing a whole bunch of Slashdot accounts as being part of a marketing campaign for Microsoft, without any evidence. GreatBunzinni has accidentally outed himself [slashdot.org] as this anonymous poster. Half the accounts he attacks don't even post pro-Microsoft rhetoric. The one thing they appear to have in common is that they have been critical of Google in the past. GreatBunzinni has been using multiple accounts to post these "shill" accusations, such as Galestar [slashdot.org] , NicknameOne [slashdot.org] , and flurp [slashdot.org] .

That's not the problem. The problem is that moderators gave him +5 Informative and are now modding down the accused, even for legitimate posts. Metamoderation is supposed to address this by filtering out the bad moderators, but clearly it's not working.

This "shill" crap that has been flying around lately has to stop. It's restricting a variety of viewpoints from participating on the site and creating an echo chamber.

Re:Organized trolling campaign by GreatBunzinni (-1, Offtopic)

QA (146189) | more than 2 years ago | (#38745186)

Fuck Off will you?

Re:Organized trolling campaign by GreatBunzinni (-1)

Anonymous Coward | more than 2 years ago | (#38745278)

Read your sig much?

Re:Organized trolling campaign by GreatBunzinni (-1)

Anonymous Coward | more than 2 years ago | (#38745408)

Sudenly this is consistently getting the firsp post, whereas the circlejerk accounts (it's just not the pro-MS comments, it's also the consistent protection of each other) were doing it before. Nice try though.

Re:Organized trolling campaign by GreatBunzinni (-1)

Anonymous Coward | more than 2 years ago | (#38745546)

it's also the consistent protection of each other

except that there aren't actually any examples of that happening

i guess you saw you were getting downmodded on your GreatBunzinni account, so now you're back to AC

Re:Organized trolling campaign by GreatBunzinni (1)

geekprime (969454) | more than 2 years ago | (#38745644)

you are posting AC to criticize posting AC?

YOUR head should asplode from the idiocy.

Re:Organized trolling campaign by GreatBunzinni (1)

lightknight (213164) | more than 2 years ago | (#38747706)

Agreed. /.ers do not need to be handheld to see any bias from the mods: that's what the meta-mod system is for. If there is any bias, it will be accounted for and nuked.

However, these consistent first posts from Anonymous Cowards with off-topic discussions is a source of annoyance. Perhaps he thinks he is doing the rest of us a public service; he is not.

Re:Organized trolling campaign by GreatBunzinni (1)

geekprime (969454) | more than 2 years ago | (#38745640)

People being paid to push a viewpoint can afford to haunt /. to make their owners point.

Re:Organized trolling campaign by GreatBunzinni (1)

geekprime (969454) | more than 2 years ago | (#38745634)

I've seen way more of your bullshit than anti-microsoft bullshit.

Come to think of it though, I HAVE seen a great deal of PRO-Microsoft bullshit lately so perhaps he has a point.

Why not for all CPUs? (-1)

Anonymous Coward | more than 2 years ago | (#38745200)

In other words, Intel are assholes.

Re:Why not for all CPUs? (5, Insightful)

tywjohn (1676686) | more than 2 years ago | (#38745226)

Really? As apposed to offering nothing?

Re:Why not for all CPUs? (-1)

Anonymous Coward | more than 2 years ago | (#38745450)

I'm offering you my middle finger. Am I a generous and good guy now?

Re:Why not for all CPUs? (1)

causality (777677) | more than 2 years ago | (#38745516)

I'm offering you my middle finger. Am I a generous and good guy now?

That depends: do I get to keep it?

Re:Why not for all CPUs? (1, Redundant)

Samantha Wright (1324923) | more than 2 years ago | (#38745674)

No, but for a fee, Intel will provide a one-time replacement of certain fingers that are damaged by overclocking and/or overvolting. It's completely optional and in addition to the original three-year standard warranty that already applies to Intel's retail boxed fingers. Intel isn't yet ready to flat-out endorse overclocking but the Santa Clara digit maker is perfectly content to provide a "limited remedy if issues arise as a result of an enthusiast's decision to enable overclocking," for a modest fee, of course. The deal applies to only to certain Extreme Edition and K-series (unlocked) fingers currently, in Intel's middle and ring families.

Re:Why not for all CPUs? (1)

causality (777677) | more than 2 years ago | (#38762486)

Yeah I read the summary. What is wrong with you?

Re:Why not for all CPUs? (1)

Samantha Wright (1324923) | more than 2 years ago | (#38762632)

Either my humour is not understood (oi moi) or you have just made a very cruel assessment of Intel's products, in which case I salute you.

Re:Why not for all CPUs? (1)

lsatenstein (949458) | more than 2 years ago | (#38760956)

Because the CPUs that are being overclocked may have defective microcode and the replacements may contain an updated design

Re:Why not for all CPUs? (5, Informative)

EnempE (709151) | more than 2 years ago | (#38745268)

Not all CPUs are suitable for overclocking. If they tested a chip at the factory and know that it won't survive if run at a higher voltage or clock speed than it is required too it would be bad practice for them to encourage you to operate the chip in a manner that will make it unstable, which could at best cost a bit of time and money when the chip fails, or worse cost you a massive amount of time when the chip operates poorly and causes intermittent failures or incorrect calculations.

