Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

AMD Phenom II Overclocked To 6.5GHz

Posted by CmdrTaco on Mon Jan 26, 2009 12:33 PM
from the freeze-your-bits-off dept.
An anonymous reader writes "During CES a group of overclockers with access to liquid nitrogen and liquid helium for the extra boost of coldness cooled an AMD Phenom II X4 chip to -232 degrees Celsius. Once they got the chip cooled to this frigid temperature, they pushed the clock speed all the way up to 6.5GHz, which is a world record for a quad-core CPU, and then dished out an astonishing 45,474 3DMark05 score!"
+ -
story

Related Stories

This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • by DigitAl56K (805623) on Monday January 26 2009, @12:36PM (#26609321)

    .. to get a decent score in 3DMark ..

  • Crunchy (Score:5, Funny)

    by x1050us (832886) on Monday January 26 2009, @12:38PM (#26609367)
    Numbers must be really crunchy at that temperature
  • from TFA (Score:5, Insightful)

    by caffeinemessiah (918089) on Monday January 26 2009, @12:38PM (#26609381) Journal

    which is a world record for a quad core CPU and they dished out and astonishing 45,474 3DMark05 score! Watch the video below to see how it was done and how history was made:

    Truly PHENOMenal, but I can't help but (cynically, I admit) think about how history inevitably mocks overclockers. Cue back to the 90s and a headline might have read "486 overclocked to 500Mhz -- history has been made!". Like Ozymandias, nothing beside remains...

    • at the time a 486 might have been overclocked to 500 mhz, it would have been a great deal. more precisely, at the time anything has been overclocked to phenomenonal mhz, it has been a great deal AT THAT TIME.

    • Stability? (Score:5, Insightful)

      by SanityInAnarchy (655584) <ninja@slaphack.com> on Monday January 26 2009, @01:07PM (#26609869) Journal

      I know I once bought a specific CPU because I knew it would be good for overclocking. It wasn't a bad idea -- a 1.8 ghz CPU that I could get running at 2.4, at perhaps half or a third the price of a similar CPU at 2.4 ghz, and I'd overclock my RAM, also.

      I learned two things:

      First, you really have to know your stuff. The RAM I had wouldn't overclock very well, and RAM which would cost a bit more. I had the BIOS helping me out, and I still had to fiddle with timings and voltages.

      And second, despite all the stress testing I did, it would still occasionally crash. I never tracked down these crashes until I clocked it back to spec. Once I got a job, I decided that shelling out another hundred dollars or so for a faster CPU was a better use of my time than trying to overclock one, and dealing with the instability once I did.

      Now, that's probably a completely different area than overclocking to 6.5 ghz, but if I really needed that, I imagine it would be much more cost-effective to buy two or three of them. It won't really help rasterized games (that'd be video-card bound), and raytraced games should scale to multiple machines.

      • Re:Stability? (Score:5, Insightful)

        by zippthorne (748122) on Monday January 26 2009, @01:54PM (#26610473) Journal

        I've always been more of a fan of underclocking, myself. Or as you say, regular-clocking.

        ten or even thirty percent just isn't that much of a difference in performance to justify a stability headache OR paying an extra couple hundred bucks.

        To be interesting, the performance improvement per dollar ought to be significantly better than linear, and at least double. Or you need an application that is CPU bound, time sensitive, and has large processing chunks. 30% isn't going to make much difference in UI performance. Spend that money on RAM.

      • Re:Stability? (Score:5, Insightful)

        by Belial6 (794905) on Monday January 26 2009, @02:05PM (#26610629) Homepage
        Overclocking is like tricking out your Honda Accord. It is a hobby in and of itself. It isn't a good idea for people who just want a computer that works well, just as constant modification to an Accord isn't a good idea for people that just want a reliable form of transportation. That doesn't mean that they are not perfectly reasonable hobbies. It just means that they are not hobbies for me, not hobbies for most people, and most people will think you are wasting your time because it is not their form of entertainment.

        Personally, I have purchased a brand new homebrew Amiga clone within the last year, and have purchased 2 C64 clones within the last 5 years. I certainly know what it means to enjoy a hobby that the vast majority of people "don't get".

        The biggest problem with overclocking for the masses is that if you don't enjoy the act of overclocking in and of itself, you can achieve better results through procrastination.
    • Re:from TFA (Score:5, Funny)

      by ahoehn (301327) <andrew@@@edgefactor...com> on Monday January 26 2009, @03:25PM (#26611995) Homepage

      Like Ozymandias, nothing beside remains...

      "My name is G1G4BY73_PU5H3R, Overclocker of Overclockers:
      Look on my 1337 benchmarks, ye n00bs, and despair!"

