Slashdot Mirror


Is Overclocking Over?

MrSeb writes "Earlier this week, an ExtremeTech writer received a press release from a Romanian overclocking team that smashed a few overclocking records, including pushing Kingston's HyperX DDR3 memory to an incredible 3600MHz (at CL10). The Lab501 team did this, and their other record breakers, with the aid of liquid nitrogen which cooled the RAM down to a frosty -196C. That certainly qualifies as extreme, but is it news? Ten years ago, overclocking memory involved a certain amount of investigation, research, and risk, but in these days of super-fast RAM and manufacturer's warranties it seems a less intoxicating prospect. As it becomes increasingly difficult to justify what a person should overclock for, has the enthusiast passion for overclocking cooled off?"

405 comments

  1. First post! by Anonymous Coward · · Score: 5, Funny

    Why? Because I've overc locked, so I'm faster than y'all!

  2. No by iB1 · · Score: 5, Insightful

    Overclock your smartphone or tablet instead

    1. Re:No by robthebloke · · Score: 5, Insightful

      which kinda defeats the point of the industries drive towards more efficient devices with longer battery life. I overclocked my netbook once. Most pointless thing I've ever done. It's now underclocked to eek out a little more battery life.....

    2. Re:No by Anonymous Coward · · Score: 5, Interesting

      I overclock my nook color from 800 mhz to 1200 mhz. I overcock my phone from 1 ghz to 1.4 ghz. My phone CPU's voltage doesn't change one bit and my Nook Color's CPU voltage is mildly higher. The CPU is far and away one of the least power consuming components of these devices -- the NC's screen uses around 1W and the cpu about 35mW. Unless you're overclocking the LCD, it the change in battery life is infinitesmal.

    3. Re:No by Anonymous Coward · · Score: 1

      If you overclock without raising voltages, or even under volt while overclocking battery life increases.

    4. Re:No by repvik · · Score: 3, Interesting

      You'll save a lot more battery if you undervoltage your CPU. Eg. My Galaxy Nexus by default runs at 1350mV. I can run it perfectly fine on 1200mV, even overclocked to 1,4GHz. By my (possibly completely wrong) logic, the faster the CPU runs, the shorter time it spends in higher voltage states. Thus, overclocking and keeping the same voltage (or even undervoltage) actually saves energy.

      (Of course, underclocking also achieves the same since the voltage is lowered automatically. But then I've got a slower device rather than a faster device, while using more or less the same amount of juice.)

    5. Re:No by Anonymous Coward · · Score: 0

      I overclocked my phone from 1GHz to 1.2GHz and simultaneously undervolted it. I get better performance and battery life now.

    6. Re:No by DeathToBill · · Score: 5, Informative

      Your logic is wrong.

      Every time a FET switches, it requires a certain number of electrons to move to or from the gate to create an electric field in the substrate to open or close a conducting pathway. This is a current flowing through a reistance and it dissipates power as heat. Assuming that the leakage current on the gate is very small compared to the switching current, the energy required to switch the FET (call it Es) is constant regardless of the clock speed. So the power dissipated by each FET (call it Pf) is:

      Pf = Es x fc

      where fc is the clock frequency in Hertz.

      Why do you suppose that frequency scaling is an effective way of saving power?

      --
      Slashdot - News for Nerds, Stuff that Matters, in ISO-8859-1 Has just realised that beta makes this signature redundant
    7. Re:No by somersault · · Score: 1

      Why?!

      What are you doing with your phone where say 10% will make much difference? Mid-range smartphones are already multi-core with hardware accelerated graphics and 512MB RAM or more. They're happily playing GTA3 now. Wait another couple of years and they'll probably be playing GTA IV. Graphics rendering is massively parallel and so easy to improve just by packing in more transistors. Better to just wait for the performance to double a few times rather than try to get tiny performance gains with exponential overheating and battery drain problems.

      --
      which is totally what she said
    8. Re:No by StripedCow · · Score: 5, Funny

      That's why you should have overclocked your battery too.

      --
      If Pandora's box is destined to be opened, *I* want to be the one to open it.
    9. Re:No by Anonymous Coward · · Score: 5, Funny

      I overcock my

      And that little slip right there says everything about the reasons for overclocking.

    10. Re:No by Anonymous Coward · · Score: 0

      Overclocking the previous generation of single core phones really extends the useful life of the device. There are plenty of original Galaxy S phones which are still rather nice devices that have a single core CPU. A moderate overclock actually brings these devices pretty close to par with the newer phones. I know several of my friends are still jealous of how slick my heavily modded Epic running CM7 is compared to their shitty locked Evo 3Ds.

    11. Re:No by jibjibjib · · Score: 2

      http://www.lesswatts.org/projects/applications-power-management/race-to-idle.php suggests that it's better to run faster for a short time than to run slowly.

    12. Re:No by Calydor · · Score: 1

      Diminishing returns, basically.

      If you for instance allowed it to slowly render that huge page you're looking at, working in the background while you were reading what was already rendered, you wouldn't have a lot of wasted power/time while just reading.

      It's like fuel economy in a car - the car has a 'best speed' for the amount of miles it'll go on a gallon. Same thing is true for CPUs.

      --
      -=This sig has nothing to do with my comment. Move along now=-
    13. Re:No by Anonymous Coward · · Score: 0

      Too bad we can't "overclock" the battery(ies) to last longer.

    14. Re:No by Anonymous Coward · · Score: 3, Interesting

      Noticed the typo just a litttttttle too late, heh.

      Actually the Nook is one of the best devices to overclock that I've ever come across -- it's really quite slow running CM7 or CM9 and the extra CPU speed helps immensely. CM9 is unusable without the extra boost, and the overclock takes CM7 from somewhat laggy to silky smooth in most operations.

      On my Epic, the overclock isn't super useful now that I'm running CM7 on it too. The normal stock Samsung build of Android has slight amounts of lag without the mild overclock, but CyanogenMod doesn't.

    15. Re:No by zAPPzAPP · · Score: 1

      Of course it does, because there is another part of power consumption for electronic devices, that does not change with the frequency.
      But this assumes you are going 100% cpu load over the whole time in both cases.
      Not very likely.

    16. Re:No by solidraven · · Score: 4, Informative

      That's total bullshit. Power consumption of a CMOS can somewhat be seen as this (incomplete formula to leave out device specific constants): Power consumption ~ Frequency * Capacitance * VCC. In other words, if you increase the frequency and VCC stays constant you're just going to burn more power for an almost irrelevant increase in performance.
      In fact I've read a lot more bullshit here but to address one more thing in specific: Most modern devices are made using CMOS technology or at the very least using FETs. A common misconception I see here is people assuming that CMOS devices use power while in a stable state. The fact is, they don't if they're well designed; The power being used is to charge the gates of the FETs. Once they're charged the only power use is to compensate for leakage currents from the FET gates to other parts of the substrate. If you don't believe me, build a small circuit using FETs (think something like a bunch of flip-flops), switch it a bit at a fairly high frequency. Stop the clock. Put a capacitor over your power supply. Then disconnect the circuit so its powered from the capacitor. It'll keep its state for weeks most likely.

    17. Re:No by pinkeen · · Score: 1

      GTA IV doesn't run smoothly on some recent PC hardware (and I'm not talking everything maxed out). You'll have to wait a long time til you can play it on a phone, cause it requires a lot of raw CPU power, not GPU, so very bad example here.

    18. Re:No by SuricouRaven · · Score: 1

      I would if I could, but I'm not that good at tablet hacking. Currently it can almost, but not quite, handle playing episodes of Friendship is Magic. I want to be able to watch those while on train journeys, but in high-motion scenes it struggles no matter what player I use - I suspect because the embedded h264 acceleration isn't being used. If I could get just 10% more processing performance, it should be able to manage.

    19. Re:No by drinkypoo · · Score: 1

      Some processors benefit from undervolting and some don't, and I have no idea what the difference is, but there must be one, because it works better for some processors than for others. It is said that undervolting most big and powerful processors makes very little difference in power consumption or heat dissipation, and only switching the transistors less (typically through clock reduction, but intel will switch off whole cores now, and IIRC AMD can shut off groups of cores if you have 6 or more of them) actually causes less heat production.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    20. Re:No by drinkypoo · · Score: 1

      When we get quad-core, 2+ GHz phones, it should be easy enough to do a scaled-back version of GTA IV. A few less pedestrians spawned slightly less far away, less bits and pieces flying off of things, it can probably be done.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    21. Re:No by Anonymous Coward · · Score: 5, Informative

      I'm sorry that you have barely passed EE 101 a week ago. The information you have is very inaccurate and quite outdated.

      In modern CMOS geometries, a large amount of power is wasted on leakage. That means that while the dynamic power scales linearly with frequency (at a constant voltage), the static power (leakage) does not.

      However, if you *can* overclock significantly at a constant voltage, there probably is power headroom the manufacturer did not use properly, or expected the devices to be unreliable with reduced voltage at the original frequency. Dynamic voltage scaling is not new.

    22. Re:No by justsayin · · Score: 5, Funny

      Maybe his l key is sticky? I wont go into any valid reasons for sticky keys,...

    23. Re:No by Anonymous Coward · · Score: 4, Informative

      There would be a massive change in operating power going from 800 mhz to 1200 mhz, even with the same voltage, because of gate capacitance.

    24. Re:No by gmarsh · · Score: 5, Informative

      Actually, voltage matters substantially.

      The gate of a FET is effectively a capacitor. Even with the FET in the on state, if you keep increasing the gate voltage it'll still keep taking electrons. And like a capacitor, energy stored in a FET gate = 1/2*C*V^2. You also have source/drain and gate/drain (miller) capacitance - source/drain has to be discharged (another 1/2CV^2 loss) and the miller capacitance has to be discharged and then charged at the opposite polarity (a CV^2 loss).

      Overall, neglecting leakage current, power loss is proportional to frequency, but it's also proportional to voltage squared.

      Power loss is also proportional to transistor count, which is why ARM is such a low power processor.

    25. Re:No by thisnamestoolong · · Score: 3, Funny

      That's why you need to be careful when overcocking...

      --
      To the haters: You can't win. If you mod me down, I shall become more powerful than you could possibly imagine
    26. Re:No by Bengie · · Score: 1

      "Some processors benefit from undervolting and some don't, and I have no idea what the difference is"

      Power consumption scales with the square of the voltage
      Power consumption scales linearly with frequency

      I under-voltaged my ATI6950 gpu. When I manually set my fan speed to 50% and stress tested my GPU with protein folding, it use to reach ~70c at the stock 1.1v. Now I have it under-voltaged to 1.085v and it's peaking about ~64c.

    27. Re:No by Anonymous Coward · · Score: 0

      Battery life is not a function of CPU power consumption alone. The dynamic power consumption of the CPU scales about linearly with the clock frequency, and so does the processing power. The less time you need the device to be active to do a particular task, the less power is consumed because the other components (the display in particular) are used less and therefore consume less power.

    28. Re:No by TheRaven64 · · Score: 2

      Add to that, a lot of work on mobile devices is bursty and the operating systems tend to be very aggressive about sleep states. If you make your CPU faster, it takes less time to do a small burst of work (e.g. rendering a web page) and then goes back to sleep. The faster processor - drawing more power when under load - can therefore use less power overall.

      --
      I am TheRaven on Soylent News
    29. Re:No by Nursie · · Score: 2

      N900 - overclocking involves lowering the voltages and increasing the maximum burst speed, theory being that if you get the work done faster you can go to sleep sooner and save power.

      I'm not sure it really works, but it does make the UI more responsive.

    30. Re:No by Anonymous Coward · · Score: 0

      People overclock mobile devices to give them pore power to play games with better frame rates, or watch media the device struggles with. Just because you are a miserable git and don't bother doesn't mean no one else should.

    31. Re:No by dgatwood · · Score: 2

      If it isn't going near 100% for a significant percentage of that time, the CPU should have been scaled back to a lower clock speed and, likely, a lower core voltage, so I would argue that not only is it likely, it should be nearly guaranteed. Am I missing something?

      Yes, I know there are sometimes performance reasons to leave the CPU going at a higher speed for short periods of time just in case it is needed for something else high-power shortly thereafter, which makes the relationship between clock speed and battery life somewhat muddy. That said, the assumption that bumping up the maximum clock speed so that you can scale back to a slower clock speed and a lower voltage sooner will reduce power seems like a pretty good first approximation, assuming that running at a faster speed doesn't require increasing the core voltage above what would have been required at the slower maximum clock speed, and assuming that you don't introduce instability in the process.

      --

      Check out my sci-fi/humor trilogy at PatriotsBooks.

    32. Re:No by JasterBobaMereel · · Score: 2

      My first computer cost £100-200
      My second computer cost £100-200
      My third computer cost £100-200
      My fourth computer cost £100-200
      My fifth computer cost £100-200
      My Smartphone cost £100-200

      All ran all the apps I wanted to when I bought it but were too slow for new ones ...

      I spent most of my time running apps that did much the same things on all of them, (web browsing, email, programming)

      Overclocking only extends the useful life of a computer, by reducing it's lifespan ...

      --
      Puteulanus fenestra mortis
    33. Re:No by dgatwood · · Score: 1

      Yes, but that's also potentially a case of premature optimization. What if the user jumps directly from page 1 to page 50? Now your device has just wasted power processing page 2 and page 3 while it could have been sitting in a completely idle state.

      Don't get me wrong—I'm not saying it's always wrong to do what you're suggesting, but it's just like read-ahead in hard drives and other similar problems in that it helps to have a pretty good sense of what the user is likely to do next; if you don't have a pretty good sense of what the user is about to do, your optimization is probably making power consumption worse, not better, on the average. And, as always, the hardest thing that you can do in computer science is predict the future.

      --

      Check out my sci-fi/humor trilogy at PatriotsBooks.

    34. Re:No by TheFakeMcCoy · · Score: 1

      Just turn off sticky keys... i always hated that feature

    35. Re:No by Anonymous Coward · · Score: 5, Funny

      My ex-wife overcocked and ended up divorced because of it.

    36. Re:No by Anonymous Coward · · Score: 0

      You do know that ideal capacitors are lossless components, right? Neglecting internal resistance, the energy that charges the capacitor is stored in the electric field, and then released when the cap is discharged. Total power loss in the ideal case = 0.

      With internal resistance, the power loss is V^2/R, where V is the voltage across the internal resistance. Nothing to do with C.

    37. Re:No by thegarbz · · Score: 1

      That's a very black and white view of the world. Just because some people sell something doesn't mean its a popular choice to buy.

      Thinking back 10 years ago overclocked PCs were everywhere. Friends would share stories about how to overclock, hell I've even seen some guys in an office overclock their work machines to squeeze some extra umph out of it. But these days the returns of doings so are greatly diminishing.

      Why would I buy an overclocked videocard or CPU? Hell why would I even bother buying a high-end videocard or CPU? I have a 2 year old video card and a 3 year old processor and most modern games run fine at 1920x1200 with settings generally quite high. Where's my incentive? 10 years ago you either upgraded, overclocked, or likely couldn't play a game if your hardware was more than a year old. The console has stagnated the graphics capabilities of games to an extent where this doesn't matter. Overclocking to run windows maybe? The HDDs have become such massive bottlenecks that you see no real improvements in office productivity from such things either.

      High end gaming rigs used to be a requirement, now they are for enthusiasts and those with too much money to burn.

    38. Re:No by Anonymous Coward · · Score: 0

      you mean eke

    39. Re:No by MXPS · · Score: 0

      Good call, because my Evo 3D definitely isn't unlocked, overclocked, and running CM7 (among other ROMs).

    40. Re:No by Anonymous Coward · · Score: 0

      The only people that I know who use it are hunt and peckers which is odd because it hardly is relevant to them.

    41. Re:No by localman57 · · Score: 1

      Sigh. If only the bad got slashdotted with the good...

    42. Re:No by localman57 · · Score: 1

      Except that you're charging the capacitor from your voltage source, then discharging the charge to ground. It's like saying a bucket is lossless in terms of water, except that you're filling up from a faucet, and pouring it down the drain.