Re:Why not for all CPUs? (4, Insightful)

fuzzyfuzzyfungus (1223518) | more than 2 years ago | (#38746952)

It is also probable that the magnitude of the additional fee required for Intel to offer this service without losing money would vary sharply between processor families.

The EE and K-series stuff is, shall we say, 'priced for the price insensitive'. Nothing wrong with that, voluntary on both sides, everybody knows that you can get 80-90% of the bang for less than half the buck by stepping back a few notches; but those parts are crazy overpriced. By contrast, their low end parts(especially in areas where they are going directly against AMD largely on basis of performance/$) aren't sold at a loss; but don't have nearly as much profit built in.

If they wanted to offer abuse insurance on value SKUs, and not lose money, the price would likely be a fair percentage of the OEM price of the CPU(very little margin on those parts, and only crazed overclockers would buy the insurance, so a high-risk pool and parts whose cost to intel is not so very different from their cost in store). Offering abuse insurance on the 'because we can' SKUs could be done at a much lower percentage of the OEM price of the CPU, because the cost to intel of that part is much lower than its price, and the entire market for those is crazed overclockers, so the additional riskiness implied by actually buying such insurance is not as great...

Re:Why not for all CPUs? (0)

Anonymous Coward | more than 2 years ago | (#38763022)

I work in "IC device reliability": this is people who figure out how long an individual transistor will last at normal operating conditions which you need to know because you don't want all your microprocessors, etc. sold to sudden fail 3 years out due to end-of-life failure unexpectedly. You want them to fail one day after your warranty period of course! Then it's someone else's problem.

So how do we predict the future short of waiting a long time? We accelerate the aging process with over-voltage and heat, and measure how long it takes to fail. Then you can apply various techniques to extrapolate that back to normal operating conditions. It's actually surprisingly accurate.

So how long will devices last on the latest 20-30 nm process at normal operating conditions (never mind over-clocking)? Right now the target that processes can expect and thus design for is 10-15 years. Not so bad you say. Well strictly correct because the processor will likely cease to be useful long before this. However, as process nodes have shrunk, this lifetime has monotonically declined. 40 years ago, you could comfortably expect 1,000-5,000 years lifetime at larger geometries. As we shrink that 10-15 year span will shrink also and at an increasing rate now.

So what Intel is doing is basically playing on this to make a buck (they are very good at these kinds of games - I used to work at Intel). I haven't looked into what design rule each of the processors is but I'd bet a large amount of money that the biggest "protection plan" they are offering is exactly on those larger geometries that need it the least because they are more insensitive to stress acceleration. Basically this is just a safe bet that plays on the physics and the market demand.

In a few years Intel will "mysterious(?!)" drop this protection plan for the simple reason they can not offer it without a financial loss. The less intelligent folks will get all huffy about "their right to protection" being taken away but what ever. You can only play to stupidity if there's money in it for you.

This also points out another absolute fact: over-clocking as a hobby is doomed in the near future because the margin for how much you need to achieve useful performance increases will run up against how much you need to age-out the processor to end-of-life failure in just a few years now. Probably by 15-20 nm.

Re:Why not for all CPUs? (0, Offtopic)

mcavic (2007672) | more than 2 years ago | (#38745270)

Intel are

Intel = singular
are = plural

Re:Why not for all CPUs? (0)

Anonymous Coward | more than 2 years ago | (#38745316)

Collective nouns!

Re:Why not for all CPUs? (0)

Anonymous Coward | more than 2 years ago | (#38745412)

mcavic = person who seeks confirmation in odd ways

Re:Why not for all CPUs? (0)

mcavic (2007672) | more than 2 years ago | (#38745550)

At least I'm posting under my name.

Re:Why not for all CPUs? (0)

Anonymous Coward | more than 2 years ago | (#38745692)

Thats a strange name you have!

Re:Why not for all CPUs? (1)

mcavic (2007672) | more than 2 years ago | (#38745936)

Or, your definition of the word "name" is too tight.

Re:Why not for all CPUs? (1)

metrix007 (200091) | more than 2 years ago | (#38746234)

Not everyone who reads slashdot posts often enough to bother registering. Why should that matter?

Re:Why not for all CPUs? (1)

mcavic (2007672) | more than 2 years ago | (#38746298)

In general, it doesn't matter. I posted as AC for a while when I was new. But if you're going to go off on someone for not giving away more free stuff, maybe you should be held accountable for your words. When I make a comment like that I usually get called out on it.

But anyway, this is just a friendly banter. I'm not upset.

Re:Why not for all CPUs? (0)

Anonymous Coward | more than 2 years ago | (#38748972)

Maybe some of us post AC because we're sick of getting modded into oblivion for defying the /. groupthink, getting negative karma, and then having all our posts default at -1.

Jesus Christ... (3, Insightful)

jamrock (863246) | more than 2 years ago | (#38745340)

It's a pleasant surprise that Intel is offering this option at all, and you're calling them assholes because they're not offering it for all CPU's? I bet you're also pissed that this optional protection plan isn't free either. You arrogant, entitled jackass.