      • Re: (Score:3, Insightful)

        by Anonymous Coward

        It is one thing to know that in two years you can regularly buy a system twice as fast as today's fastest system

        I doubt it. CPU speeds haven't really increased that much in the past few years. We're reaching the limits of what we can do
        with a CPU as far as speed goes. Even shifting to 32nm wont increase the speeds that much... mostly just lower power usage.
        They wont be running too much faster but we'll have CPUs with many more cores and it'll take more than a couple of years to shift
        to properly distribut

          • Re:from TFA (Score:5, Interesting)

            by Endo13 (1000782) on Monday January 26 2009, @03:53PM (#26612489)

            What you're talking about isn't really "speed" or "faster". The term you are looking for is "more powerful". The GHz is how many hertz per second the CPU runs at... or in other words, its speed. Newer CPUs get a lot more done per cycle than the first 2-3GHz CPUs. And because they get more things done per cycle, they get things done faster, but that's not the actual CPU speed.

            OP is probably right - 3GHz is probably about the practical limit of what CPUs can run at for everyday use. Speeds higher than that so far seem to increase heat too much to be useful for most applications. Think of GHz like RPMs for a car engine. Most car engines run somewhere between 2000-6000 RPM at driving speed, however some get a whole lot more horsepower at the same RPMs and therefore make the car go faster... but the engine is still running at the same speed. There WAS a time though when they were getting more horsepower by increasing RPMs.

  • But... (Score:4, Funny)

    by LWATCDR (28044) on Monday January 26 2009, @12:46PM (#26609503) Homepage Journal

    Can you run FSX and Cryis at 60FPS?

  • by VinylRecords (1292374) on Monday January 26 2009, @12:47PM (#26609513)

    I'm not sure if it's quite -232 Celsius in my apartment but it's pretty close. They probably could have achieved 6.0GHz overclocking using an air-cooled system in my living room alone.

    • Re: (Score:3, Funny)

      by hattig (47930)

      You can't be married. Any woman would have turned on the central heating to max by now, and filled every room with electric heaters on top.

      And they'd still complain about it being cold, even as you sat there sweating like a pig, wearing a wife-beater, with your feet in cold water, and a cold can of wife-beater in your hand.

  • I was there (Score:5, Informative)

    by Rinisari (521266) on Monday January 26 2009, @12:50PM (#26609567) Homepage Journal

    I was there, too. The coolest it got was approximately -242 degrees C; the warmest was approximately -218 degreesC, at least while I was watching.

    The party was the XtremeSystems.org [xtremesystems.org] party at its LV headquarters, and it was sponsored primarily by AMD, DFI, Gigabyte, Cooler Master, and Thermaltake. It seems to me that Commodore had a presence there, too.

    See ThinkComputers' blog [thinkcomputers.org] for some more pictures (disclosure: my article).

      • Re:I was there (Score:4, Informative)

        by horza (87255) on Monday January 26 2009, @01:01PM (#26609765) Homepage

        Why doesn't -242C exist? -273C exists.
        http://en.wikipedia.org/wiki/Absolute_zero

        Phillip.

      • Metric ? (Score:5, Informative)

        by DrYak (748999) on Monday January 26 2009, @01:14PM (#26609949) Homepage

        I don't know what weird kind of units you are using in your part of the world. But the rest of the planet is using Celsius for everyday temperature measures and Kelvin for scientific measures (same step size, different zero).

        And on our scale, absolute zero [wikipedia.org] (0K) is -273C.

        Thus -242C (aka 31K) is pretty legal and possible temperature. (Although maybe not a very common one outside university labs and mad overclocker's basements)

        Now please stop using Réaumur scale and start using what everybody else is using around.

        --

        PS: I checked, -242Ré is indeed impossible on Réaumur scale - 0 K is -218Ré

      • by Hordeking (1237940) on Monday January 26 2009, @01:16PM (#26609969)

        -242C is a temperature that doesn't exist, unless your religion allows temperatures below absolute zero. All we need now is a campaign for Intelligent Cold. ;)

        In this house, we obey the laws of thermodynamics!

      • Re:I was there (Score:4, Informative)

        by Too Much Noise (755847) on Monday January 26 2009, @02:00PM (#26610569) Journal

        Not only are you wrong about where 0K is, you're also wrong about negative temperatures. Temperature is a statistical measure. A positive temperature corresponds to an equilibrium population distribution across a bunch of energy levels, which will have occupancy probabilities decreasing exponentially with energy. A negative temperature is obtained when the population distribution is inverted, for instance in a 2-level system where an external energy source resonantly pumps up the occupancy of the higher level. Presto, $\exp(-\beta E)$ greater than 1 requires negative $\beta$.

        • Re: (Score:3, Funny)

          by dex22 (239643)

          Yeah, and this was the mistake I made. 237, 273... Ah the price I pay to make a joke about "Intelligent Cold"! :)

          Shoulda got me some "Intelligent Dumb!"