    43. Re:No by ladoga · · Score: 1

      which kinda defeats the point of the industries drive towards more efficient devices with longer battery life. I overclocked my netbook once. Most pointless thing I've ever done. It's now underclocked to eek out a little more battery life.....

      There are lots of cases where overclocking actually increases the battery life. If CPU does the calculation faster (considering that voltage stays the same) it spends more time idle and thus saves the battery.

    44. Re:No by maxwellmath · · Score: 1

      Typically, increasing the core voltage will speed up the processor. Decreasing the core voltage will save power but will slow it down. This is because a higher voltage is able to more quickly charge up the capacitances associated with MOSFETs (there is a capacitance at the gate of every FET which slows down the operation of the transistor. There are also diffusion capacitances on the source/drain side of the FET which also slow things down). The time constant associated with a FET is T = RC where C is the capacitance and R is the resistance associated with the transistor when it is on (and possibly the connections between transistors as well). A simplified model of the charging process is V = Vd * (1 - e^(-t/T)) where t is time, and Vd is your core voltage. Since a transistor must reach some threshold voltage in order to turn on (call it Vt, typically around 0.6V for typical CMOS processes), you can see that if you increase Vd then V will reach the transition voltage more quickly.

    45. Re:No by Anonymous Coward · · Score: 0

      Your logic is wrong.

      Every time a FET switches, it requires a certain number of electrons to move to or from the gate to create an electric field in the substrate to open or close a conducting pathway. This is a current flowing through a reistance and it dissipates power as heat. Assuming that the leakage current on the gate is very small compared to the switching current, the energy required to switch the FET (call it Es) is constant regardless of the clock speed. So the power dissipated by each FET (call it Pf) is:

      Pf = Es x fc

      where fc is the clock frequency in Hertz.

      Why do you suppose that frequency scaling is an effective way of saving power?

      No, Your logic is wrong, fc = Force Close

    46. Re:No by Zuriel · · Score: 1

      If it has a hardware decoder, check that the video matches the hardware decoder's capabilities. Some can't handle h264 main or high profile and only baseline profile works smoothly. http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Profiles

      Software decoders, you only need to worry about CPU performance. Hardware decoders are fussy.

    47. Re:No by Anonymous Coward · · Score: 0

      Define efficient.

      To me, efficient means raw performance. I couldn't care less about having more battery life.

    48. Re:No by blackraven14250 · · Score: 2

      If you overclock the CPU and don't need to have the screen on for as long as a result (i.e. waiting for an app to load takes 1 second instead of 1.5), you're clearly ahead of the game in power consumption because the screen is using far more energy than the CPU. This adds up over time too, so silky smooth operation usually equates to lower battery consumption overall.

    49. Re:No by jandrese · · Score: 0

      Near the end of its life I overclocked my original model iPhone (I only replaced it a couple of weeks ago) to extend its life slightly with various game apps. The battery life took a hit, but since I tossed it on the charger every night anyway it wasn't unmanageable. It basically became an iPhone 3G once you overclocked it. Amusingly, it wasn't actually an overclock in the traditional sense, Apple underclocked the CPU to save battery life and all I did was revert the chip to its native clock rate.

      Other than that, the last time I messed with overclocking was on my old Pentium 75 (it ran smoothly at 100Mhz) because it was dogging on some games I wanted to play. Overclocking for the sake over overclocking never seemed worth it to me. It's mostly just a free way to wring a few more months/years out of a system before upgrading.

      --

      I read the internet for the articles.
    50. Re:No by Anonymous Coward · · Score: 0

      Oh, look we have here ........ This argument.

      "Voltage matters" .... "No it doesn't. Heat dissipation is the concern" .....

      You two, get a room already!

    51. Re:No by cod3r_ · · Score: 0

      i like you

    52. Re:No by Anonymous Coward · · Score: 1

      My ex went off halfcocked.

    53. Re:No by solidraven · · Score: 4, Informative

      Sadly you're wrong there. We spend hours calculating the optimal frequency of systems in a lot of cases and then spend the rest of the day designing a PLL to work on that frequency. Believe me when I say you're not going to increase energy efficiency by overclocking the base setup without extensively modifying the hardware.

    54. Re:No by Anonymous Coward · · Score: 0

      Did you seriously place hunt and peckers 2 degrees of separation from overcocking? Bravo!

    55. Re:No by Runaway1956 · · Score: 1

      People who care about nothing except raw power never have, and never will, understand "efficiency". You'll have noticed by now that no one is willing to waste time defining "efficient" for you.

      --
      "Windows is like the faint smell of piss in a subway: it's there, and there's nothing you can do about it." - Charlie Br
    56. Re:No by Guspaz · · Score: 1

      Not to mention that the Nook's processor uses a hell of a lot more than 35 milliwatts when doing work. It might survive on that when idle, but not active...

    57. Re:No by deblau · · Score: 1

      I don't see anything particularly nefarious about pressing the Shift key five times in a row...

      --
      This post expresses my opinion, not that of my employer. And yes, IAAL.
    58. Re:No by solidraven · · Score: 1

      It doesn't work that way sadly.

    59. Re:No by BancBoy · · Score: 0

      Checking the usernames on this subthread, it would appear there is a conspiracy of Ravens in here!

      solidraven
      TheRaven64
      blackraven14250

      --
      [UID-HeinzIntel]
    60. Re:No by Anonymous Coward · · Score: 0

      which kinda defeats the point of the industries drive towards more efficient devices with longer battery life.

      Not really. With on-demand clock speed changes its an entirely different game. Add to that the ability to overclock, underclock, and under volt, many people get their cake and eat it too. These days, over clocking only kicks in when you have a CPU intensive task for longer than non-trivial durations. Which basically means, faster, more responsive UI but battery saved when on a phone call or when the device is multitasking in the background on an otherwise idle device.

    61. Re:No by izomiac · · Score: 1

      Underclocking at the same voltage worsens your battery life. The concept is called "race to idle". Basically, if the processor takes 2 V under load, it's better to finish in 1 second at 1000 MHz than 1.7 seconds at 600 MHz. Since modern processors use nearly zero power in idle, the former is close to 60-70% more efficient.

    62. Re:No by Anonymous Coward · · Score: 0

      No man, look at the datasheet. Idle uses about 15 mW.

      ARM devices are awesome.

    63. Re:No by ChrisMaple · · Score: 1

      The R in question is the sum of the wiring resistance and the resistance of the transistors in conduction. Since switching time is inversely proportional to resistance, resistance drops out of the equation. (The only way around this is inductors, which are not practical.) Internal resistance cannot be neglected, because it is essential to the actual loss mechanism.

      --
      Contribute to civilization: ari.aynrand.org/donate
    64. Re:No by kimvette · · Score: 3, Funny

      Is your name Lorena?

      --
      The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
    65. Re:No by Mike+Buddha · · Score: 1

      Most phones and tablets are underclocked out of the box for that very reason. It's a matter of how much processing power is necessary for the device to appear to be responsive. There are tricks that Apple uses to make iOS appear more responsive than it is. I wish Android would catch up in that regard.

      --
      by Mike Buddha -- Someday the mountain might get him, but the law never will.
    66. Re:No by Anonymous Coward · · Score: 0

      I hope she wasn't getting dual-cored

    67. Re:No by __aasehi2499 · · Score: 2

      Um, gamers need to turn off sticky keys, and have learned to do so after a few times getting their application focus changed to a message saying "You have just turned on sticky keys by pressing * key 5 times, would you like to learn more?" --bite me stupid worthless function, you just cost me the game.

    68. Re:No by Guspaz · · Score: 1

      Well, a standard high-performance dual-core Cortex A9 (clockspeed of up to 2GHz) has a TDP of 1.9W, while the low-power version (clockspeed of 800MHz) has a TDP of 0.5W. In either case, that's way more than the 0.035 watts being given as the peak usage.

    69. Re:No by snowgirl · · Score: 0

      You do know that ideal capacitors are lossless components, right?

      Yes, and frictionless bearings, and surfaces also do not slow down objects. Guess which of these three items exist in the real world: hint, it's zero.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    70. Re:No by bemymonkey · · Score: 1

      Can you elaborate? Which optimal frequency? The idle frequency (lowest multiplier)? The max. frequency (full tilt)?

    71. Re:No by DMUTPeregrine · · Score: 1

      I have an Android phone (Samsung Vibrant) and I both over and under clock it. SetCPU profiles underclock when battery gets low/temp gets high, overclock a bit when under load, and a lot when plugged in and under load.

      --
      Not a sentence!
    72. Re:No by DemonGenius · · Score: 1

      I overclocked my N900 back when I was using it and the speed increase was noticeable. The apps ran faster and the screen was more responsive to touch. However, the increased voltages to the CPU totally killed my reception and made it nearly impossible to use it as an actual phone. I would imagine other smartphones would have similar issues with the components so close together. Tablets, on the other hand, would not have this issue, so in this case, overclock to your heart's content!

    73. Re:No by Anonymous Coward · · Score: 0

      Power = Ceff * V^2 * F.

      So the power of the processor goes up 50% as you increase F by 50% (800 MHz to 1200 MHz)

      Before Overclocking:
      1W + 35mW = 1.035W

      After Overclocking:
      1W + 1.5*35mW = 1.052W

      So with overclocking, power increases 1.6%.

      What are we assuming here:

      1. the only other source of power draw that doesn't change with cpu overclocking is the screen. This ignores other obvious users of power like wifi, i/o circuits, etc.

      2. The only source of power draw that does change with cpu overclocking is the cpu. If memory or other component's frequency must also increase, then these component's power will increase as well.

      3. V doesn't change as F changes. If overclocking requires voltage to increase, the power used by overclocking will increase as the square of the voltage increase.

      Accounting for (1) increases the strength of parent poster's argument. Accounting for (2,3) decreases the strength of parent poster's argument.

    74. Re:No by Forty+Two+Tenfold · · Score: 1

      Yeah, it's a real turn-off.

      --
      Upward mobility is a slippery slope - the higher you climb the more you show your ass.
    75. Re:No by mjwalshe · · Score: 1

      exactly that is why there is a market for overclocked and water cooled super record 2 MB bundles for the finance industry.

    76. Re:No by DrEldarion · · Score: 1

      He probably pressed the shift key too many times in a row.

    77. Re:No by solidraven · · Score: 1

      Nevermore...

    78. Re:No by Anonymous Coward · · Score: 0

      And myopic people who believe that the world revolves around them never have and never will understand that "efficiency" is completely subjective. What I noticed is that nobody defined "efficient" because they understood this and agreed. Efficiency is getting my work done instead of twiddling my thumbs, waiting for my computer to catch up. Then, maybe your job is pretty worthless and you don't have such responsibility.

    79. Re:No by blackraven14250 · · Score: 1

      How not? The raw energy needed to burn through those calculations remains the same, while the time the screen is on decreases in the use case of picking up the phone to find a small piece of information. If you have something like JuiceDefender turning off data connections when the screen turns off, you're going to have a similar effect from using 4G vs 3G - quicker load times, and less overall energy used in the time between sleep states.

    80. Re:No by solidraven · · Score: 1

      In complex systems as these the power usage is often a complex formula of several components. We can often model a relative simple formula for these "small" embedded systems anyway (think portable devices). Based on this the ideal operating frequency is determined to minimize power use while keeping the performance acceptable. This isn't related to multipliers or maximum frequencies most of the time. We design the system to operate near these optimal frequencies.

    81. Re:No by solidraven · · Score: 1

      Not nearly as much as you'd think. Switching is still the major source of power usage. If what you are saying was true we could just as well switch to bipolar again.

    82. Re:No by Anonymous Coward · · Score: 0

      That's total bullshit. Power consumption of a CMOS can somewhat be seen as this (incomplete formula to leave out device specific constants): Power consumption ~ Frequency * Capacitance * VCC.

      The voltage term is squared. And that's only the dynamic power component, you didn't calculate static power at all. Despite your claims, static power is a major concern for a wide variety of devices.

      In other words, if you increase the frequency and VCC stays constant you're just going to burn more power for an almost irrelevant increase in performance.

      From this and other things in your posts I suspect you are a young engineer who suffers from just-graduated-and-still-thinks-he-knows-it-all disease. Not all systems are the same -- you might well be working on one where increasing frequency doesn't buy you anything, but that doesn't mean it's universally true.

      In fact I've read a lot more bullshit here but to address one more thing in specific: Most modern devices are made using CMOS technology or at the very least using FETs. A common misconception I see here is people assuming that CMOS devices use power while in a stable state.

      Well, that's a safe assumption, because they do. All CMOS devices need at least some steady state current. Sometimes that current is very small, other times it isn't.

      The fact is, they don't if they're well designed; The power being used is to charge the gates of the FETs. Once they're charged the only power use is to compensate for leakage currents from the FET gates to other parts of the substrate.

      "Well designed" has nothing to do with it. As with many things in engineering, there are tradeoffs. In this case, the tradeoff is that it's possible to make an ultra low leakage FET, and it's possible to make a FET capable of switching states extremely quickly, but you cannot build one device which is both fast and low leakage. So if you want to build high performance logic, like, say, a fast CPU, you end up using fast, leaky FETs in performance critical circuits, and slower less leaky FETs wherever you can get away with them, but probably all of them are significantly more leaky than FETs optimized for nothing but low leakage current. The resulting product will have a non-trivial amount of leakage current.

      If you don't believe me, build a small circuit using FETs (think something like a bunch of flip-flops), switch it a bit at a fairly high frequency. Stop the clock. Put a capacitor over your power supply. Then disconnect the circuit so its powered from the capacitor. It'll keep its state for weeks most likely.

      Data retention for weeks on a cap powering a handful of discrete 74HC series flops (let's say ~100 general purpose FETs in the whole system) does not in any way imply that you could do the same with millions of high performance logic FETs in a fast microprocessor. The operating principle of FETs is always the same, but that doesn't mean that all FETs are created equal. Nor can you ignore the magnitude of the systems you're comparing!

      The dude you replied to was not bullshitting, he was correct:

      If you overclock without raising voltages, or even under volt while overclocking battery life increases.

      P = Pstatic + Pdynamic

      If you raise the frequency without raising the voltage, the two components in this equation behave differently. Pdynamic should scale roughly linear with frequency. The effect on Pstatic is indirect -- extra power creates more heat, and leakage power scales with junction temperature. It's not a powerful effect until high temps, and cooling systems can keep temperature under control, so in many circumstances Pstatic scales sub-linear WRT frequency.

      Given that, if the performance gain of an OC is linear with the frequency delta, and the net power gain is less than linear, computational efficiency (perfo

    83. Re:No by Anonymous Coward · · Score: 0

      And myopic people who believe that the world revolves around them never have and never will understand that "efficiency" is completely subjective. What I noticed is that nobody defined "efficient" because they understood this and agreed. Efficiency is getting my work done instead of twiddling my thumbs, waiting for my computer to catch up. Then, maybe your job is pretty worthless and you don't have such responsibility.

      You're an ass. Plenty of people have important, responsible jobs which don't depend on the computer being super fast, but DO depend on it being available because it's their primary mobile communications device. If their Megadroid Razr HAL-9000 is ultra fast but burns through battery, they'll have to be tethered to AC power to get anything done, which is the opposite of efficiency for something that's supposed to enable you to move and stay in touch.

      What's more, you're doubly an ass because going super fast these days requires being efficient. Today, even high performance desktop CPUs are limited by power. Intel and AMD both experienced practical problems trying to push CPU thermal power above 140-150W or so (you start to need liquid cooling above that), and generally speaking ~100W or less is a much better idea. It keeps cooling system costs under control and failure rate down. Given this fixed power budget, a more efficient CPU is a faster CPU. So you are actually an efficiency fan, even though you didn't know it.

    84. Re:No by bemymonkey · · Score: 1

      So you're talking about the "operating frequency", i.e. 1 GHz on a 1st-gen Snapdragon? Because my Snapdragon spends about 5% of its time at 1GHz, and the rest clocked down to 250MHz...

  3. It's not dead, it's fun! by X-Power · · Score: 5, Informative

    For me, It's fun and I could care less what some dude did with liquid nitrogen.