Re:Jesus Christ... (-1)

Anonymous Coward | more than 2 years ago | (#38745364)

Intel is a business. Businesses are assholes that try to fuck people over. Nothing can change that.

Re:Jesus Christ... (0)

Anonymous Coward | more than 2 years ago | (#38748664)

And customers are assholes that try to fuck over businesses. Nothing can change that.

Re:Jesus Christ... (1)

lsatenstein (949458) | more than 2 years ago | (#38761058)

It's a pleasant surprise that Intel is offering this option at all, and you're calling them assholes because they're not offering it for all CPU's? I bet you're also pissed that this optional protection plan isn't free either. You arrogant, entitled jackass.

My view is that the models that are being protected may have flawed microcode, For $35, the cost to Intel to replace, it is a good deal for both customer and Intel.

Usually all microprocessors (AMD, INTEL, other), have some instructions to fix defective instructions or add some new ones. If this space in the chip is exhausted, the faulty instruction may require code in the operating system's supervisor (kernel) to run in kernel mode. Much better to provide a corrected or improved processor.

Re:Why not for all CPUs? (2)

rhook (943951) | more than 2 years ago | (#38745630)

Because other CPUs are not unlocked and therefore not very good for overclocking.

Re:Why not for all CPUs? (1)

geekprime (969454) | more than 2 years ago | (#38745660)

The unlocked chips are sold at a premium and are with a sold to the overclockers,
Personally I'm an AMD guy unless performance requirements justify the extra $ but I think it's excellent that Intel is extending real support to the overclockers, they are after all the "bleeding edge" of enthusiasts.

Re:Why not for all CPUs? (1)

Opportunist (166417) | more than 2 years ago | (#38750832)

Because, from Intel's POV, you're not supposed to overclock CPUs that they could not sell for more, despite being essentially the same silicon but just didn't pass all the tests well, so it had to be downstepped. It might work with OC, it might not, but in any case Intel would be very stupid (or you would be, depends) to offer that kind of protection for CPUs that are not the peak performers of their line. They have to sell you a chip that would rate at X GHz with Y cores for Z money if everything worked flawlessly, but it didn't, so they have to sell you a chip for X-x GHz with Y-y cores for Z-z money (x, y, z >=0). If you could now undo x and y, they'd either have to charge z (which would mean that you end up with a plus/minus of zero), less than z (which would be stupid on their side because you'd get the same CPU cheaper) or more than z (which would be stupid on your side since you'd buy insurance you wouldn't need if you bought the "good" CPU in the first place).

If they now sell you their top notch CPUs with OC insurance, they not only get some additional dough (rest assured that they calculated VERY exactly how to make money that way, too. Insurance is a numbers game, not a gamble), they can also boast the fastest CPUs in various e-peen benchmark tests people run and upload to comparison pages, plus your rather expensive CPUs most likely won't last as long as they would have without OC.

I hope whoever came up with this brilliant idea got his fair share, that's a million dollar PR idea.

times change (5, Interesting)

Formalin (1945560) | more than 2 years ago | (#38745222)

Never thought I'd see intel go for something like this, although I don't bother with overclocking these days.

from TFA, since the summary neglected it:

Processors in which you can purchase a Protection Plan include:

        Intel Core i7 3960X: $35
        Intel Core i7 3930K: $35
        Intel Core i7 2700K: $25
        Intel Core i7 2600K: $25
        Intel Core i5 2500K: $20

Seems fairly affordable if you plan on burning one up, I suppose.

Re:times change (4, Interesting)

jimmydevice (699057) | more than 2 years ago | (#38745330)

Having tested 8 core processors with and without cooling while testing at Intel, I would say it's a sucker bet.

Re:times change (1)

Anonymous Coward | more than 2 years ago | (#38745380)

It's just a nice option to have and adds some goodwill towards Intel with next to no cost on their part (and potentially some profit) given the durability of their processors.

And for those rare few who do somehow destroy their CPU, it'll give them another one to use more cautiously.

Re:times change (2)

dotancohen (1015143) | more than 2 years ago | (#38746182)

It's just a nice option to have and adds some goodwill towards Intel with next to no cost on their part (and potentially some profit) given the durability of their processors.

Sorry, but Intels are not "durable" processors. I managed to fry a DuoCore2 at 72 degrees in a laptop. This wasn't overclocked, but rather was crunching some serious Matlab. The fan was running at the time.

Contrast that with the Duron that I had up to 108 degrees because I forgot to reconnect the fan wire. I learned of the problem by _smell_. Not a scratch on the thing, it continued to run for at least two years after that. Now, that Duron I did assemble myself with Arctic Silver and the DuoCore was a stock Dell build. But I cannot call Intels durable, at least not compared to really durable silicon like the bottom-of-the line Duron.

Re:times change (0)

Anonymous Coward | more than 2 years ago | (#38746228)

I'm surprised someone who would have use for MATLAB would also make sweeping generalizations on a sample size of 1.

Why not? (0)

Anonymous Coward | more than 2 years ago | (#38746290)

This is Slashdot. The place where Maxtor drives are awesome, even though they die by the thousand, just like Seagate, Hitatchi, IBM (back in the day), speaking of back in the day, does anyone remember Connor?

Uh, right. Where was I?

Anyhow, everything sucks!