  • FIRST POST (Score:5, Funny)

    by A. B3ttik (1344591) on Monday January 26 2009, @12:56PM (#26609677)
    ...with my lightning-fast 486!!!
  • Light Distances (Score:5, Interesting)

    by SuperAndy (1414157) on Monday January 26 2009, @01:32PM (#26610155)
    What I think is really amazing about this is that at a clock speed of 6.5 GHz, each cycle takes around 15 nanoseconds (15 * 10^-9 seconds) to complete. In this time frame light can only travel around 5 cm. Electrical signals travel close to this speed themselves, so the limit of clock speeds is being reached, since the chip itself is on this same order of distance. It is around the point where one side of the chip will not be able to communicate with the other side in a single clock cycle.
    • Re:Light Distances (Score:5, Informative)

      by diablovision (83618) on Monday January 26 2009, @02:13PM (#26610761)

      You're off by two orders of magnitude. 6.5ghz is 153 picoseconds per cycle.

    • Re: (Score:3, Informative)

      by Orestesx (629343)
      With pipelining, a signal does not have to travel across the entire length of the chip in one clock cycle. In modern processors, there are always several (usually 10-20) instructions in process at any given moment.
  • by w0mprat (1317953) on Monday January 26 2009, @01:41PM (#26610285)
    I'm highly surprised and intrigued the chip even worked at -242C (31K!) for a long time it was speculated in overclocking circles that weird things would happen to current silicon much below the temperature of liquid nitrogen. It does seem liquid helium has been tried a few times but this is the lowest reported temperature I have ever seen on a overclocked CPU. It might not mean much for people who don't care about overclocking but I think this is a significant achievement.

    I'm also intrigued by the possibility this chip could have gone faster, it may have become bound by motherboard reference clock and multipliers at this speed. It's not uncommon for the motherboards ability to deliver current to become the limiting factor.
    8ghz is reportedly the outright world record http://www.nordichardware.com/news,5505.html [nordichardware.com] Although I think this was reset to 8.2ghz not long after.
  • Obligatory (Score:3, Funny)

    by smcdow (114828) on Monday January 26 2009, @02:25PM (#26610949) Homepage

    Imagine a Beowul... oh, never mind.

    • Re: (Score:3, Informative)

      by Thornburg (264444)

      An slightly overclocked Core i7 965 (Extreme Edition) in a similar rig (in terms of video cards, etc) scored about 26,000 in the same benchmark (3DMark05).

      So, no, they didn't have to go to liquid helium to be competetive, but going to liquid helium did allow them to set a world record (although I don't see any Guiness Book or other "official" information about this).

    • by Merovign (557032) on Monday January 26 2009, @01:07PM (#26609865)

      AMD doesn't make any $1200 chips.

      Like it or not, that's just not the market they're in. They're doing well at the $200 level, though.

      I'm not particularly concerned that there's little competition in the segment I'd never pay for anyway. I mean, it's nice that there are Maybach Mercedes and McLaren F1's, but that doesn't mean I'm worried about competetiveness in the segment.

      Whereas I'd be worried if there was only one mid-priced performance sedan, especially if it was sub-expectations in some way.

      I don't think AMD is ashamed to have set a record with a $235 chip, in a world previously dominated by $1000+ chips.

    • Re:NASA Processors? (Score:4, Informative)

      by ChrisMaple (607946) on Monday January 26 2009, @02:14PM (#26610779)

      Deep space may be cold, but vacuum is a superb insulator. The chips can't be pushed hard without extensive and expensive heat sinks. Considerations on deep space probes are reliability and low power consumption, and there isn't a lot of need for speed. Reliability, radiation hardness, and low power consumption all have requirements that oppose speed.

      Furthermore, since space probes take a long time to develop and use only very well established technology, they are using nearly-obsolete semiconductors by the time they're launched. They're really old when they get where they're going. It's not fast stuff by today's standards.

    • Re:Okay.. (Score:4, Insightful)

      by cide1 (126814) <`moc.liamg' `ta' `trebrehguod'> on Monday January 26 2009, @02:23PM (#26610929) Homepage
      I generally don't let these types of things affect the CPU I use for work. I have found that in order for a system to be fast, all components much be equally matched. When the CPU is overclocked by a factor of 2, and the memory is not, the amount of time spent waiting on memory will increase significantly. If a designer knew the chip would be run at the higher speed, more cache would generally be included to make up for the disparity between CPU speed and memory speed. A good rule for buying new systems is to upgrade in two halves. I generally buy motherboard, RAM, CPU, and power supply at the same time for compatibility reasons. A year or two later, I will update my storage and video card. I buy a motherboard that supports the fastest memory made, I buy a lot of memory, and I buy a CPU that is at a point on the price to performance curve where spending more doesn't yield much more performance. In a year or two when software starts to actually use this capacity, Ill upgrade storage and video for a bit of a boast. Unfortunately, faster hard drives only make a bit of difference.
      • Re:Okay.. (Score:4, Insightful)

        by shaitand (626655) on Monday January 26 2009, @06:31PM (#26614997) Homepage Journal

        'Unfortunately, faster hard drives only make a bit of difference.'

        I suppose a 'faster' hard drive doesn't make a big difference compared with a 'fast' hard drive. But a fast hard drive compared with a slow hard drive makes a HUGE real world performance difference. Clunky and slow drives are the primary reason that laptops are so doggedly slow compared to desktops.

        Of course 'speed' is defined by rpm's in this case, not throughput.