    First computer, I just used Asus Overclock and felt I got more for my money.
    Second computer, I started fiddling with manual settings.
    Third computer I pushed it until I couldn't get rid of the heat with air cooling.
    Fourth and current computer, water cooled and running awesome (6 cores at 4.3 GHz).

    Each time I felt the progress, it's like leveling your character, but the character is you, and the game is real life!

    1. Re:It's not dead, it's fun! by jones_supa · · Score: 3, Interesting

      For me, It's fun and I could care less what some dude did with liquid nitrogen.

      About this post, it's hard to determine whether this should be "could care less" or the classic "couldn't care less". :)

      You could be interested about liquid nitrogen as you are an overclocker or, you're not as you don't want to go to such an advanced level it just being a fun hobby.

    2. Re:It's not dead, it's fun! by somersault · · Score: 2

      I always found that overclocking was very anti-climactic. It's noticeable if you have a really awful computer, but if your computer is already running okay then it makes no difference to add a little extra performance. It's like adding more RAM. There are less slowdowns, but technically nothign is actually speeding up.

      Also to me it still sounds like you're levelling something external, ie your computer. Levelling your knowledge very, very slightly too, but it's nothing compared to the levelling you'd feel if ate more healthily and did a bit of regular exercise (even just going for a decent walk a few times a week will make you feel much fitter). I wish that in games like Skyrim I could just control the character directly - get some exercise while still enjoying the game world. I think the ideal system would involve something like a bungee harness and a giant trackball under your feet. Maybe I should patent that? :P

      --
      which is totally what she said
    3. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 1, Informative

      You must be new here. Americans say "could care less" when they mean "couldn't care less". They spell "colour" without the "u", too, and say "sidewalk" instead of "pavement".

      I'm not aware of any dialect of English in which people say "could care less" to mean "I care about this somewhat", which is obviously what "could care less" actually means.

    4. Re:It's not dead, it's fun! by Kevin108 · · Score: 5, Insightful

      Morons with no actual understanding of the language say "could care less." It's just that there's a lot of them.

      --

      It's a perfect time for being wasted.
      A perfect time to watch the stars.
      - Burden Brothers, "Beautiful Night"
    5. Re:It's not dead, it's fun! by reve_etrange · · Score: 1, Flamebait

      Or, the sarcastic phrase became a common idiom.

      --
      .: Semper Absurda :.
    6. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      The only beneficial overclock I ever experienced was running my single-core Athlon64 at closer to 2.4GHz than its default 2.2GHz.

      It meant I could play the 3D version of Xevious in MAME smoothly versus not being able to play the 3D version of Xevious in MAME smoothly.

      I have a dual Opteron 2.6GHz system now. It only doesn't emulate a Wii or Android device at a decent speed.

    7. Re:It's not dead, it's fun! by VinylPusher · · Score: 5, Funny

      All I do these days if I'm feeling daring is activate the 'high performance' power profile in Windows.

    8. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 1, Funny

      >> there's a lot of them.

      there are a lot of them, not there is a lot of them

    9. Re:It's not dead, it's fun! by TheRaven64 · · Score: 4, Informative

      Overclocking my old Cyrix CPUs made a noticeable difference in a lot of games. These days? I just don't care. My laptop has a quad-core 2.2GHz i7. It is insanely fast. Even big compile jobs can run with -j4 and it's still responsive. My tablet has a 1.2GHz dual-core processor, and it's faster than the desktop that my mother uses - why would I bother overclocking it?

      --
      I am TheRaven on Soylent News
    10. Re:It's not dead, it's fun! by jones_supa · · Score: 1

      Also to me it still sounds like you're levelling something external, ie your computer. Levelling your knowledge very, very slightly too, but it's nothing compared to the levelling you'd feel if ate more healthily and did a bit of regular exercise (even just going for a decent walk a few times a week will make you feel much fitter).

      That, and along just tweaking some megahertz number (and cooling systems), having a wider electronics or programming hobby is a more rich way to level up someone's skills.

    11. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      A good possibility. How disappointing.

    12. Re:It's not dead, it's fun! by swalve · · Score: 1

      "Could care less" means that there exists a possible situation where one's level of caring would be even lower than now. Which means the phrase "could care less" as an indication of apathy is useless. Whether you had almost no concern at all, or you really loved something, you could care less for it. No matter how you inflect it, you are indicating that you hold the concept in some esteem.

      "Couldn't care less" is a valid indication of apathy, because there is no possible situation where your level of caring would be lower. It is the least amount of concern/interest possible.

    13. Re:It's not dead, it's fun! by swalve · · Score: 1

      The only time I've seen overclocking be worthwhile was back in the day of the slocket pentiums, where you could make a celeron 266 run at 400 mhz. It also helped if you could overclock enough to gain a whole multiplier, or if you could get your memory and PCI devices to go up a notch and lose a multiplier. Getting pc100 memory to run at pc133 made a huge difference.

    14. Re:It's not dead, it's fun! by Stickybombs · · Score: 1

      Pavement is the material itself, such as asphalt, concrete, etc. The word "pave" means to cover a surface. Sidewalk, road, driveway, and such refer to the use of the paved area. You don't typically call your house "bricks" or "wood" or whatever it may be constructed from, so why call a sidewalk "pavement".

    15. Re:It's not dead, it's fun! by somersault · · Score: 1

      Definitely. I go through phases though and I haven't done a whole lot of programming as a hobby since I was a teenager (though I now program as part of my job). I think about it occasionally, but haven't had any project ideas that really grab my attention.

      That idea for the human trackball is probably the most attractive project idea yet - it probably wouldn't even be that expensive or difficult to build a giant trackball and have a harness on a frame above it, and the device drivers wouldn't be that difficult to do. Get a yoga ball mounted with some bearings to let it run smoothly, and have a webcam or even just a normal laser mouse as the movement sensor. But I don't have space for setting up something like that in my flat, it would have to be over 2.5 metres high :/

      --
      which is totally what she said
    16. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Perhaps you should go walk on the pavement instead of the sidewalk.

    17. Re:It's not dead, it's fun! by oakgrove · · Score: 1

      The GP has a good point. Concede with dignity.

      --
      The soylentnews experiment has been a dismal failure.
    18. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Irregardless(*), alt.usage.english actually lists both as completely acceptable in their FAQ, probably because there's too many people who don't realize they are both equally acceptable and correct.

      http://alt-usage-english.org/excerpts/fxcouldc.html

      In fact, the so-called "incorrect" usage of "could care less" sort-of appears in literature dated prior to "couldn't care less": "You know nothing and you care less, as people say." (Mansfield Park (1815), Chapter 29). Knowing that most pedants base their preferences on historical usage, the defensible position would be to suggest "couldn't" is the wrong word.

      Concede that perhaps either you or whoever taught you english is closed-minded and probably also believes the only dictionary in the world is the OED.

      (*) Used on purpose so you'll look up the word and find that it really is in the dictionary (just not yours), and therefore a word. ;-)

    19. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Well, quite. But if we're going to be all patronisingly pedantic, perhaps we could learn to distinguish between apathy and indifference. Hmm?

    20. Re:It's not dead, it's fun! by Stickybombs · · Score: 1

      "You know nothing and you care less, as people say." is saying that you care less than nothing, which kind of implies that you couldn't care less. How does one 'care less than nothing'?

    21. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      I always found that overclocking was very anti-climactic. It's noticeable if you have a really awful computer, but if your computer is already running okay then it makes no difference to add a little extra performance. It's like adding more RAM. There are less slowdowns, but technically nothign is actually speeding up.

      I depends on the workloads. If you just use your PC for web browsing/e-mail/word processing, then yeah, not much point in overclocking. But If you're doing any 3D rendering, compiling, video editing/transcoding, or other CPU intensive tasks, you will get quite noticeable gains out of overclocking. I overclocked my i7 920 CPU from 2.67 GHz to 3.6 GHz (mild) - and believe me, it's noticeable.

    22. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      "Care less" and "could care less" have different meanings.

      If "I care less" then I care less than whatever I am comparing to, e.g. "I care less than you about your mother."

      "I could care less" implies that for whatever reason I care somewhat but there also exists the possibility that I could stop caring as much.

      "I couldn't care less" suggests that in no way whatsoever do I care leaving no possibility that I could care less.

      To my way of thinking "irregardless", with its two negatives, must mean something like "with regard to."

    23. Re:It's not dead, it's fun! by amRadioHed · · Score: 1

      Moron with no actual understanding of language in general complain about language not being logical.

      --
      We hope your rules and wisdom choke you / Now we are one in everlasting peace
    24. Re:It's not dead, it's fun! by zzsmirkzz · · Score: 0

      Morons with no actual understanding of the language say "could care less." It's just that there's a lot of them.

      You might want to be a little more careful in what you say about common expressions as I take it you really haven't put much thought into it. To say "you couldn't care less" about something means you care a certain amount (which is left unsaid) about it already and couldn't care any less about it as it is (apparently) important. However, saying you "could care less" about it means you care a certain amount (again left unsaid) about it already but could care even less about it because it is that insignificant to you. In one expression you've reached the bottom and in the other there is no bottom. Picking from the two, I'd say the one that could be cared less about is less important that one that couldn't.

    25. Re:It's not dead, it's fun! by YttriumOxide · · Score: 1

      Pavement is the material itself, such as asphalt, concrete, etc. The word "pave" means to cover a surface.
      Sidewalk, road, driveway, and such refer to the use of the paved area.
      You don't typically call your house "bricks" or "wood" or whatever it may be constructed from, so why call a sidewalk "pavement".

      Personally I use the word "footpath" for the path that I walk on beside the road. "Sidewalk" sounds foreign, but I don't consider it bad at all. "Pavement" I'd agree to me sounds like a reference to the material (that with which it is paved).

      For reference, New Zealand English is my native variant, but I've spent over half of my life in other countries, so I may have been influenced somewhat in various directions.

      --
      My book about LSD and Self-Discovery
      Also on facebook as: DroppingAcidDaleBewan
    26. Re:It's not dead, it's fun! by qualityassurancedept · · Score: 1

      There are of course prescriptivist and descriptivist approaches to grammar. What you did in your post is called "overcorrection": you take a perfectly comprehensible idiomatic phrase and claim that it is not grammatical for logical reasons and then add in the parts that would make it a proper phrase. In fact, from a descriptivist point of view, the grammar of English is: what native speakers actually say when they speak English... not what they SHOULD say if they spoke English correctly. In short, lots of people say "could care less" and because of that it is the idiom. It's not like the idiom should be "couldn't care less" because the addition of the "not" makes the idiom more logical. In other words, your standard of grammaticality starts with logic and you are claiming that the phrase should be made grammatical by the addition of the correct logical elements. This is only one way to approach grammar.

      --
      if your life is such a big joke then why should I care?
    27. Re:It's not dead, it's fun! by vrt3 · · Score: 4, Informative

      No, "I couldn't care less" means "it's not possible to care even less than I already do, even if I wanted to". It means I care the least possible amount. I have reached the bottom.

      "I could care less" (but I don't, meaning that I do care a certain amount) means that there is still a margin between the amount I care and the least possible amount of care. I haven't reached the bottom. Nothing is said about the size of that margin, so this statement really doesn't say anything.

      --
      This sig under construction. Please check back later.
    28. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Yeah, that's the "giant hashtable" approach to understanding language (or pretty much anything else for that matter). See a phrase you haven't seen before? Look up cases where someone else has used it and fuck logic.

    29. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0
      irregardless/irigärdlis/
      Adjective: Regardless.
      Nartural language does not need to follow the rules of logic. Stop shoehorning the world around you to your limited intellectual capacity.

      To my way of thinking "we don't take no prisoners", with its two negatives, must mean something like "we take prisoners."

      Think less and look it up?

    30. Re:It's not dead, it's fun! by zzsmirkzz · · Score: 0

      No, "I couldn't care less" means "it's not possible to care even less than I already do, even if I wanted to". It means I care the least possible amount. I have reached the bottom

      That is one possible, valid interpretation, but not the only one. It is also possible to interpret it to mean that it is not possible for me to care less than I do now because it is significant and I must care the amount I currently do. The "it's not possible to care less than I already do" is a given with the statement. However, the reason why it's not possible is assumed, and both are grammatically valid assumptions. When said in speech, the tone of voice and context fill in this assumption for you.

      "I could care less" (but I don't, meaning that I do care a certain amount) means that there is still a margin between the amount I care and the least possible amount of care. I haven't reached the bottom. Nothing is said about the size of that margin, so this statement really doesn't say anything.

      Again you're filling in reasons that aren't provided. "I could care less" means that it is possible for you to care less than you do now. It may even be likely for you to do so. Saying "I could care less" means that it's not very important, it's possible for it to be cared less for. It is also usually said as a response to someone inquiring how important something is, and in that context it is pretty obvious what the person means.

      The problem with semantic arguments around common expressions is that they are a moot point. It just so happens that both of these sayings mean the same thing even though they are grammatically opposites. Mostly because they sound similar and people interchange them without thinking about what they are actually saying. It is this familiarity with the phrase that causes people to make the assumptions (like you made) about what is meant and the correct message is conveyed.

    31. Re:It's not dead, it's fun! by gl4ss · · Score: 1

      it depends on how long compile, rendering or whatever runs you run with it.

      1 minute off from 5 minutes might matter.

      1 hour off 5 hours matters much more.

      --
      world was created 5 seconds before this post as it is.
    32. Re:It's not dead, it's fun! by j33px0r · · Score: 1

      Morons with no actual understanding of the language say "could care less."

      Morons with an actual understanding of language would describe the term "could care less" as a slang phrase. It means exactly what the original poster implied. Changing the phrase to "couldn't care less" does not change the original poster's intent, something that you were very well aware of before you started your rant to make yourself feel better about your language skills. This is all beside the point since the English language and subcultural phrases change from region to region.

       

      It's just that there's a lot of them.

      Are you a member of the APA, MLA, or similar group because each has its own set of guideline (with multiple versions) of what is an acceptable practice. Make sure that you have the right qualifications before you sign up to be a member of the language police. For example, what does "it's" refer to? If you want to be technical, your sentence reads "It is just that there is a lot of them." Since there really is no "it," your sentence must be read, "there is justice in that there is a lot of them."

    33. Re:It's not dead, it's fun! by AngryDeuce · · Score: 1

      Yeah, really. I hear people say "for all intensive purposes" far more often than the correct, "for all intents and purposes".

    34. Re:It's not dead, it's fun! by Rogerborg · · Score: 1

      No, I'm still pretty sure it's all down to morons.

      --
      If you were blocking sigs, you wouldn't have to read this.
    35. Re:It's not dead, it's fun! by Keith_Beef · · Score: 1

      Pavement is the material itself, such as asphalt, concrete, etc. The word "pave" means to cover a surface. Sidewalk, road, driveway, and such refer to the use of the paved area. You don't typically call your house "bricks" or "wood" or whatever it may be constructed from, so why call a sidewalk "pavement".

      In British English, the terms "pavement" and "paved" also tend to have a strong sense of a surface being covered with "paving stones" as opposed to being covered by a continuously poured material such as tarmac, asphalt or concrete, a process known as "metalling". So the pavements are paved, and the roadway is metalled.

      We adopted metalling for the roadway long before using it on the pavements, and it is still very common to find pavements that are flagged (i.e., covered with big flagstones) .

      But we also use "footpath" and sometimes "causeway" (pronounced kz.i in some areas) for the bit reserved for pedestrians.

      K.

    36. Re:It's not dead, it's fun! by billcopc · · Score: 1

      Do I care what some Romanian attention whore benched with LN2 ? No. I care about real-world, sustainable results in a permanent installation. I can't do my work if I have to pause every 10 minutes to pour more nitrogen into an open-air cooling column. This would be an example of masturbatory overclocking, kind of like blowing $50k pimping out an old Shelby, then carting it out to the Sunday car show in a trailer "look at me, I did this, aren't I cool ?". There's a big difference between TFA and normal overclocking like you do. Even your water cooled rig, you can still use it as a regular PC, play your games, transcode video and render 3D scenes of a yet another coffee table with plastic fruit. Buddy boy in TFA, on the other hand, will be lucky if the board is still usable after his little experiment. The funny thing about thermal stress is it's great at taking all the impurities and turning them into tiny little cracks, which then cause the board to fail prematurely, but hey at least he got to show off his ePeen, right ?