Or rather, with manufacturers of everything producing at such volume, one should surely take all claims of durability or lack thereof with a grain of salt. No matter what you're discussing, you will find no shortage of people who have had no problems with a brand, and no shortage of people who had problems with every product they've purchased from said brand.

Re:times change (1)

Richard_at_work (517087) | more than 2 years ago | (#38746368)

Equally, I blew an Athalon XP back in the day just by seeing if the system would post after I installed the memory and CPU. However, I forgot the heatsink...

System was up all of 30 seconds, if that. Posted fine, got to the "no system disk" bios screen and everything.

However, on second boot, the bios informed me through the beeps of death that the CPU was dead, and that was that.

Re:times change (1)

lightknight (213164) | more than 2 years ago | (#38747736)

Thou shalt not forget the heatsink on an AMD processor.

You committed a cardinal sin.

Re:times change (1)

level_headed_midwest (888889) | more than 2 years ago | (#38753202)

The "AMD processors start on fire if you don't put a heatsink on them" bit hasn't been true in a long time. Yes, early AMD CPUs that needed customer-applied heatsinks did not have any sort of catastropic overheat protection, namely the K6s. However, by the time AMD came back from the SECC sold-with-a-heatsink-attached Slot A CPUs to Socket A CPUs, the platform was supposed to have a catastropic thermal shutdown feature. Not all motherboard vendors actually implemented it though, and that's where the infamous Tom's Hardware Guide video where they roasted an old Duron came from. AMD mandated the catastrophic overheat shutdown feature in K8s and some/all Stars core (K10) and later AMD CPUs can throttle like P4 and later Intel chips when they overheat as well.

Re:times change (1)

JamesP (688957) | more than 2 years ago | (#38750812)

Most likely it wasn't the processor that 'passed away'

Notebooks have several areas where excess heat can cause damage (especially on most cheap laptops today). At 72F? Not the processor

It's most likely a power source failure, or another area getting too hot and melting the solder

Re:times change (1)

dotancohen (1015143) | more than 2 years ago | (#38750926)

Most likely it wasn't the processor that 'passed away'

Notebooks have several areas where excess heat can cause damage (especially on most cheap laptops today). At 72F? Not the processor

It's most likely a power source failure, or another area getting too hot and melting the solder

Thanks, I had not considered that. I have no idea where the motherboard on that thing came from.

Re:times change (0)

Anonymous Coward | more than 2 years ago | (#38746832)

Pft, just means you weren't overclocking hard enough. Time to break out the liquid nitrogen :D

Re:times change (0)

mea_culpa (145339) | more than 2 years ago | (#38745566)

Makes some sense actually. This cost seems more likely to cover the actual cost of the processor (the die on the wafer). The $1000 that the enthusiast paid is more of a license and the processor is merely the media.

Re:times change (0)

Anonymous Coward | more than 2 years ago | (#38745818)

No. No it is not at all like that.

Shove your license up your ass.

Re:times change (0)

Anonymous Coward | more than 2 years ago | (#38746126)

In the future maybe we can subscribe to a processor license and for a yearly fee you can always have the latest version of the cpu sort of like those software subscriptions from Autodesk or Adobe! Holy shit, sweet idea for a patent or what?

and they will change the socket or some stuff so y (1)

Joe_Dragon (2206452) | more than 2 years ago | (#38747700)

and they will change the socket or some stuff so you have buy a new MB + maybe new ram as well.

Re:times change (1)

makomk (752139) | more than 2 years ago | (#38746210)

I don't suppose it covers motherboard damage due to something like the unreliable LGA socket fiasco a while ago. (A lot of the sockets were just good enough to work for a while at stock clocks but destroyed themselves and the processors quickly if you overclocked. I think there may still be motherboards for Sandy Bridge on sale with this problem actually.)

Re:times change (1)

tverbeek (457094) | more than 2 years ago | (#38748456)

I don't know why they're calling it a "protection plan", when it's really just "overclocking insurance". Everybody pays into the pool, and the people who get "injured" get payouts.

Does this even happen much? (5, Informative)

ZorinLynx (31751) | more than 2 years ago | (#38745230)

How often do CPUs can fried by overclocking these days?

Modern CPUs have complicated temperature monitoring onboard that will throttle down the chip if it starts to overheat. Shouldn't this protect against 99% of possible damage scenarios?

Re:Does this even happen much? (2)

Formalin (1945560) | more than 2 years ago | (#38745250)

Yeah, throttling and whatnot should make it fairly impervious to heat effects, I would think. Over-volting on the other hand, not as much.

Re:Does this even happen much? (4, Informative)

broken_chaos (1188549) | more than 2 years ago | (#38745348)

Overvolting, last I checked, was the only actual thing Intel won't warranty replace for. If you don't overvolt (outside specs, not outside 'standard' voltage -- on my i7 standard is ~1.20v and overvolting is >1.35v) and the processor dies, it'll be replaced whether it was overclocked or not. And you can get a huge bump on clock frequency on most processors without a single bit of extra voltage (in my case, >700MHz without touching voltages at all).

Re:Does this even happen much? (1)

alexo (9335) | more than 2 years ago | (#38747746)

It's an interesting topic.
What is considered a "safe" or "within specs" overvolting for i5 / i7 CPUs and how much of a frequency bump can one usually get from it?

Re:Does this even happen much? (1)

broken_chaos (1188549) | more than 2 years ago | (#38774712)

Look up the processor's voltage VID range -- it should be on the Intel spec sheets, such as the one for the i7-980X [intel.com] . In the case of that processor, anything between 0.800V and 1.375V is considered 'safe'.

As far as how high a boost in speed you can get out of a processor without changing the voltages, it depends on your specific chip and how close to overvolting you get. Most of the time you can get 1-1.5 GHz extra (sometimes more) out of the chip before overvolting it (rarely do any overclockers actually overvolt chips -- there's little need, unless you're going for some sort of record), and an easy 0.5 GHz or more without even touching the voltage at all.

Re:Does this even happen much? (1)

BenJCarter (902199) | more than 2 years ago | (#38745756)

How often do CPUs can fried by overclocking these days?

I think your base are belong to us.

Buy insurance at your own risk...

Re:Does this even happen much? (1)

DarthVain (724186) | more than 2 years ago | (#38748866)

Yes they all come with Over volt, heat, surge protection these days.

However that is presuming those features actually work. Like any fab, they likely check a few, and the rest are assumed to be OK, which may or may not be the case. In addition, the usual suspect of CPU fail is heat. If your "protection" is also on chip, which is likely also susceptible to heat failure, you just hope that it doesn't fail "first". Also it may be that these protections only work a few times, and degrade the more stress you put them under.

Anyway just guessing here, I have zero actual technical knowlege of this stuff, but that would make sense to me. These features make it much more tolerant than it use to be however, and in the past if you even went a little too far, ZAP. You're done.

They make it better, but is by no means a guarantee, which is now what they are offering. Is it worth it? I am doubtful. If you do moderate OC, probably not. If you buy really high end stuff and extreme OC it? Yes it likely is. Call it, 250$ with 5% fail for 35$, or 600$ with 40% fail for 35$. Depends on situation. What would be interesting to me is how long is that warranty for? It say's 6 months, but I think that is the offer, not the service. As high OC and higher temps do reduce lifespan of CPU no matter what generally speaking (most people that do, likely intend to replace long before fail anyway).

target audience: met! (2)

muel (132794) | more than 2 years ago | (#38745236)

Who are the folks buying high-end processors? Us! Ppl who know their OC business. This is no loss and all gain for Intel in a product category whose ability to differentiate is practically nil for the target savvy audience. Good on them for throwing us a worthwhile promotional bone.

Re:target audience: met! (1)

tlhIngan (30335) | more than 2 years ago | (#38749046)

Who are the folks buying high-end processors? Us! Ppl who know their OC business. This is no loss and all gain for Intel in a product category whose ability to differentiate is practically nil for the target savvy audience. Good on them for throwing us a worthwhile promotional bone.

The high-end processors have unlocked dividers. Sure you can overclock the cheaper chips but that involves running everything at the faster speed. In the old days, that included your PCI and AGP busses, and it could mean that spiffy new videocard forced your system to run slower because it couldn't handle the faster clock. Or memory, which was often also a limiting factor in clock speeds. Heck, you often had PC fun because the hard drive would get corrupted from the faster clock - so things worked, but a week later you're battling complete data loss, and everything works just fine otherwise (there's nothing wrong with the hard drive - it just couldn't talk SATA or IDE that fast).

Intel has marketed for a number of years the unlocked divider chips as "Extreme Edition" - you can set the divider any way you want and run the busses at a normal speed.

Otherwise you might have ot invest in faster RAM and looser timings and all that jazz. Still possible, but it's just easier getting wild overclocks with the unlocked processors than locked ones.

You're already overpaying (0)

Anonymous Coward | more than 2 years ago | (#38745256)

So this is like gravy on the gravy train that's sailing on a gravy boat floating on an ocean of gravy.

Only one thing doesn't make sense...who puts a train on a boat?

Seriously though, the top-end CPU's are way out of a sane budget for their performance, so Intel can probably afford a number of replacements if they get the sales boost out of it from people who don't know any better anyway.

Re:You're already overpaying (0)

Anonymous Coward | more than 2 years ago | (#38745492)

That's what happens when AMD can't compete. And this particular announcement makes it clear that Intel could manufacture processors with higher clock speed guarantees for everyone, if the market situation actually required it. But since AMD has no competitive offering, Intel can rest on their laurels and improve their clock speeds in small increments, while charging high-end users and early adopters $$$ for every 100 MHz increment.

Re:You're already overpaying (0)

Anonymous Coward | more than 2 years ago | (#38746620)

But since AMD has no competitive offering, Intel can rest on their laurels and improve their clock speeds in small increments, while charging high-end users and early adopters $$$ for every 100 MHz increment.

AMD does the same thing with its server CPUs. The last 100 MHz increment is particularly costly.
 

Re:You're already overpaying (1)

lightknight (213164) | more than 2 years ago | (#38747780)

Dude, the Black Edition of AMD's CPUs can be overclocked to ridiculous speeds just with air cooling. With liquid cooling, even further.