      Fact is, my PC is now a work machine, not a toy anymore. I just don't relish the idea of losing a week to overclocking endeavours anymore, because I have more pressing shit to do with my PC, like getting paid. Actually my previous PC had one of those sealed water rigs from CoolIT, and sure enough, it overclocked quite well with minimal effort. My current rig, well it's running at stock speeds and my board doesn't even have OC settings in the CMOS setup (it's a server board). I still sell overclocked PCs to those who want them, that's been a decent chunk of my business for nearly a decade, but my personal builds tend to favour noise reduction and energy savings over Ghz. If I need more compute power, I'd rather assemble an $800 node than spend a few days overclocking the ones I already have. Cluster > clock speed.

      --
      -Billco, Fnarg.com
    37. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Go stereotype someone else. Like your sisters. They couldn't care less. ;)

    38. Re:It's not dead, it's fun! by LordLimecat · · Score: 4, Funny

      Can we quantify "care" with units? Im still having trouble grasping this.

    39. Re:It's not dead, it's fun! by ChrisMaple · · Score: 1

      One problem with ambiguous language, which you seem to support, is legal. In principle, a person could be hanged for answering "yes" to the question "Didn't you murder that woman?", which literally means "Did you not murder that woman?" instead of the colloquial "Did you murder that woman?" Or maybe that person could be hanged for answering "no". Rest assured that the prosecution will strive to distort your answer.

      --
      Contribute to civilization: ari.aynrand.org/donate
    40. Re:It's not dead, it's fun! by snowgirl · · Score: 1

      Morons with no actual understanding of the language say "could care less." It's just that there's a lot of them.

      The tone of the phrase is sarcastic, though. The negation is represented through tone, and voice, rather than being explicit in the wording. But then Grammar Nazis never actually do research into the rationality of their position, they're just driven by knee jerk rote lessons: "double negatives make a positive!"

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    41. Re:It's not dead, it's fun! by snowgirl · · Score: 1

      No, I'm still pretty sure it's all down to morons.

      It is... the morons being the ones who cannot understand the sarcasm in "I could care less".

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    42. Re:It's not dead, it's fun! by snowgirl · · Score: 0

      Overclocking my old Cyrix CPUs made a noticeable difference in a lot of games. These days? I just don't care. My laptop has a quad-core 2.2GHz i7. It is insanely fast. Even big compile jobs can run with -j4 and it's still responsive. My tablet has a 1.2GHz dual-core processor, and it's faster than the desktop that my mother uses - why would I bother overclocking it?

      I read that the best -jX to run at, is X = 2 * number of cpus (which would now be cores). Something to do with compiling being I/O intensive, so swapping to another task while the previous one is I/O blocked or something like that.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    43. Re:It's not dead, it's fun! by tomthegeek · · Score: 1

      Sidewalk is a specific type of pavement, in that it is on the side of the road for use by pedestrians (and arguably bicycles). Pavement is pretty much any type of concrete or blacktop surface.

    44. Re:It's not dead, it's fun! by bemymonkey · · Score: 1

      But overclocking would speed up that compile time significantly. Video encoding and the like that also benefits immensely... Depends on what you're doing, really - if your video is currently taking 30 seconds to encode, a 5 second improvement might not be that big a deal. If it's taking 30 hours to encode, shaving 5 hours off of that is pretty awesome... :)

    45. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      "I couldn't care less" => "I already care the very minimum amount"
      "I could care less" => "I don't care very much and in fact if I thought about it, I could care even less than I do now, which is to say not very damn much"

    46. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      A pointless debate, I know, but hey, I'm bored.

      It is also possible to interpret it to mean that it is not possible for me to care less than I do now because it is significant and I must care the amount I currently do.

      That would make sense if the statement was "I shouldn't care less" or "I mustn't care less". But it's not. It's "I couldn't care less", which can only mean that it's not possible to care less. Being not possible implies that desire is irrelevant. In other words, the unspoken addendum to "I couldn't care less" is "because it's impossible to do so". It's impossible because I already care the least that is possible.

      "I could care less" means that it is possible for you to care less than you do now. It may even be likely for you to do so. Saying "I could care less" means that it's not very important, it's possible for it to be cared less for.

      That makes even less sense. If it's possible to care less, then you do at least care a bit. This does not suggest that it's not very important. It suggests that it's at least a little bit important, or more specifically, more important that something that you care the least about. Something that you care the least about would be something about which you couldn't care any less than you do, because you're already at the bottom of the caring scale.

      Therefore, the only logical interpretation of those two statements is "I couldn't care less" means "I care the least I possibly could", while "I could care less" means "I care at least a little bit more than the least I possibly could". Both statements suggest that you don't care all that much, but the former suggests that you care even less than the latter.

      Put it into a different context. I care a great deal about my wife. I could, theoretically, care less for her than I do. In fact, there's lots of room between the amount that I do care for her and the complete absence of any caring, so I could actually care a whole lot less for her. I could even hate her, and be a total ass to her. But I don't. I care a lot for her. I could care less for her, but I don't. Therefore, I definitely do care for her.

    47. Re:It's not dead, it's fun! by Jstlook · · Score: 1

      As far as the 'irregardless' goes, let me say this:

      Just because a dictionary concedes that some people will use a word despite a lack of coherent meaning, and include it as a "non-standard" word, does not mean that it is truly a word in the language. Language is meant to convey information, and unless you're trying to convey the fact you're an idiot, you should probably choose where you use that. Here, as an AC is probably an excellent suggestion.

      Again this falls back to the wiki argument. Just because you can find information posted on the Internet doesn't mean that information is correct, and certainly doesn't mean you should use it just because Google showed it to you.

      --
      ---jstlook ---For that is the way of Elves, for they say both yes AND no, and mean every word of it. --- J.R.R.T.
    48. Re:It's not dead, it's fun! by TheRaven64 · · Score: 1

      You can use heuristics like that. Or you can try the really novel approach of actually testing. On this machine, -j4 to -j8 are about the same speed (although use more RAM, so with C++ stuff that takes 512MB+ per compiler process it can start swapping, which completely kills performance). Any higher numbers and it gets slower. -j4 is the optimum.

      --
      I am TheRaven on Soylent News
    49. Re:It's not dead, it's fun! by Tastecicles · · Score: 1

      When I'm elected, people who use the word "irregardless"* and the phrases "all of the sudden", "a whole nuther" and "to all in tents and porpoises" (or similar sounding bullshit) shall be relocated to work camps.

      That is all.

      *: Citation needed for any claim that this word is in any modern, widely-used dictionary; further that any such entry lists it as correct English rather than a double-negative.

      --
      Operation Guillotine is in effect.
    50. Re:It's not dead, it's fun! by Tastecicles · · Score: 1

      I'd drop a clue, but I can't be bothered.

      --
      Operation Guillotine is in effect.
    51. Re:It's not dead, it's fun! by Tastecicles · · Score: 1

      Some places use the "sidewalk" for human powered vehicles (cycles, skates) with the pedestrian walkway situated in between that and the vehicle track (also referred to as the "road"). So to take pedanticism to a whole new level, it can't be called a "sidewalk" because you wouldn't want to walk there.

      Oh, want a citation? Here you go - kick up Google Earth and look here:

      52d55m25s55N, 1d8m11s87W

      --
      Operation Guillotine is in effect.
    52. Re:It's not dead, it's fun! by UttBuggly · · Score: 1

      For me, It's fun and I could care less what some dude did with liquid nitrogen.

      First computer, I just used Asus Overclock and felt I got more for my money.
      Second computer, I started fiddling with manual settings.
      Third computer I pushed it until I couldn't get rid of the heat with air cooling.
      Fourth and current computer, water cooled and running awesome (6 cores at 4.3 GHz).

      Each time I felt the progress, it's like leveling your character, but the character is you, and the game is real life!

      My history of chasing the clock is much like yours, although my first effort was taking a Pentium 90Mhz CPU up to...wait for it...100Mhz! Some time later, jacked an AMD K6 from 350 to 450.

      Since then, I've tweaked every box I've built some amount, but nothing newsworthy by any stretch.

      Currently, have a Core Duo CPU at 2.5Ghz that I've gotten to 3.2Ghz on air. Have a relatively new MSI 560Ti GPU that I created a profile for when I run the Tribes Vengeance beta that jacks the card up a good bit and requires the dual fans to be at 75% to keep a 60C ceiling. My full size ATX case has 7 fans, so it's cool...and loud as hell. I don't care; the CPU never exceeds 40C under load and the GPU is as I mention above.

      My reasons for overclocking are entirely practical; I don't want to build another box right now but have SC2, COD MW3, and the aforementioned Tribes title that run more smoothly with the increased clocks than the defaults. So I'm not trying to prove anything...I'm just frugal.

      --
      I am my own gestalt.
    53. Re:It's not dead, it's fun! by snowgirl · · Score: 1

      You can use heuristics like that. Or you can try the really novel approach of actually testing. On this machine, -j4 to -j8 are about the same speed (although use more RAM, so with C++ stuff that takes 512MB+ per compiler process it can start swapping, which completely kills performance). Any higher numbers and it gets slower. -j4 is the optimum.

      Well, naturally, practical testing > heuristics. Being I/O bound is probably less of a big deal anymore anyways. Thanks for letting me know that I might want to update my heuristics.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    54. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      People only say that it's sarcasm when they get called out and realise how stupid they are. People like you.

      The Caring Continuum

    55. Re:It's not dead, it's fun! by Kevin108 · · Score: 1

      I didn't really read what you wrote but 6 for one, half a dozen the other.

      --

      It's a perfect time for being wasted.
      A perfect time to watch the stars.
      - Burden Brothers, "Beautiful Night"
    56. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Remember what you wrote here the next time you correct anyone for anything, Mr. Hypocrite.

    57. Re:It's not dead, it's fun! by snowgirl · · Score: 1

      People only say that it's sarcasm when they get called out and realise how stupid they are. People like you.

      Yeah, like I could care less what you think.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    58. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Okay, place "care" on a scale of 1 to 10, with 1 being a new album by Ace of Base, and 10 being Natalie Portman or Hugh Jackman (depending on your gender preference) naked at your front door. "I couldn't care less" indicates a 1 on the scale, the lowest possible score, the album by that band that was inexplicably popular for a few months fifteen years ago. "I could care less" could be anything from a 2 to 10, from "there's still half a bag of stale cheetos in the back of the cupboard I could eat" all the way up to "Hello, Natalie!"

      Does that help?

    59. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      I give 4.3 microshits about the units.

    60. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      All I do these days if I'm feeling daring is activate the 'high performance' power profile in Windows.

      Jokes aside, that would be a dumb thing to do. Especially if you overclocked.

      Your computer spends most of its time idle, even when under load you will probably only be putting load on the CPU or GPU or HDDs, not all at once. Windows' normal power management ensures that the system runs cooler and doesn't cost so much in electricity, running with "high performance" will just reduce the operational life and increase your bills.

    61. Re:It's not dead, it's fun! by VinylPusher · · Score: 1

      My server spends /some/ time idle, but a lot of the time it is handling 4 Java processes.

      These processes do low-latency transactional stuff. Typically, each tick of a job is so small that it doesn't trigger a frequency stepping of the CPU. Windows doesn't schedule these very well, typically making the single threads hop between CPU cores and actual CPU's.

      It all conspires to make performance suck unless I keep the CPU's running at full speed.

      Yes, I've tried setting processor affinity. It's of limited use.

    62. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Can we quantify "care" with units? Im still having trouble grasping this.

      Let me help you with a car analogy...

    63. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Haha does anyone else feel old, reading this?

      Back in my day, there was no internet, and overclocking didn't have a name until the internet. It was just something I figured out by reading my motherboard manual and changing the jumpers for the bus speed from 33 to 40. The DX2-80 was born. Duke 3d could handle more sprites.

      Later, I found out that what I had done had been done before, and it had a name: overclocking.

    64. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      Can I also get that quantification in Libraries of Congress?

    65. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      That's the different between "less" and "fewer" right there.

    66. Re:It's not dead, it's fun! by Anonymous Coward · · Score: 0

      It doesn't matter if its "obvious" what it means. It doesn't make it right to completely fuck up a saying simply because it's "close enough".

      http://google.com.au/search?q=could+couldnt+care+less
      It is agree that it is nonsensical and is a corruption of *n't care*

      The first person I ever questioned about this claimed that it was sarcasm, proving that they didn't even know why they were saying it.
      Saying "Americans say it this way" sounds insulting to me. I know many people from the USA and not from the USA who say it either way.

      The phrases "I do care about that" and "I don't care about that" are complete opposites.
      A bunch of people saying or doing something wrong does not make it right.

    67. Re:It's not dead, it's fun! by Burning1 · · Score: 1

      Can you be 20% more awesome?

      Some things just dont have units.

  4. Maybe, maybe not... by RogueyWon · · Score: 5, Interesting

    From a gaming perspective (typically one of the big drivers of overclocking), a few factors that might argue "yes, it's over":

    1) For quite a few years now, PC games haven't been forcing the kind of upgrade cycle that they did over the previous 20 years. When Crysis appeared in 2007, it was a game that gave many people an "upgrade or don't play it choice". And after that... the industry retreated. Consoles were the primary development platforms at the time and few PC games pushed significantly past the capabilities of the consoles. Not only did we not see any games more demanding than Crysis, but the vast majority of PC games released were substantially less demanding. As a gamer, if you had a PC that could run Crysis well, you did not need an upgrade. This situation lasted 4 years.

    2) Performance has become about more than clock-speeds. The main advances in PC gaming technology over the last few years have come from successive versions of directx. You can't overclock a machine with a directx 9 graphics card so that it can "do" directx10. Same goes for dx10/11.

    3) As the entry barriers to PC gaming get lower, the average knowledge level of users fall. PC gaming is, in general, easier and more convenient than it has been at any time in the past. Pick up an $800 PC, grab Steam and off you go. If you just want to play games and are using an off-the-shelf PC from a big manufacturer, you don't need to worry about switching around graphics drivers, sorting out hardware conflicts or any of the other little niggles that used to make PC gaming such a "joy". You can even find cases where PC gaming is easier than console gaming; the PS3, with its incessant firmware updates and mandatory installs has taken us a long way from the "insert game and play" roots of console gaming. People who are new to PC gaming just won't be coming from the kind of mindset that even considers overclocking as something you might even remotely want to do.

    4) Among "old school" PC gamers, I think there's been a growing recognition that overclocking has its downsides as well. In an economic downturn, when money is tight, you don't necessarily want to go risking a huge reduction in the lifespan of your expensive toys.

    That said, there are a couple of factors that might argue the other way (closely connected to the earlier arguments):

    1) System requirements are finally on the move again. After years in stasis, 2011 has seen the release of a number of games with equivalent or higher requirements than Crysis. Bulletstorm started the trend, but Battlefield 3 and - to an even greater extent - Total War: Shogun 2 have really started to push the envelope on PC hardware. A lot of developers openly admit to being bored with console hardware. Even though they still get most of their sales from the consoles, they are using the PC to push beyond what they can achieve there, both to get their studio noticed and to get themselves ready for developing for the next round of console hardware.

    2) The downturn also means that people feeling a squeeze on their budgets may be looking to get as much bang for their buck in terms of performance as possible. If you think that your new, overclocked PC will last long enough that you will be able to afford a replacement when it does start to give out, then why not take the risk?

    1. Re:Maybe, maybe not... by Anonymous Coward · · Score: 0

      I built my PC about 3 years ago with quality but not bleeding edge components. My graphics card is a couple of generations old (GT275) and cpu is a CoreQuad, not an i[insert odd-number here]. Not a single game I've played in the past 3 years has come close to taxing my setup, and most of them I play at max resolution (1920x1200) with most if not all high-end graphical options turned on. Hell I 2-boxed EvE for 6 months on the same machine and only had to turn down half of the graphical options to make things smooth for both.

      I HAVE overclocked this setup, but...I don't see the point anymore. GPU's are doing more and more of the heavy lifting, the CPU/Ram/Disk have ceased being the bottleneck.

    2. Re:Maybe, maybe not... by ripdajacker · · Score: 5, Insightful

      I think you're right. I've overclocked my i5 750 from 2.66 to 3.15, and the speed increase is.. well hard to spot. In benchmarks I certainly see it. It was much easier to do than in the good old days where it was jumper settings.

      I think the gist of it, at least for me, is that there's fun in it anymore. I have relatively high end gear, at least at time of purchase, and it all basically guides you to overclocking. It's not as bad ass as it used to be.

      This may be a bit biased since I now have much larger sum of disposable income compared to when I was overclocking.

    3. Re:Maybe, maybe not... by Anonymous Coward · · Score: 1

      Not just that, the whole recession thing combined with the really nasty moves by Microsoft with the release of Vista put a horrible taste in everyone's mouth, including their good friend Intel who were pretty pissed at being lied to over Vista.
      This kept a considerable number of people on XP for a long time, some still are even after 7 has come out.
      This put a huge dent in upgrades, which led to a freeze in R&D both for hardware and software since nobody was buying anything.

      Now there is a whole host of more confusion with respect to Windows 8, and the apparent internal war and distrust within Microsoft itself. (shown to some extent by those court records)

      Ever since Bill moved away from the top, things have just went downhill.

    4. Re:Maybe, maybe not... by pinkeen · · Score: 1

      Confirmed here. I can play almost all recent games on my old GTX260 in 1920x1080 fine. Sometimes I turn-down AA or disable SSAO (really, I don't see any justifiable difference with SSAO enabled). Only exception is BF3, haven't even tried it yet, I will once I upgrade.

      IMHO Games to be more visually appealing should go into the direction of improving character animation and polygon count. More full-screen shaders and effects like SSAO won't do much good, when there's fast paced action and immersive gameplay. You just don't see the difference any more.

    5. Re:Maybe, maybe not... by Anonymous Coward · · Score: 0

      @RogueyWon... you should DEFINITELY make that a journal article. A brilliant read!

    6. Re:Maybe, maybe not... by VinylPusher · · Score: 1

      I've only recently swapped my 7800GTX 256MB for a GT440 (a nice GDDR5 version).

      New card uses half the energy and performs quite a lot faster. When was a 7800GTX brand new? I paid >£300 for it, that much I can remember.

      Next upgrade will be when I can buy a card requiring no more than 75W (i.e., no external PCIe power) yet runs slightly faster than a GTX580.

      IMHO games to be more visually appealing need to ditch the idea of polygons altogether and be comprised of voxels or truly curved surfaces and implement at least some elements of ray-traced lighting and material properties.

      Fairly sure this will be happening within 20 years.

    7. Re:Maybe, maybe not... by alen · · Score: 1

      most games use the same few engines which are used for years at a time. only time the upgrade cycle starts again is when a developer releases a game with a new dev kit and new engine. otherwise some CoD games are based on a 5 year old engine and play just fine on old hardware even if it's this year's CoD game

    8. Re:Maybe, maybe not... by Anonymous Coward · · Score: 0

      You forgot 3 things.

      First, Intel isn't going to make the mistake of another Celeron 266 that can be overclocked to 500mhz with reasonable cooling gear and beat out the processor three times the cost performance wise.

      Second, as clock rates and transister counts go up, power draw increases. That celly 266 probably drew no more than 40 watts OC'd; most processors nowadays draw 90-130 minimum. OC that to the same level and you're talking 5-6 times the heat in the same area. We don't have the cooling tech yet to make a sustainable solution for that.

      Finally, why spend cash on OCing when you can spend it on Scaling? Go look at the prices for Dual-processor AMD Server boards with 8 slots of memory and a dozen SATA-2 ports that supports RAID5. Why blow $500 on cooling gear when you can blow the extra money to upgrade the entire system to a server-level unit?

    9. Re:Maybe, maybe not... by mrchaotica · · Score: 1

      No kidding. I recently upgraded my desktop (only because I had to; the motherboard quit working), and decided to get something fast enough for Skyrim (a brand new game). Even then, the CPU cost only $50.

      I only upgraded the CPU in the first place because I was going from DDR2 to DDR3; otherwise I'd have kept using my Athlon64 2800+.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    10. Re:Maybe, maybe not... by LoRdTAW · · Score: 2

      2) Performance has become about more than clock-speeds. The main advances in PC gaming technology over the last few years have come from successive versions of directx. You can't overclock a machine with a directx 9 graphics card so that it can "do" directx10. Same goes for dx10/11.

      Not necessarily, video cards are the dominant force in today's gaming rigs. The CPU has taken a back seat to the GPU as both graphics and Physics calculations are ran on the GPU. If anything, GPU overclocking should be the focus. Memory speed is not as big a deal since its already fast enough to move data between the graphics memory and main memory. Hell I have one of those AMD A8 APU's with dual channel DDR3 1866 and its amazingly fast. The GPU is center stage.

      4) Among "old school" PC gamers, I think there's been a growing recognition that overclocking has its downsides as well. In an economic downturn, when money is tight, you don't necessarily want to go risking a huge reduction in the lifespan of your expensive toys.

      I think its because hardware is so cheap and fast its not worth it anymore. In the days when the CPU, memory and motherboard ran you $1000 and a full gaming rig cost 2 granny, it made sense to buy mid/low end CPU's, add a hundred or so bucks in cooling gear and have the same performance as the $2000 rig. Now you can get an 8 core 3.6 GHz AMD for what, $270 bucks? Then throw in a decent motherboard for about $100 and 8GB of ram is just $50 bucks. A pretty powerful video card is about $200. Hard drive/case/ps/dvd are the only accessories left and you can get those for about $200-250 total. So what you are now spending what, 800-900 dollars. THAT IS FUCKING NUTS! Go back 4 or 5 years and tell people we would have such powerful and cheap hardware and they'd laugh at you. I have a spare PC that has an AMD A8, 8GB ram 1TB disk with the case/dvd/ps that I put together from newegg for 400 dollars. FOUR HUNDERD DOLLARS. That is twice the cost of an XB360 and way more flexible and powerful. It runs most of my games just as well as my older quad core AMD rig with a PCIe Radeon 5770. Its just not worth it to overclock anymore.

      Sure there are new crazy CPU's like the Socket 2011 i7 from intel, but in the end that isn't going to give you a huge boost in frame rate, the GPU will. Id love the new super quad channel i7 but what will I do with it? And AND gets the job done well for so little money. Low price high performance hardware killed overclocking.

    11. Re:Maybe, maybe not... by wisty · · Score: 1

      >> Not a single game I've played in the past 3 years has come close to taxing my setup

      Try Dwarf Fortress ;)

    12. Re:Maybe, maybe not... by RogueyWon · · Score: 1

      Agreed on animation. Way too many games have fantastically detailed character models, which looks and move like a bad special effect from an old movie. When "realistic" games are moving, we're still very much stuck in the uncanny valley.

      Dark Souls is a little bit better. Characters move with a sense of weight and momentum that you don't normally see in games. Plus the skeleton monsters look and move just like the ones from Jason and the Argonauts - which is to say, a good special effect from an old movie.

    13. Re:Maybe, maybe not... by synapse7 · · Score: 1

      ..and then there was Crysis 2, and Crysis 2 with the large texture pack.

    14. Re:Maybe, maybe not... by pinkeen · · Score: 1

      Now it occured to me that the trick might be less repetition. Take for ex. "rage" - megatexture maybe still immature (not ready for production IMHO) but it really makes the environment look way more realistic. It may be so because human mind is trained to recognize patterns, repetitions. The same goes for animation. It should have at least some randomness. The trick is to add randomness without making the movement appear even more artificial. I think they accomplished this to some degree in BF3.

    15. Re:Maybe, maybe not... by billcopc · · Score: 1

      I've never had an overclocked rig "burn out" before it's end-of-life, so the whole lifespan argument seems unfounded to me. Yes, I know in theory we're stressing these components a lot harder than spec, but I've never seen it happen myself. I've beaten the crap out of past PCs, then turned them into perfectly stable servers a few years in, or sold them to lesser minds to make room for the new hotness. The only CPU I actually killed was an old Athlon T-bird, the kind that would hit over 9000 degrees under normal operation. I don't think a 40% overclock did much to hasten its demise; if nothing else, the upgraded heatsink and deafening Delta fan probably bought more time, compared to the shit cooler it came with.

      But you're right about the games though, there hasn't been anything revolutionary on the graphics front since Crysis. Even Crysis 2 was a half-assed sequel that looked worse than the original. Quite a few recent A-list titles have been nothing more than cheap console ports that look like ass, and countless PC-native games are so poorly programmed. they make your balls-out SLI/Crossfire rig chug like an ATI Rage Pro. Those boneheaded game houses couldn't even begin to deliver Crysis-level graphics 4 years later, because they can barely push 10k single-textured polys on bleeding-edge hardware. PEBKAC, I say!

      --
      -Billco, Fnarg.com
    16. Re:Maybe, maybe not... by bemymonkey · · Score: 1

      "IMHO games to be more visually appealing need to ditch the idea of polygons altogether and be comprised of voxels or truly curved surfaces and implement at least some elements of ray-traced lighting and material properties."

      Quick question: Why do games NEED to be visually more appealing? Hell, I played through Modern Warfare 3 last weekend on a big monitor with surround sound and a big subwoofer, and was astonished at how immersive the experience was... the graphics are good enough to convey character emotions and the intensity of the story - so why would you need better graphics?

      Where do you draw the line between, say, an action game and a war simulation?

    17. Re:Maybe, maybe not... by VinylPusher · · Score: 1

      Would you prefer that we take e.g. the new ATI 7xxx series GPUs, set that as a "this is fast enough and has as many features as we ever need" benchmark and focus any future advancement on making GPU's that are exactly the same, only they use less power?

      The 'need' for advancement comes not just from making players feel more immersed, it's about meeting the requirements of HPC in order to provide realtime data visualisation and rapid number crunching. A lot of drug research is done via simulation of protein interaction, which lends itself perfectly to running on GPU's (hence the Folding@Home project).

      In 10 years time, most of us will be enjoying our Super-HD screens. Some of us may be enjoying our Super-HD, QLED, 48 bit colour, 120Hz, 3D, high dynamic range screens. Some of us may even have 3 of them in a multi-monitor config. That could potentially mean a GPU having to drive a (virtual) display resolution of 12288x2560. That is a minimum 180MB frame buffer needing to be refreshed 120 times per second.

      It's the difference between having to process 7GB/s versus today's reasonable maximum of 0.35GB/s. That will require some advancements in (GDDR) memory technology as well as obvious improvements to GPU performance. If we maintain the current pace of advancement, this may be possible within the next 5 years.

    18. Re:Maybe, maybe not... by bemymonkey · · Score: 1

      Resolution, sure... gimme gimme gimme!

      But do the actual models and animations used in the games really still need that much more improvement? I hit my peak graphics desire some time around CS:Source and UT2004...

      New gameplay tweaks and styles are the only things that have kept me playing games... the graphics have looked great for ages.

    19. Re:Maybe, maybe not... by VinylPusher · · Score: 1

      Your great, my great and everyone else's great may be entirely different ;)

      I've been waiting for true curved surfaces in games since... well, since there were 3D cards that did both 2D and 3D.

      Improved lighting, reflections (including reflected light/shadow casting), caustics and fluid motion... these are things I want to see in games. These are also all things not requiring much more effort from modelers and artists but /do/ require more horsepower from GPU's.

  5. Pointless in most cases by 1s44c · · Score: 4, Insightful

    Few people have any real need to sacrifice stability for a little more speed. Overclocking is pretty pointless for anyone with a modern CPU.

    1. Re:Pointless in most cases by Anonymous Coward · · Score: 1

      I certainly disagree. First, if you know what you're doing, you don't sacrifice stability. Second, most (as in virtually all) 2500K and 2600Ks easily reach 4.5 GHz, with little trouble. That's a third of extra performance for free, and it takes very little time to get it set up. A little more to stability test properly, but that can be done using a background process, so it can literally take less than an hour of active work to get those 33% for free.

    2. Re:Pointless in most cases by Anonymous Coward · · Score: 1

      If it doesn't sacrifice stability, then why doesn't Intel sell those CPUs @ 4.5 GHz ?

    3. Re:Pointless in most cases by Anonymous Coward · · Score: 0

      I overclocked, took me half an hour to get it right with the easy tools they have now, and it hasn't crashed a single time in the 6 months since. And my processor now benchmarks as well as ones that cost well over $100 more. Yes, I notice a difference in performance, and I'm pretty darn happy with it.

    4. Re:Pointless in most cases by VinylPusher · · Score: 1

      Buy the new Intel Core i7 2800K 4.5GHz!!*

      *New PSU required. TDP = 200W

    5. Re:Pointless in most cases by Kjella · · Score: 1

      I think it depends what you use it for... if it's primarily a gaming machine then the only thing that matter is perceived stability, if it appears to run the games stable then for all intents and purposes mission accomplished. Better frame rates, cheaper parts, bleeding edge drivers, freshly released game still being patched heavily... it's 99% stability all the way.

      If you care about producing some kind of output that needs to be correct, then you care a lot more about stability and correctness. I don't want crashes, I don't want corruption, I want a machine that's operating exactly as it should, if there's any kind of issue I don't want "is this because I overclocked it?" to complicate matters. It's actually more that it adds many more possible causes than that it actually is the problem. Just a few rounds of clocking back to spec, trying to see if the problem doesn't occur now then clocking back up as it turns out that had nothing to do with it wastes time.

      --
      Live today, because you never know what tomorrow brings
    6. Re:Pointless in most cases by 1s44c · · Score: 2

      'If you know what you're doing' excludes most computer users.

      It's not for most people who just want to run office tools, read web pages, and maybe play games. They can't handle the possible instability and won't notice the gains in most cases. You always sacrifice stability when overclocking, you are just hoping the difference is too small to be noticable.

    7. Re:Pointless in most cases by Anonymous Coward · · Score: 0

      Few people have any real need to sacrifice stability for a little more speed.

      Few people ever had that need.

    8. Re:Pointless in most cases by Manfre · · Score: 2

      For some family of chips, they are literally the same chip fused to disable cores or drop the clock speed. This means that the chips are designed to do more, but are intentionally crippled. It's cheaper for them to have a single manufacturing line than one for each chip in the family.

    9. Re:Pointless in most cases by 1s44c · · Score: 1

      You should haved logged in because that is a +5 insightful comment.

    10. Re:Pointless in most cases by Anonymous Coward · · Score: 0

      And that comments shows that you have really no understanding of overclocking.
      Most CPUs that are sold today are UNDERCLOCKED by default. If you overclock properly, there will never be any stability issues.

    11. Re:Pointless in most cases by adisakp · · Score: 1

      Few people have any real need to sacrifice stability for a little more speed. Overclocking is pretty pointless for anyone with a modern CPU.

      Not true... For example, Intels i7 2700K is an unlocked chip basically intended for overclocking. The stock speed is 3.5 GHz but with a simple Corsair Water Cooler just about anyone can get a stable overclock to 4.7GHz and that makes a huge difference in any CPU limited games and video transcoding. There are people going as high as 5.2GHz and stable on simple all-in-one water coolers like the H100.

    12. Re:Pointless in most cases by ChrisMaple · · Score: 1

      Part of the reason is safety margin. Another part is test vector coverage: a user's stability testing may not cover that single instruction with that worst case data pattern that fails. Yet another part is operating condition: Intel guarantees the chip to operate at maximum temperature at a given voltage less a specified tolerance; the owner can boost speed by boosting voltage and lowering temperature. A fourth part is the complex game of maximizing profit: not all the top speed chips can actually be sold at a premium price, so it's better to sell them at a lower price and a downrgraded label than not to sell them at all.