Re:You're already overpaying (1)

bemymonkey (1244086) | more than 2 years ago | (#38745712)

So $200 for a 2500K is overpriced? I've been out of the PC hardware game for a while, but a 4-core Sandy Bridge chip with an unlocked multiplier for $200 seems pretty awesome.

Yes, the highest-end CPUs are hella expensive, but that's always been the case - not many of those are sold anyway.

Why Overclock? (3, Interesting)

Anonymous Coward | more than 2 years ago | (#38745276)

With the performance of today's processors, I really don't see any reason to overclock beyond "my clocks are bigger then yours".
Overclocking is a great way to ruin perfectly good hardware that costs a pretty penny to begin with.
Undervolting, underclocking, that I can get behind. Less power consumed, less heat produced, lower energy bills.

When my cheap AMD Quad Core can handle HD Multimedia encoding in a decent length of time, why push it beyond it's capacity for a few seconds, minutes off of that time? For a production studeo, sure, but for a home user? get real.

Re:Why Overclock? (4, Funny)

Formalin (1945560) | more than 2 years ago | (#38745346)

I agree with the sentiment, but the answer is mostly just because.

Why drive 80 in a 60? Why have double the bacon on that cheeseburger? Why is there a market for breast implants and 'male enhancement' pills? Why do billionares want more cash? Why do douchebags have trucks jacked up higher than the roof of my car?

Just because. MORE!

Re:Why Overclock? (2, Insightful)

Anonymous Coward | more than 2 years ago | (#38745592)

I understand what you are saying. My take on overclocking these days is: more is less. Yeah, you may get a tiny performance increase, but your energy efficiency goes straight to hell, and your cooling solution gets more complex, expensive, and (usually) louder. Then factor in the extra electricity bills to feed the computer, and the AC bills in the summer. It seems better to spend the money on a better chip from the start then to try to push something beyond its specs to get the illusion of something for nothing.

Re:Why Overclock? (0)

Anonymous Coward | more than 2 years ago | (#38745620)

The better chips aren't that much better. The best chip can be pushed further. Also, on the side, you forgot to account for the savings in the winter.

Re:Why Overclock? (2, Informative)

bemymonkey (1244086) | more than 2 years ago | (#38745798)

"your energy efficiency goes straight to hell, and your cooling solution gets more complex, expensive, and (usually) louder. Then factor in the extra electricity bills to feed the computer, and the AC bills in the summer."

So faster versions of the same chip don't use more power when they're sold as the faster version? Even though they're both rated at the same TDP, the 3.4GHz version of the chip might (and probably will) use [(3.4/3.2)-1]% more power than the 3.2GHz chip of the same type at full load.

Modern example: Will a Core i7 2700k use less power than a 2600k with its multiplier bumped up to the 2700k's stock multiplier? I HIGHLY doubt it.

Also: It's long become possible to overclock without losing any power-saving functions - things like Speedstep still work, so the only time you're using more power is when the machine is at full load - and at full load, the overclocked machine will be done faster, offsetting the jump in your power bill. IIRC power usage actually scales linearly with clock speed (provided all other factors are the same, and forgetting efficiency issues), so 30% more clock speed => 30% more power usage for the processor, but the required time drops to ~77% for CPU-limited applications... which exactly balances out that 30% bump in power consumption.

And when you consider that the CPU only makes up, say, half or 3/4 of the PC's full-load power consumption, that figure gets better - say the other components make up half of the full load power consumption, so that 30% overclock only results in a 15% net power increase - with a 30% speed boost. Even if the CPU uses 75% of the machine's power, you're still only looking at a 22.5% power usage increase.

I know this is all a very amateurish calculation, but I very much doubt that overclocking (within certain logical limits, like no adding voltage) a CPU will increase your power bill if done with efficiency in mind.

As for "complex cooling solutions": Intel chips from the last few years hardly need 'em unless you're going for a very high overclock (with overvolting and so on) or are bothered by the sound of rushing air. 45nm Core 2 Duo gen 45nm chips will often happily run at 50C full load with a 30% overclock on stock cooling... without any noticable increase in fan noise. I've been told it's similar with current Sandy Bridge chips...

Re:Why Overclock? (0)

Anonymous Coward | more than 2 years ago | (#38746798)

This
I have a i5-2500k OC by a modest 1Ghz, a 30% increase in speed on a .05v increase in voltage, a ~4% increase. And the processor still has Speedstep enabled and sits at its minimum clock rate most of the time, using no more power than it would normally. And when processing spikes hit, it spikes speed and finishes the work sooner and returns to minimum clock faster.

Re:Why Overclock? (1)

Kenja (541830) | more than 2 years ago | (#38745352)

Why install Linux on your toaster? Cause you can.

Re:Why Overclock? (1)

ThatsMyNick (2004126) | more than 2 years ago | (#38745672)

Why settle for less? With todays processors and their temperature sensors, that can throttle the clock if necessary, you cannot ruin your hardware.
 
About the benefits, I do like the tiny bit of snappiness when I use slashdot (and some of the JS heavy websites). You would see significant benefits for any single threaded task you perform. For encoding you might want to try overclocking your GPU and see the difference. I can guarantee it would be significant percentage increase, and not seconds.