      --
      Contribute to civilization: ari.aynrand.org/donate
  6. No, it simply doesn't provide the extras it used t by djsmiley · · Score: 4, Interesting

    It used to mean windows would run faster, games would run faster, everything was FASTER MAN!!!!!111one.

    But now overclocking for the at home folks is a case of hit a button in your bios, or in some cases a physical button on the motherboard, and it'll do some overclocking for you, automatically. As its become more automated, the news worthy stuff becomes more and more expensive to implement and show off, and so most things are less news worthy and so it appears "overclocking" happens less. In reality I'd expect it happens alot more, and maybe even when people aren't fully aware of what they are doing.

    Also systems being so much faster now, generally provide the speed that users require of them, unless they are the kind of users to be pushing systems to overclock simply for the hell of it, like the guys who get in the news. However you don't see these guys then gaming and getting 200fps on these systems, or anything exciting like that anymore. Its simply overclocked, and shown it to be "stable" at said speed. No one ever goes "lets see how many FPS can we get outta this baby now!", its all become very much a concept thing rather than actually running systems at these speeds for any sensible amount of time.

    --
    - http://www.milkme.co.uk
  7. Most people don't understand that it's a bad idea. by Toasterboy · · Score: 5, Informative

    Look, digital electronics are still subject to analog limitations. When you overclock, you squeeze the hysterisis curve, increasing the probability that your chip incorrectly interprets its the state of a particular bit as the opposite value. i.e. you get random data corruption. This is why you eventually start crashing randomly the more you overclock.

    While overclocking a chip that has been conservatively binned simply to reduce manufacturing costs but is actually stable at higher clock rates is reasonable, trying to overclock past the design limits is pretty insane if you care at all about the data integrity. Also, you tend to burn out the electronics earlier than their expected life due to increased heat stress.

    I never overclock.

  8. The law of diminishing returns applies by DeathToBill · · Score: 4, Insightful

    Ten years ago, CPU and RAM speed were really big factors in how fast your PC felt. We've spent the last ten years optimising hell out of them, while still using 7200RPM spinning disks (if you're lucky). So, surprise surprise, today disk IO is what limits your PC's performance. Why overclock your RAM? It makes (almost) not difference to your IO speed.

    I got a new laptop just over three years ago. It had a 2.4GHz processor. I got my next new laptop a few weeks ago. It has a... 2.5GHz processor. Clock speeds have become almost irrelevant. What makes the new sucker fly is the SSD. Unfortunately, there is no BIOS setting, however risky, to change from disk to SSD.

    --
    Slashdot - News for Nerds, Stuff that Matters, in ISO-8859-1 Has just realised that beta makes this signature redundant
    1. Re:The law of diminishing returns applies by djsmiley · · Score: 1

      Hmmmm 5 years ago I was running a single core 1.4ghz

      Now I'm running 6 cores at 2.4 ghz.

      But, yes, your point is valid ;)

      --
      - http://www.milkme.co.uk
    2. Re:The law of diminishing returns applies by GNious · · Score: 1

      How do I overclock my SSDs? :D

    3. Re:The law of diminishing returns applies by Anonymous Coward · · Score: 0

      Tell that to the poor guys that still have to wait hours or days to render images. Im not saying that someone is going to overclock a render farm but to think that teir 3 disk IO is the only bottle neck is pretty un informed.

    4. Re:The law of diminishing returns applies by Anonymous Coward · · Score: 0

      For most people, the speed of their Internet connection is much more obvious than the speed of their machine.

    5. Re:The law of diminishing returns applies by Type44Q · · Score: 1

      I got a new laptop just over three years ago. It had a 2.4GHz processor. I got my next new laptop a few weeks ago. It has a... 2.5GHz processor. Clock speeds have become almost irrelevant.

      Processor speeds are very much relevant - and proportional to - processor performance. Your claim that they've become irrelevant is like claiming that the rpm's of an automobile engine have become irrelevant because a a car you just bought has a similar redline to one you bought decades ago. Logical fallacy?

      Disclaimer: sorry for the car analogy; I thought I wasn't supposed to have to resort to those here. :P

    6. Re:The law of diminishing returns applies by Tomji · · Score: 1

      Almost irrelevant for games. I got 6 or 12GB of ram. The whole game will be loaded into RAM, no need for an SSD.
      I am sorry I do not care how "Fast" my Word and Excel apps feel.

    7. Re:The law of diminishing returns applies by Anonymous Coward · · Score: 0

      Hmmmm 5 years ago I was running a single core 1.4ghz

      Now I'm running 6 cores at 2.4 ghz.

      But, yes, your point is valid ;)

      So 5 years ago you were running a (now) 10 year old cpu.
      5 years ago I had an Athlon 64 at 3.0GHz, which I believe outperforms your 6-core at any single-core task like most video games (given the same graphics card ofc).

    8. Re:The law of diminishing returns applies by greed · · Score: 1

      Ten years ago you could count on maybe 42 MB/s from that 7200 RPM hard disk--top of the line, premium device. Would have been, what, in the $250 for 80GB range?

      Today, you can easily get 145 MB/s from a 7200 RPM hard drive--near bottom of the range, commodity price. But--prior to the floods--it would run you $120 for 2TB. (Heck, I just run the 5900 RPM units, they're plenty fast enough.)

      High RPM drives don't actually give you much higher sustained read speeds--there's less data per rotation. Their random access performance may be better--but I find this year's 7200 RPM drive beats last year's 15,000 RPM drive easily.

      And I can buy a whole, whole, whole lot of 7200 RPM units, plus RAM for cache, save a bunch on power, and wind up with more storage for the department compared to the 15,000s. And spares are easier to source.

      But we do love our new SAS SSDs. They do random I/O like it was going out of style. Sustained rates in the 450 MB/s range.

      So, most of us use 7200 RPM drives because we need to store lots of stuff. Then we can add SSDs to handle highly random workloads.

    9. Re:The law of diminishing returns applies by Anonymous Coward · · Score: 0

      The newer processors run about 4 times faster at the same clock speed (not to mention the additional cores) than say, P4s. When you overclock, you go further. I OC my desktop and see a significant improvement in performance. Occasionally, my OC fails due to a BIOS reset at boot. I notice it even with something not highly sensitive to CPU such as browsing. What does not make economic sense is when people buy CPUs without great performance/$ and then spend extra on pricier motherboards, RAM, coolers etc so that they can OC. The final equation turns out only modestly advantageous, if at all... especially with all the reading one has to do.

    10. Re:The law of diminishing returns applies by Tastecicles · · Score: 1

      Holy hell, five years ago I was still running a single core at 1GHz (Pentium III Mobile - in a Dell).

      Now I run two cores at 1.6GHz with HD graphics core on the same die. Cost less than the PIII.

      I still play (nominally) the same games: Yuri's Revenge, Homeworld 2, UT2004, FS2000 and Railworks 3, though recently I've taken up Battlestar Galactica Online and Brothers In Arms (because frankly the PIII just couldn't even load it, never mind play it). These days the PIII is sitting in a box in a dark place because my dual core netbook (with a 1.66GHz Atom which cost less than an external hard drive) pisses all over it.

      --
      Operation Guillotine is in effect.
  9. Competitive Overclocking != all Overclocking by Anonymous Coward · · Score: 5, Informative

    That's like saying competitive soccer going broke would impact on EVERYONE EVER from playing soccer with their friends.

    Not everyone overclocks to beat a record.
    Hell, "overclock" a toaster if you have to. 2 second cold toast anyone? (the best toast)
    But really, there are still plenty of things you can overclock to beat records, such as what iB1 mentioned up there, overclock a smartphone or tablet.
    Overclock a Beagleboard, or a Raspberry Pi when it comes out, Arduinos. All these compact computers are pretty much sitting around waiting to be hit by the overclocking spirit.

    1. Re:Competitive Overclocking != all Overclocking by djsmiley · · Score: 2

      OMG You've just reminded me of the hell in a kettle overclocked kettle video on youtube.

      http://www.youtube.com/watch?v=rGL67coOdOk for those wondering.

      --
      - http://www.milkme.co.uk
  10. No by lga · · Score: 3, Informative

    No. Next question.

    Seriously though, both Intel and AMD sell multiplier-unlocked CPUs as a feature, and the winners of tests in PC Pro magazine are overclocked by the system builder. You can even buy upgrade bundles pre-overclocked. My latest motherboard came with one-click overclocking software and can adjust the clock speed through a web page while playing a game. Liquid coolers are mainstream. Overclocking is definitely not dead.

  11. Huh, no by buserror · · Score: 3, Informative

    "has the enthusiast passion for overclocking cooled off"

    Not from my 5.0Ghz Core i7 2600k anyway -- The tools have become better, the mobo are generally better built and more tolerant to punishment (some have 2 Oz copper), the power rails are a LOT more controllable than before, and in general the IC companies that make the power ICs have progressed a lot too in that time, so you can overclock easier, quicker, get better results and in general, extract quite a bit more, without nitrogen.

    And, I compile distros all day, to me going from 3.8Ghz max to 5.0Ghz stable (and quiet!) is awesome; make -j10 FTW !

    1. Re:Huh, no by TheGoodNamesWereGone · · Score: 1, Insightful

      I have a 2600K too, and I've squeezed 4.2Ghz out of it with air. That said, the most demanding thing a person usually runs is games. And since games are more and more becoming nothing but shitty console ports (I'm looking at YOU, Bethesda!) I wonder, what's the point? Maybe I'll try some compiling myself.

    2. Re:Huh, no by Anonymous Coward · · Score: 3, Interesting

      Why the fuck would anyone "compile distros all day" on their personal computer? If you're doing it for work, use the work machines. If you're doing it for a hobby, dude, get a better fucking hobby.

    3. Re:Huh, no by ifiwereasculptor · · Score: 1

      If you're doing it for a hobby, dude, get a better fucking hobby.

      Arguably, any hobby that involves fucking is better.

    4. Re:Huh, no by Dishwasha · · Score: 1

      I guess you've never heard of Gentoo...

    5. Re:Huh, no by Anonymous Coward · · Score: 0

      If you're compiling the whole of Gentoo daily then you're doing it wrong.

    6. Re:Huh, no by daspriest · · Score: 1

      Because he can.

    7. Re:Huh, no by bemymonkey · · Score: 2

      Which games do you know of that are CPU bound at 100FPS? There aren't very many... if you're getting crappy gaming performance, you'd be better off overclocking your GPU.

    8. Re:Huh, no by Anonymous Coward · · Score: 0

      Personally, I always seek the approval of slashdotters before starting a new hobby. I wouldn't want to be criticized by someone who knows nothing about me.

  12. Gains aren't there by DNS-and-BIND · · Score: 5, Insightful

    In the days of the 300MHz Celeron, you could overclock it to 450MHz and gain 50% improvement. That extra 150MHz represented several hundred dollars straight to Intel, which you kept in your pocket by overclocking. These days, a few percent? It's just not worth the trouble any more.

    --
    Shutting down free speech with violence isn't fighting fascism. It IS fascism!
    1. Re:Gains aren't there by kermidge · · Score: 1

      Why was this rated -1?

    2. Re:Gains aren't there by Rockoon · · Score: 2

      Thats his default karma.

      Trolling it a bitch.

      --
      "His name was James Damore."
    3. Re:Gains aren't there by chichilalescu · · Score: 1

      Are you saying that overclockers are stealing megahertz from Intel? damn pirates!

      --
      new sig
    4. Re:Gains aren't there by Anonymous Coward · · Score: 0

      Its not 100% true. My Phenom 2 6-core overclocked from 3.2 to 4.2ghz on air. Gain of around 25%

    5. Re:Gains aren't there by Anonymous Coward · · Score: 0

      Same thing today. ...sat time i was buying a processor, I bought the cheapest C2D model and overclocked it to the specs of a much more expensive product thus saving myself some monnies. :)

    6. Re:Gains aren't there by Nimey · · Score: 1

      The gains are still there, it's just not as easy as the legendary Celeron-300A, which itself was an anomaly[1]. I've OC'd my i5-2500K by around 1 GHz, which isn't as big a percent gain (vs. stock 3.3 GHz) as the old Celery, but is still "free"; granted I had to pay around $10 more for an OCable processor, but it's small enough vs the cost of a system to be essentially free.

      Even a non-OCing i5 will automatically overclock one or two cores if you're running a program that's not sufficiently multithreaded to use all four cores but is close to saturating the cores it can use.

      [1] Srsly, why do you think people always get starry-eyed about that processor? Nothing else in its time would OC near as well, especially at that price point, and /especially/ with that Abit motherboard that let you run two CPUs which were never intended for SMP. I kind of regret going for a standard P2-350 at the time.

      --
      Hail Eris, full of mischief...

      E pluribus sanguinem
    7. Re:Gains aren't there by scharkalvin · · Score: 1

      The crap really hit the fan back then when Abit came out with their dual socket 370 MB that allowed you to run TWO over clocked Celerons in SMP. Intel took a fit since they never intended for the Celeron to run in dual cpu mode. IIRC the Celeron 300 or 330 would overclock to 500mhz by increasing the bus speed from 66mhz to 100mhz and changing to PC-100 dram. Some outfits starting to sell "1ghz" computers, which were really two Celeron processors running at 500mhz. Problem was you couldn't take advantage of an SMP system then unless you ran Linux or BSD as Windows didn't yet support it (well maybe NT did).

    8. Re:Gains aren't there by bemymonkey · · Score: 1

      50% is pretty rare these days, but 20-30% isn't all too uncommon...

    9. Re:Gains aren't there by kermidge · · Score: 1

      Thanks. Now I know where to look.

    10. Re:Gains aren't there by sheddd · · Score: 1

      That era was when I first assembled a pc from parts. Fun times :) I was broke so I had the no cache 266 Celeron with no L2 Cache; It ran @ 400 or so IIRC.

  13. Hahahaha by aitikin · · Score: 3, Interesting

    "...has the enthusiast passion for overclocking cooled off?"

    That's like saying, "Do nerds no longer need a proxy for phallic measurement?" As long as there's still testosterone (even if it is minimal in some here) there'll still be people (men mostly) looking to say "We did it first!"

    --
    "Don't meddle in the affairs of a patent dragon, for thou art tasty and good with ketchup." ~ohcrapitssteve
    1. Re:Hahahaha by Viol8 · · Score: 2

      I think even most geeks think of overclockers as a little bit obsessive and kind of out there.

  14. Re:Most people don't understand that it's a bad id by nzac · · Score: 3, Insightful

    You just don't get the overclockers mentality.
    Either is all part of the fun adding to the risk or you are getting the most out what you paid for and are still within stable limits.
    I don't think many overclockers care about random data corruption unless they blue screen or they turn it off when they need stability.

  15. Can't notice the difference anymore by dan_barrett · · Score: 5, Insightful

    I think CPU speed is less of an issue these days; eg Core2 onwards processors are generally "fast enough" for most users.
    Compare the change in noticeable speed between a 386 and 486, or even Pentium vs Pentium 2 or 3, to today's Core2/Athlon vs Core i5/Phenom.
    Most people don't notice the jump in CPU performance on modern processors.

    The other traditional bottlenecks are rapidly disappearing too, eg a midrange Directx10 graphics card is good enough to play all but the most demanding games these days, and memory and disk speed and capacity are generally outpacing most people's demand.

    People will still overclock for the challenge of it, but I think there's no tangible day-to-day benefit anymore.

    As someone above mentioned, the real performance battle has moved to portable devices, eg how much performance can you get from a tablet or phone, given a fixed battery capacity?

    1. Re:Can't notice the difference anymore by TheGoodNamesWereGone · · Score: 1

      I recently upgraded from a Phenom II 810 (quad-core, 2.6Ghz) to an i7-2600K (quad-core w/hyperthreading, 3.4Ghz). I saw a BIG difference in games. But you're right, most users are content with 'fast enough'.

  16. Overlocking was only ever a dick waving contest... by Viol8 · · Score: 1, Funny

    ... between a tiny bunch of geeks who had more money than sense. Someone should have told them that if they really wanted to play their first person shooter faster they should overclock the graphics card, not waste time on the CPU.

    Yeah , I 'll get modded down for offending the high priest overclockers who read this, but really, if you spend 1000s on a special cooling system for your CPU just so it runs 25% faster so you can get even more unnoticable frames per second you really need to get out more.

  17. Re:No, it simply doesn't provide the extras it use by Anonymous Coward · · Score: 0

    In the old days when CPU speed was relatively low, every % speed increase made a big difference. Now the base speeds are so high you aren't really gonna notice the difference.

  18. As in the words of Sir Arthur C Clarke! by NSN+A392-99-964-5927 · · Score: 3, Insightful

    When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is probably wrong.

    Therefore, I rest my case.

    Cheers Arthur a true friend who is missed but still there :)

    --
    All cows eat grass!
  19. Car analogy by qxcv · · Score: 1

    When a land based automobile broke the sound barrier, was amateur automobile racing over? When Amundsen reached the South Pole, was adventuring over? Rather than eliminating enthusiasm for these ventures, such achievements tend to make these things more popular and accessible. Hell, anyone can get into higher-end over-clocking nowadays without having to worry about it taking too much of a toll on their wallet. A few months ago I glued a PVC tube to the lower half of a stock 775 cooler and threw some dry ice in to try and see what I could get the machine too. I mightn't have clocked my RAM up to 3600MHz but it was still enjoyable seeing my old box get put on cyber-roids (also playing with the dry ice afterwards - there's a lot of fun things you can do with it ;) ). Maybe it's not a high stakes game of processor roulette anymore, but geeks tend to be experts at doing stuff "just because they can".

    --
    "The most dangerous enemy of a better solution is an existing codebase that is just good enough." -- Eric S. Raymond
  20. It's now mostly irrelevant... by r3verse · · Score: 1

    I'm in my late 20s. Fifteen years ago, in the heady days of personal computing, I spent long hours squeezing every last Kb and clock cycle out of my (underpowered) machine. It was new. It was fun. But it also crashed, and burned. These days I want things to 'just work'.

    And mostly they do. In 2011, we are not wanting for processing power, in most any electronic device we own. Xbox, PS3, iPod Touch. People of the current gen., financially well-endowed and relatively unburdened, face the same challenge I faced, and proceed to ascertain how quickly they can prestige on CoD. The paradigm has shifted.

    ---

    In retrospect I have unintentionally managed to blame all three of the unholy trinity for the death of power-user computing. Go subconscious, go!

    /end_thread

  21. I bet you're the life and soul of a party by Viol8 · · Score: 5, Funny

    [Sultry babe walks up]
    "Hello, and what do you do?"

    [nasal voice]
    "I compile distros all day. Yes, did you know that Slackware on average compiles 20% faster than Debian for 64 bit but if I overclock my Core i7 by raising power rail voltage and tweeking the quantum flux capacitor.... hello, where are you going..hello? Hey, come back, did I mention its a 2600K? Hello?"

    1. Re:I bet you're the life and soul of a party by buserror · · Score: 3, Insightful

      Is this really "slashdot.org" where "nerds" used to be around ? You know, nerds, who do technically oriented stuff "just because they can" ?

      The various comments on this topic -including the one up- makes me wonder really, or has "nerd" become more of a "I'm such a nerd, babe, look, I installed an app on my smartphone".

      Or /. has been mirrored to "hipster.com" and I'm accessing the wrong portal

    2. Re:I bet you're the life and soul of a party by Viol8 · · Score: 1

      Oh dear. Someone's had a sense of humour bypass.

      You did rather set yourself up for it saying you compile distros all day. I mean, even for a nerd thats a bit of an odd thing to do. Once a week/month to rebuild a kernel sure, we've all done that at some point, but every day building entire distros? Why??

    3. Re:I bet you're the life and soul of a party by buserror · · Score: 2

      Ever heard of embedded development ? Or, maybe you think that distros themselves just appear magically as an ".iso" file brought by father xmas ? I'm sure you're very proud of having recompiled your kernel at some point, and that seems to have given you enough insight into general software development to make large, broad statements about it all.
      Actually, I /do/ find it funny, but not in the way you probably intended.

    4. Re:I bet you're the life and soul of a party by Anonymous Coward · · Score: 0

      oooooooooooooooooooooh

      *lifts handbag*

    5. Re:I bet you're the life and soul of a party by Anonymous Coward · · Score: 0

      Nerd Alert!
       
      /troll

    6. Re:I bet you're the life and soul of a party by Anonymous Coward · · Score: 0

      I see what you're saying, but I have to admit that I got a geek hard-on when the GP said "make -j10".

  22. Re:Overlocking was only ever a dick waving contest by fitteschleiker · · Score: 2

    Don't worry those guys then convince themselves they have a better visual perception than normal people so they don't feel stupid that they paid said 1000's. Much like the audiophile who buys a massively expensive sound system. Anyone asks them the hell why, they subtly (or not so much) imply that they have a hearing range that is far in excess of your standard human, and of course a better appreciation for music anyways :)

  23. Re:No, it simply doesn't provide the extras it use by djsmiley · · Score: 1

    Yep, thats one reason.

    Others are its simply too expensive to do the kind of overclock you'd of done 5 years ago, for the same increase in %. I guess chip creators are getting better at running the chips at their limit already, rather than shipping them lower than they "could" go, tho I don't have any stats to back this up.

    --
    - http://www.milkme.co.uk
  24. Translation: by ProfanityHead · · Score: 2

    PC's are so fast these days for our simple minds there is no longer a need to overclock.

    And GET OFF MY LAWN!

  25. Huge difference for game development by llZENll · · Score: 4, Informative

    Well let me dig up my test results spreadsheet from when I first got my 2500K CPU, times are in seconds to complete my task in Visual Studio 2010, first set of numbers is the system at stock clock, second set is overclocked at 5GHz, during my game development most of my day consists of building the game, loading the game and testing out changes or additions, therefore the reduction from doing that in 32s vs 21s is absolutely huge, even doing code changes that don't require a total rebuild I am waiting 3s less. It may not sound like a lot but when you are focused any time saved is very important, you can only be focused for so long.

    build debug from clean 12.9 6.9
    built already, go and load all effects and units 8.2 5.6
    at title screen all loaded, start medium map 19.6 14.3
    modify main.h build load to splash scrn 3.4 2.1
    modify main.h load into medium map 31.9 20.9
    modify main.h optimal load no sound, small map 16.9 10.3
    running in game, modify main.h apply changes 10 6.7
    average 14.7 9.5

    system is 2500K, C300 SSD, 16GB memory

    1. Re:Huge difference for game development by jamesh · · Score: 1

      There's always someone who comes along and spoils an argument with facts and evidence :)

    2. Re:Huge difference for game development by xmas2003 · · Score: 1
      Just to add another data point, I was able to push my 2500k to 5.0GHz (liquid cooling) and it ran Prime95/etc. stable ... but did get fairly warm ... and there was an intermittent bug on restore from sleep that may be fixed in the next version of ASUS Bios.

      So I backed off to 4.7GHz ... runs a lot cooler and has been rock-solid stable. So basically got an extra GHz in performance for free ... and yea, as the OP says, all those reductions in time add up and make for a more pleasing experience.

      --
      Hulk SMASH Celiac Disease
    3. Re:Huge difference for game development by drinkypoo · · Score: 1

      Restore from sleep, I find that very interesting. I just installed a bios update to my Gigabyte Ga-MA770T-UD3P 1.0 and fixed a sleep problem that I never had before overclocking with AMD overdrive. At least, I think that's what fixed it, I didn't change anything else, but there could have been a windows update in there someplace.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    4. Re:Huge difference for game development by scsirob · · Score: 1, Insightful

      I think you should consider going the opposite route. Optimize your game so it runs 50% faster on the same hardware. Unfortunately it seems that software optimization is an art long gone.

      --
      To Terminate, or not to Terminate, that's the question - SCSIROB
    5. Re:Huge difference for game development by bjackson1 · · Score: 2

      I agree with this. As a gamer myself, I am tired of how long it takes to build the game from source in Visual Studio. When will they optimize?!

    6. Re:Huge difference for game development by Anonymous Coward · · Score: 0

      Optimizing the compiler used to build your game is a helluva lot different than optimizing your game.

    7. Re:Huge difference for game development by llZENll · · Score: 1

      I do all of the above. I spent a considerable amount of time optimizing the loading code path, I have the details somewhere, using memory caching techniques, ensuring the cache isn't blown, serializing disk access, etc, all led to massive performance improvements. When you are reaching a point of diminishing returns on that, why not upgrade the computer? I built my 2500k over a year ago and it was around $1500, pretty much a no brainer upgrade for that price.

    8. Re:Huge difference for game development by Anonymous Coward · · Score: 0

      If you build 1,300 times over the life of the project, that 11 seconds will save you four hours total, mathematically. Focused or not, your time and attention between builds will still be fragmented and interrupted by outside factors. Think of it as gaining an extra 5 FPS by OC'ing your video card: jump from 100 to 105 FPS. You won't actually see it, you won't actually realize any gain. Same for your builds. Sorry, but that's drop-in-a-bucket false savings.

    9. Re:Huge difference for game development by BLKMGK · · Score: 1

      That's also not a long build time. Overclocking my machine and moving from a 2 core Hyper machine to a 4core Hyper saves me HOURS when I'm encoding video. A long multithreaded compile is likely much the same I'd imagine. Judging from what I see on my two differently clocked Atom machines that compile XBMC even small changes make a difference on longer compiles...

      --
      Build it, Drive it, Improve it! Hybridz.org
    10. Re:Huge difference for game development by Anonymous Coward · · Score: 0

      I don't consider his minimal time differences of under 15 seconds in all cases to be a very compelling argument. Let's increase the chance of data corruption and component failure because we're that impatient! Yeah, no thanks.

  26. Re:Overlocking was only ever a dick waving contest by llZENll · · Score: 2

    Not everything is about gaming, I overclocked for faster compiling cycles, and it makes a HUGE difference. And for gaming the market has long figured it out, hence the huge market for overclocking GPUs, you can get custom water blocks, heatsinks, memory coolers, all specifically for video cards. The market for overclocking GPUs dwarfs the market for overclocking CPUs in the yesteryear, people have been overclocking GPUs since they very first came out from 3dfx.

  27. I'm usually doing the opposite... by damn_registrars · · Score: 3, Insightful

    I tend to underclock more often now, to reduce power consumption on my systems. Of course, I don't play any games on my systems, so I am almost never pushing the capabilities of the hardware.

    --
    Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
  28. Re:Overlocking was only ever a dick waving contest by Viol8 · · Score: 2

    Isn't that the truth! It reminds me of people who still claim that LPs sound better than CDs even though the LP stereo system is a hack that doesn't always reproduce phase properly and the audio before its recorded on an LP has go to through a compressor first to limit the amplitude because of physical restrictions in the offset of the groove and also the high frequency response is limited because the goove simply can't be machined to undulate enough accuratly enough to reproduce them especially at 33rpm closer to the centre of the record. But hey, its analogue so it must be better, right?

  29. we wouldn't need over clocking if our code were fa by MichaelCrawford · · Score: 1

    -ster.

    the hardware vendors devote tens of billions of dollars every year to keep up with moo re's law. this has lead to the fallacy that CPU power, ram and storage are infinite, so many of today's coders don't even bother to optimize their code.

    consider initializing a 2d array in a nested loop. if you increment columns in the inner loop, the memory cache will speed you up. but a dumb mistake could increment rows instead. in that case the cache actually slows your code down dramatically.

    it is wrong that few coders ever learn assembler or hardware architecture anymore. not so you can write assembly code, but so you can understand the effect that your java or perl has on the underlying hardware. the only code that ever touches the metal, after all, is machine code.

    --
    Request your free CD of my piano music.
  30. In this discussion. by richy+freeway · · Score: 1

    1. People who do overclock and reap the benefits.

    2. People who don't and like to moan about it.

    1. Re:In this discussion. by Anonymous Coward · · Score: 0

      3. People who underclock so they can get a silent PC.

  31. Re:Overlocking was only ever a dick waving contest by jamesh · · Score: 1

    if they really wanted to play their first person shooter faster they should overclock the graphics card, not waste time on the CPU.

    I always thought they did that too?

    Yeah , I 'll get modded down for offending the high priest overclockers who read this, but really, if you spend 1000s on a special cooling system for your CPU just so it runs 25% faster so you can get even more unnoticable frames per second you really need to get out more.

    Maybe it's been a while since you were a kid but there's something enticing about "sticking it to the man"... robbing Intel of those few $$$ by taking a cheap CPU and running it as fast an expensive CPU. Intel (and AMD probably) know exactly what's going on and how best to make money out of it though :)

  32. It's not about using it. by petes_PoV · · Score: 4, Insightful

    For a small proportion of the population (but, possibly, a large proportion of slashdot-ers) a PC is not a platform for doing useful work or serving entertainment, it's a source of "fun" in its own right. In past decades the people who like to play with their computers would be out in the yard, covered in oil, fiddling with a junky old car, or tuning a valve radio. Now they get their satisfaction from squeezing the last few MHz out of their PCs - whether there is any need or use for those few extra cycles, is immaterial.

    And for those with a more software bent, than a hardware leaning, there's always OSS - which serves a similar purpose.

    --
    politicians are like babies' nappies: they should both be changed regularly and for the same reasons
    1. Re:It's not about using it. by Anonymous Coward · · Score: 0

      Ah yes, the BIOS clock multiplier settings. I mean that thing has like, six different settings! It would take you a lifetime to explore the possibilities! And don't forget the excitement of it all... Power on! Now hit F10! Aw shit, too slow, try again! Ok quick now, RightArrow RightArrow Enter Down Down Enter! Now select the "2.5" setting! Enter again! Save and reboot!

      We're having fun now, bitches!!!

    2. Re:It's not about using it. by Anonymous Coward · · Score: 0

      Except that tinkering with a PC is such a low skill thing, these kids with the "rigs" are fucking clueless idiots with any sort of real world devices such as cars or radios.

      There is vastly more knowledge needed to rebuild an engine than to buy a bunch of parts at frys and screw them into a case. "PC building" is not a skill. Its tinker toys or legos. It either fits, or it doesn't. It works, or it doesn't. "Building rigs" is retarded and easy, and is not a skill. Thats all kids know how to do these days.

      I replaced a controller chip on my C64 with a soldering iron and a replacement chip found at a surplus electronics store. THAT is "modding" a computer. I had to solder on the damn main board. I also taught myself assembly language before I knew it was called assembly language. By looking up byte codes in the processor manual. So I could give my guy extra lives in a game, by tracing the code to the proper memory locations and modifying those values. When I was 9.

      These days kids are so fucking stupid about everything around them. Putting neon lights on a car is not modifying it. Buying a new GPU isn't "building a rig". Its tinker toy consumerism.

      The tide has turned, and kids these days are actually DUMBER than previous generations.

  33. Missing the point. by Artea · · Score: 1

    I see people asking what practical value overclocking has. Why do people tune up cars? Superchargers? They do it because they can, not because "Hey, I can reach the speed limit 5% faster". It's a hobby, something fun to do. It's not about how much you spend, or if you are going to really benefit from that 600Mhz boost. It's all about the point of view, whitewashing it as a dick waving contest, or trying to apply a practical need to it is pointless really.

  34. Re:yes, there is - RAID by Anonymous Coward · · Score: 0

    Damn. Have to post as AC now or I'll undo a swag of mods.

    Unless you actually logged out, your mods were undone. Simply selecting "Post Anonymously" is insufficient to protect the mods.

    This misfeature should be better documented.

  35. Hehe by Arancaytar · · Score: 1

    with the aid of liquid nitrogen which cooled the RAM down to a frosty -196C [...] has the enthusiast passion for overclocking cooled off?

    I see what you did there.

  36. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 3

    While "low end" chips may be conservatively binned to reduce manufacturing costs, thats really not the whole story at all.

    When the highest end chips can be clocked from 3.8 to 4.5ghz and higher using the stock cpu cooler, doesnt it make you wonder why Intel/AMD do not sell any 4.5ghz versions of these chips? Its because the OEM's fuck up case cooling every single time.