Re:Why Overclock? (0)

bemymonkey (1244086) | more than 2 years ago | (#38745730)

"When my cheap AMD Quad Core can handle HD Multimedia encoding in a decent length of time, why push it beyond it's capacity for a few seconds, minutes off of that time? For a production studeo, sure, but for a home user? get real."

When an encode takes 5 hours or longer, cutting that time by 30-40% is pretty awesome.

Yes, there is a limit that "normal" users should not go past (nice low temps if you're using the machine for something strongly CPU-limited.

Re:Why Overclock? (1)

bemymonkey (1244086) | more than 2 years ago | (#38745750)

WTF? Slashdot ate my post (and no, the machine I'm posting from isn't overclocked :P). Last sentence was supposed to be:

Yes, there is a limit that "normal" users should not go past (nice low temps, no overvolting), but if you're using the machine for something strongly CPU-limited, not using that untapped potential is a waste.

Re:Why Overclock? (1)

rrohbeck (944847) | more than 2 years ago | (#38745864)

It makes a huge difference if you really need cycles. I'd need dual Opterons or Xeons for the equivalent compilation performance of my FX-8150 at 4.5GHz. That means $2k+ vs $250.

Re:Why Overclock? (1)

Anonymous Coward | more than 2 years ago | (#38746128)

There's isn't an earthly reason why any normal user would want to overclock. It's hobbyists doing it for fun and enjoying the challenge of designing/building coolers etc. Good luck to them... but unless you *are* doing it fun... forget it. You aren't doing yourself any favours by thinking you are getting something for nothing.

Re:Why Overclock? (-1)

unixisc (2429386) | more than 2 years ago | (#38746386)

With the performance of today's processors, I really don't see any reason to overclock beyond "my clocks are bigger then yours". Overclocking is a great way to ruin perfectly good hardware that costs a pretty penny to begin with. Undervolting, underclocking, that I can get behind. Less power consumed, less heat produced, lower energy bills.

When my cheap AMD Quad Core can handle HD Multimedia encoding in a decent length of time, why push it beyond it's capacity for a few seconds, minutes off of that time? For a production studeo, sure, but for a home user? get real.

I agree w/ this. Previously, overclocking made sense when multi-processing techniques weren't mainstream in software, which is not the case today w/ dual, tri, quad, oct-cores. The best way to architect a CPU is to come up w/ the most optimal - not maximum - speed for the CPU, and then replicate it as many times as needed in an MCP. That way, one gets the best bang for buck.

I too would rather underclock chips if their MTBF can be extended significantly, in return for having more cores. Undervolting however - what voltages to these CPUs run internally anyway? 1.8V? Below that, I believe the logic levels would fail, thereby making them in-operable. Same thing goes for overvolting the thing. Who's the idiot who first thought of that idea - overvolting? Overclocking, I can understand, but not this.

Re:Why Overclock? (0)

Anonymous Coward | more than 2 years ago | (#38746796)

Just because you don't stress your processor to the max, does not mean others don't also.

Many games are still single process affairs. I can get 1ghz boost to all four cores at the click of a button now, making games that play ok: liquid smooth.

Re:Why Overclock? (0)

Anonymous Coward | more than 2 years ago | (#38752098)

Where I work we run big simulations that eat up every cpu cycle we can give them.
If we could get 10% more out of a machine, it would save us money.

Is there a pointer as to how much these CPUs can be speedup, the reliability of the results, and the performance improvements?

Sounds to me they are relying on fuses (0)

BlueCoder (223005) | more than 2 years ago | (#38745342)

All in all it's hard to fry a modern processor. Probably some time back they put fuses in the processors just in case of short circuit since then know people overclock them. It's probably sophisticated enough now that they can actually recycle the processors. Simply tear down the package and replace the micro fuses and install in a new package. They might even rely on this mechanism in their processor testing and slotting. So if they are already recycling processors with blown fuses why not insure extreme edition processors as well for a little extra. What they insure the processors for covers the costs probably but it makes buying and overclocking the expensive premium extreme edition a little less scary for those that might put down the money.

Re:Sounds to me they are relying on fuses (3, Informative)

germansausage (682057) | more than 2 years ago | (#38745468)

Not one word of your post is true...Sorry.

Re:Sounds to me they are relying on fuses (5, Informative)

Khyber (864651) | more than 2 years ago | (#38745676)

"Simply tear down the package and replace the micro fuses and install in a new package."

Would you like to know how I know you don't have experience in this field?

Chip lithography is very much a one-time thing. Once it's made, you aren't adding on anything else. Spare silicon is gone. If it breaks, you're screwed, get a new one or nothing at all, those are your only answers.

The only recycling likely to happen will be melting the package down to get the metals out, and Intel would leave that to a reclamation company. There would NEVER be a refurbishing plant made, I can almost guarantee you this, as it's cheaper and easier (plus more logistically sound) to make a new one.

Re:Sounds to me they are relying on fuses (2)

artor3 (1344997) | more than 2 years ago | (#38745740)

Well, technically you can repair some types of damage to the metal or add new connections (with FIBs for example), but it's expensive, and might induce some latent failure modes. Plus with CPUs you need to repackage it (glob top wouldn't make a good thermal connection), which can be extremely difficult. Useful for failure analysis, but no one would even think of repairing a customer return.

Re:Sounds to me they are relying on fuses (1)

iamacat (583406) | more than 2 years ago | (#38745826)

You are forgetting the possibility of soft fuses that can be reset or rerouted without opening the package.

Re:Sounds to me they are relying on fuses (1)

lightknight (213164) | more than 2 years ago | (#38747814)

Any evidence?

Brilliant business move (4, Interesting)

bikin (1113139) | more than 2 years ago | (#38745824)

This is a brilliant business move from Intel in every sense. This is what should go to the Harvard Business Review instead of Use Case Studies that can mostly be attributed to luck.
  • It encourages people who know what they are doing to overclock already powerful CPUs, which means they can demonstrate machines that will hardly be surpassed by the competition.
  • It is pretty low cost, because the user pays the protection AND their variable costs on new CPUs are low (most of their costs are fixed, in development, factory building, manufacturing line assembly, etc.).
  • Generates good will.
  • An overclocked processor will either fail soon or not fail at all... which means replacements will happen while the processor is still being manufactured.
  • By the time the processor fails, is sent, comes back, etc. a lot of time is lost, and the processor value is likely to have gone down, which will likely discourage fraud by sellers trying to pass overclocked processors to unsuspecting clients.

Re:Brilliant business move (1)

rolfwind (528248) | more than 2 years ago | (#38746862)

An overclocked processor will either fail soon or not fail at all... which means replacements will happen while the processor is still being manufactured.

By the time the processor fails, is sent, comes back, etc. a lot of time is lost, and the processor value is likely to have gone down, which will likely discourage fraud by sellers trying to pass overclocked processors to unsuspecting clients.

These last two reasons seem at odds at each other. It will fail soon or not all all... but so much time will be pass simply sending it back?

It doesn't matter when the chip comes back to the house from intel's side, just when they send it off. With priority mail, even from the other side of America from Intel, it will take 3 days maybe to reach them at a cost of $5 or so. Give them a couple of days to process it and it's still under a working week.

I don't think that is enough time to drive down market price in most cases, the world is not moving that fast.

A Limited Remedy? (1)

guttentag (313541) | more than 2 years ago | (#38745932)

A limited remedy if issues arise as a result of an enthusiast's decision to enable overclocking, for a modest fee.

Just how limited is this remedy? For this modest fee, do they send an engineer in a bunny suit to your home/office to laugh at you and suggest that you not do that again?

Extended warranty! How can I lose? (1)

shoppa (464619) | more than 2 years ago | (#38746704)

This is just Intel jumping on the "extended warranty" bandwagon. "Extended warranty" always means gigantic profits for the guy selling the extended warranty.

Ever notice how hard they push extended warranties at the electronics and computer stores? There's a good reason, there's a huge profit margin in them. I bet they pay out $1 for every $20 they take in.

Only chumps buy the extended warranty. Maybe this is a sign... overclockers are chumps?

Insurance premium or replacement fee? (1)

badzilla (50355) | more than 2 years ago | (#38746744)

Do you buy this "insurance plan" in advance, just in case? Or wait until you fry your processor then sign up quick and make a claim? How will Intel know?

Just how far.. (2)

delta98 (619010) | more than 2 years ago | (#38746874)

do you want to over clock? I have pushed machines hard and I found that once we hit dual cores @ 2g things ran just fine;I had no need to overclock a machine since.The hardware is fine now,lets get the wetware up to par.

Re:Just how far.. (2)

drinkypoo (153816) | more than 2 years ago | (#38746908)

I OC'd my Phenom II X3 720 from 2.8 to 3.4 GHz with a small voltage bump. I anticipate it will last longer than I own it even so. It drops to the same clock rate when idle. I can actually notice the difference in some CPU-intensive applications, like strategy games or data compression. I had to install a $20 heat pipe cooler to accomplish this, but I wanted to anyway as the stock cooler sounds stupid.

Re:Just how far.. (0)

Anonymous Coward | more than 2 years ago | (#38747222)

got ya. I was wondering/thinking out loud.Of course ymmv and it was a general question.I'll get slammed for this but the average has changed in as far as the level of the "bar" is concerned. I like the idea of protection but I still ask XXXXX nevermind. I'm still thinking.I might just rethink as I type .

The nineties are calling... (0)

Anonymous Coward | more than 2 years ago | (#38746958)

Who does still overclock?

I will never buy a used PC.

Mind that Intel is not Maecenas (1)

Artem Tashkinov (764309) | more than 2 years ago | (#38747874)

I've read several times that Intel CPUs cost a little over $30 to manufacture, so don't think of Intel as Maecenas. Of course, they are lowering their profits by doing this, but they also give a lot more people the incentive and opportunity to overclock without fearing consequences like burnt $1000 CPUs.

No OC need these days. (1)

g0bshiTe (596213) | more than 2 years ago | (#38748430)

There is absolutely no way anyone can justify OC'ing with today's hardware. Back when you could squeeze 50 - 100 MHz out of a CPU, that's a bit different. Today there is really no need to unless you have older hardware and are trying to avoid a pricey upgrade.
Load More Comments
Slashdot Login

Need an Account?

Forgot your password?