    If Intel sold a 4.5ghz i7, Dell would still put it into a case with horrible venting and only a single fan that has been poorly placed, and then Intel would be footing the bill for loads of warranty replacements. The reason the i7 980X's cost so much isnt just because Intel was taking advantage of performance enthusiasts irrationality.. its because the Dell's of the world fuck up cooling every single time. The sandy bridge i7's perform nearly as well but run a lot cooler so can survive the harsh conditions the OEM is going to hamstring them into, and THAT is the main reason why they are so much cheaper than the 980X's.

    --
    "His name was James Damore."
  37. Its fun? by Xerp · · Score: 1

    Overclocking is fun. Maybe you'll find a new new cooling method too! http://www.bbc.co.uk/news/science-environment-16285036

  38. Re:Most people don't understand that it's a bad id by Ardyvee · · Score: 1

    Sure. That's why you do test when you overclock in order to ensure it's still stable. You should also know that sometimes the CPU is capable working faster than what it was designed because it so-happened-to-be one of those that was produced with a quality greater than expected.

    Or sometimes the cpu is already at the end of it's life and you're just giving it a hand to last a little while longer, which is my case, as I have an Intel Pentium D 3.2GHz dual core. It really depends on the case.

    --
    I don't care if I'm wrong. I only care about everyone obtaining something from the discussion.
  39. To ruin the sound of course by Anonymous Coward · · Score: 0

    Nothing makes more sense than buying highly noisy cooling systems so you can overclock your CPU and GPU for getting better graphics. And the sound coming from your expensive sound card is better when you can't heard it anymore due to all the noise coming from the cooling system anyway.

  40. Re:No, it simply doesn't provide the extras it use by drinkypoo · · Score: 1

    Speaking as someone who was overclocking Cyrix chips and AMD K6s, I'm super-glad that now I can just run a program and have the computer overclock and stress test while I sleep. I bought a 2.8 GHz processor and it goes 3.4 GHz for no additional cost. That's a small bump, but it cost me nothing, so it's very difficult to complain. Every car is different and some are just a little better built than others and could take more tuning, and lo and behold, the car's computer is self-tuning, and tunes itself for efficiency continually. The computer ought to do the same thing. I shouldn't have to run AMD Overdrive manually, it should (optionally) do a short run every time I shut down for the night. The next generation should dynamically overclock itself at all times, monitoring onboard sensors to determine when it nears the thermal envelope of reliability.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  41. It reminds me old time... by dolmen.fr · · Score: 1

    I remember how in 1998 it was said that overclocking is dangerous and that it can kill your processor.
    My dad still use nowadays my old overclocked Intel Celeron 300A. I only had to reduce the RAM usage (one bank is dead and searching for a new one is not worth it).

    1. Re:It reminds me old time... by drinkypoo · · Score: 1

      I remember how in 1998 it was said that overclocking is dangerous and that it can kill your processor.

      back then, it could. If you had clocked up those 300As just a little more you could have burned them up real good. modern CPUs have thermal protection and it's relatively difficult to kill them by overclocking.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    2. Re:It reminds me old time... by BLKMGK · · Score: 1

      Ummm no. For starters you don't know how high his was clocked. Mine were clocked up pretty good and when there were cooling failures the machine always locked up before any damage was done. I had one friend bring me back a machine I'd built for her because it was locking up. It was a P2 machine I think, had a slot processor card like the 300A did. Anyway, I am looking at the thing and cannot quite figure out what's wrong, why does the heatsink look funny? I reach in and grab the heatsink promptly leaving burned fingerprints on it! Turns out the fan had become unplugged and was no longer providing any airflow. The machine had apparently been like this for weeks before she brought it to me. All I did was plug the fan back in and it becamestable. That machine was overclocked about 40% and there was never an issue with it afterwards either.

      --
      Build it, Drive it, Improve it! Hybridz.org
  42. I overcocked my girlfriend by Chrisq · · Score: 0

    I overcocked my girlfriend

    1. Re:I overcocked my girlfriend by Anonymous Coward · · Score: 0

      I overcocked my girlfriend

      You need to either upgrade to a faster model or have a cluster

  43. Re:Overlocking was only ever a dick waving contest by Viol8 · · Score: 1

    Sure, you stick it to Intel. But this grand gesture then means you give twice the extra money you've have spent on an equivalent out-the-box CPU to some other faceless corp who provide overclocking kit and who may be just as venal and grasping as Intel/AMD/whoever. Plus you reduce the life of your CPU. *shrug*

  44. Re:Overlocking was only ever a dick waving contest by VinylPusher · · Score: 0

    Actually, my hearing range is a little bit extended at high frequencies. I can tell the difference between CD quality and 96KHz/24bit.

    I wish I couldn't, because I'm still paying for my sound system :-(

  45. Re:Overlocking was only ever a dick waving contest by drinkypoo · · Score: 1

    Yes, in fact, any decent graphics card now has overclocking options right in the settings, so that you don't have to use a third party tool that may do it wrong and confuse the driver. For nVidia, for example, you just bring up the nVidia control panel and select "Performance" -> "Device settings" from the tree and you can fiddle with the clocks and on some cards even the voltages.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  46. Re:Most people don't understand that it's a bad id by VinylPusher · · Score: 2

    I'm sure it's more to do with the fact that Intel do not want to advertise a CPU with a TDP of 200W.

  47. Re:No, it simply doesn't provide the extras it use by VinylPusher · · Score: 1

    It's costing you in power usage.

    I read some recent CPU / system reviews and the increase in power consumption whilst overclocking some of the latest CPU's and GPU's is... scary.

  48. Re:Most people don't understand that it's a bad id by jampola · · Score: 1

    Dad? You're on /.?

  49. you wouldn't need to continue your first sentence by drinkypoo · · Score: 1

    in your comment if you didn't begin the comment in the subject line.

    See how annoying that is? The subject line is for a subject line, not for the beginning of your comment.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  50. Overclocking has plenty of uses... by jampola · · Score: 1

    I have a friend who overclocks 2 of his PC's specifically for compiling duties. When it saves an extra 20% of time when compiling a Linux Kernel, then more power to ya (no pun intended of course)

  51. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 1

    ..yeah.. thats it.. Intel wouldnt want to advertise and sell faster CPU's... because of a number that consumers dont understand or care about...

    Are you not from earth or some shit?

    --
    "His name was James Damore."
  52. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 0

    You are right, I do not (not the OP).
    I use my PC for work and fun. Recovery from a corrupted svn database, for example, could be a nightmare which is not worth 10% extra speed from an already fast CPU.

    So while I can understand that one wants to overclock some PCs. I will never understand why someone would risk the corruption of personal data, unless they are of no value to you.

    How old are you? Have you produced anything of value yet?

  53. Re:Most people don't understand that it's a bad id by VinylPusher · · Score: 1

    Perhaps they are content to derive £800+ from CPU's that run at a nominal 3.33GHz for now.

    If you think marketing and pricing strategies are so simple, perhaps you ought to tell us more about how you could improve Intel's market position and profitability by making some simple changes to the way they do things?

  54. Re:No, it simply doesn't provide the extras it use by drinkypoo · · Score: 1

    I read some recent CPU / system reviews and the increase in power consumption whilst overclocking some of the latest CPU's and GPU's is... scary.

    The problem with that idea is that most of the time the processor is idle and then it runs at the same 800MHz that it always did, no more, no less. There should be slightly more leakage with the slightly higher (under 10%) voltage, but the difference in temperature at idle is nonexistent. And most of the time, the computer is asleep, and using precisely the same amount of power no matter how I clock it. The TDP of my whole system is less than some graphics cards of today, so it's a non-issue for me in general.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  55. I agree by Anonymous Coward · · Score: 0

    (see subject)

  56. Overclocking CPUs => overclocking GPUs? by Lazy+Jones · · Score: 1

    CPUs are no longer really interesting for number crunching, having been replaced by GPUs. They are also being overclocked with liquid nitrogen: 3dmark record from November '11. I bet we will see more of this kind of insanity as GPU support for all kinds of number crunching increases.

    --
    "I love my job, but I hate talking to people like you" (Freddie Mercury)
  57. Re:First post by Anonymous Coward · · Score: 1

    Unfortunately yes. If you had overclocked your computer, you would have made it.

  58. Mac IIsi by GrahamCox · · Score: 1

    I upped a Mac IIsi from 20 to 25MHz once. It was well worthwhile. In fact, that's a 20% increase - I bet you'd be hard-pressed to get a reliable 20% improvement on anything these days.

    1. Re:Mac IIsi by Anonymous Coward · · Score: 0

      A 25% increase, even.

    2. Re:Mac IIsi by BLKMGK · · Score: 1

      An Intel 2600K from 3.6ghz to 5ghz isn't completely uncommon and that's 4 real cores plus 4 added hyperthreads. C2D were pushing a 30%+ overclock too and that's two generations older.

      --
      Build it, Drive it, Improve it! Hybridz.org
    3. Re:Mac IIsi by Mal-2 · · Score: 1

      I have an AMD 6-core Thuban 1090T Black Edition. 3.2 GHz out of the box, 4.0 stable easily attainable with a $30 aftermarket cooler and a voltage bump (from 1.375V to 1.475V). That's a 25% overclock out of a $169 CPU -- hardly what I'd call "hard-pressed". 3.8 with no voltage bump was stable on the stock cooler, though I didn't much care for the temperatures.

      --
      How is the Riemann zeta function like Trump rallies? Both have an endless number of trivial zeros.
  59. Re:No.... yes by mynicknamewasused · · Score: 0

    a mobile cpu switches between frequencies very fast on demand, so if you increase the top speed, it will spend less time in that state.., so for example we are comparing (40% full@1ghz 60% idle@250mhz ) vs (30%full@1.4ghz 70%idle@250mhz) . so yes, you can save power overclocking in the typical usage of a mobile device.

  60. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    200W is really hard (and expensive) to cool. Not many people want to deal with a lot of loud fans or a big liquid-cooling unit, or the expense of paying for same.

    The people that do can overclock the shit out of a more pedestrian CPU and take their own risks.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  61. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 0

    How old are you? Have you produced anything of value yet?

    Are you serious? For one thing, not everyone produces things of "value" (subjective) on the same computer that they overclock. Second of all, just because someone has different priorities than you doesn't mean that they're young.

  62. Re:overc locked by TaoPhoenix · · Score: 1, Insightful

    No actually you subconsciously slipped into the results of overclocking: something breaks if it's not perfect!

    --
    My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
  63. Insignificant by Anonymous Coward · · Score: 0

    Overclocking has become insignificant. It use to make a huge difference, but today, it is almost pointless, as you can't really notice an improvement in performance.

    This is mostly due to chip manufacturers like Intel and AMD focusing on putting more cores, than on increasing the speed as they use to. You can't end up with more cores by overclocking. This also has a negative effect on the CPU markets. To digress a little, It's just like Windows XP, people have a tendancy to stick with what they have for longer and the sales reflect that. People stayed with Windows XP because it was called XP and not 2001. Otherwise people may have switched earlier. People still tend to stick with XP because Vista and Windows 7 have some major bugs in them (along with some serious privacy concerns) and people want a reliable system, with security, and privacy they can control.

    Adding more coors is great, but the manufacturers have forgotten a basic rule; "If it's not broke, don't fix it". They had a good formula of increasing the speed of the processors, and everybody knew that 40 Mhz was twice as good as 20 Mhz. If that trend had continued we would be using 10 Ghz processors right now, and although it might actually be slower and less powerful than what we actually had, the result would still be more sales.

    Another mistake was made in the CPU industry and Intel lead the way on it. They made identifying the right processor for end users "simpler", "easier". Wrong wrong wrong wrong WRONG. It is more complicated. If you run multimedia, you get one type of processor. If you use it for gaming, you use another processor. But what happens if you do many things, and all of them with extreme needs? I do not acceopt taking a performance hit on one function only to have another one do better. Going from a 20 Mhz processor to a 40 Mhz processor was a no-brainer since it was double the speed no matter what you were doing. and it increased the demand for faster memory, and storage, etc.

    The new models from the big manufacturers like Intel and AMD have a huge negative impact on many industries.

    For now, yes overclocking is dead. But if the industry gets it's act together it will come back.

    Supply and demand uses the concept of more demand = more supply. But if there is no supply of new technology, all the demand in the world does not help, and that is where we are today.

    Where is my flying car? Where is my heads up display projected on the lense of my glasses? Why can't I buy a ticket to the moon?

    There are lots of demands, but the industrial world need to get their act together and start inovating instead of simply milking the same old lame junk.

  64. Data integrity is more important by Anonymous Coward · · Score: 0

    I don't overclock, nor underclock, because I want as correct data as possible. That's why I only use ECC memory and AMD.
    Back in the day when computer games actually did have steep hardware requirements I used to overclock everything to its limit. BSOD was quite common after a few hours of gaming.

  65. Re:Most people don't understand that it's a bad id by Anrego · · Score: 1

    It's a hobby like anything else..

    You don't see businesses overclocking their systems (at least I've never seen this.. ). It's something people do at home to get more enjoyment out of their purchase (which was likely the point of the purchase to begin with). Most people who are into extreme overclocking are primarily using their box for gaming.. not preparing year end financial reports for work. A little data corruption isn't the end of the world (and for the record, I've never seen that happen.. the system either works fine or is obviously unstable.. and I've _never_ seen a processor "wear out").

    These days I tend to limit myself to the safe limits of the chip, just overclocking enough to get what I paid for.. but I can see the appeal.

  66. I overclock my hard disk drives by Anonymous Coward · · Score: 0

    i put my drive up to a real motor that can spin them much faster. my access times are in the 9000.

  67. Re:yes, there is - RAID by swalve · · Score: 1

    Not necessarily. Depends on whether the RAID processing is fast enough to be able to take advantage of it. And whether the data bus can handle it. And whether the overhead of having larger stripe sizes makes up for the faster speed. (IE, if your disks have 512 byte sectors, to get one sector you have to read 512 bytes from each disk, concatenate them and then extract the needed data. A single sector read takes just as long.)

  68. Re:Overlocking was only ever a dick waving contest by JustNiz · · Score: 0

    I not even close to being an audiophile but even I can certainly tell the difference between CD and DVD audio of the same thing. (CD being significantly worse).
    I'm assuming its because of the lower relative sample rate of CDs but its irrelevant, because it shows that CD's aren't that great, for whatever reason.
    I think you're missing the point by claiming analog isn't that accurate so therefore can't sound as good. I don't think its about accuracy, as much as the way the sytstem deals with the lack of it.
    The tonal 'color' that using analog technology gives to the sound is more pleasing than CD, regardless of how accurately it reproduces the original signal.

  69. sorta off-topic, but still kinda related by Anonymous Coward · · Score: 0

    When you are single-threaded and CPU-bound, overclocking might be your only solution.

    As pointed out in this article by Herb Sutter in 2005, "The Free Lunch Is Over: A Fundamental Turn Toward Concurrency in Software" (http://www.gotw.ca/publications/concurrency-ddj.htm), the number of transistors has kept pace with Moore's Law, but clock frequency has not. We were on an upward climb until around 3.4 Ghz cores then we flattened-out. Overclocking is one way to get past this barrier until we have new tech. Another way is multi threaded programming. make -j10 ftw as someone said in this thread, already. lol. Or, nowadays, make -j24 wtf! Luckily gnu make is multithreaded, but after you've compiled, if your code is single-threaded then having 24 cores doesn't matter. By overclocking you can get more clock cycles from a single core.

    I'm sure this is all elementary to you guys, but I haven't heard anyone say that it really depends on the application. In some cases 32 bit is preferred to 64-bit, slower cpu is preferred to faster, more cores preferred to fewer, and vice versa. What do old ladies taste like? Depends. Intel is re-releasing pentiums. Did you hear about that?

    I think overclocking is a more valuable tool than ever, if you need to have extreme throughput for one thread; waiting 6 months for the industry to build you a faster core is not an option anymore. (But you should really be writing that program as multi-threaded, in the first place.)

    The gamers are the Tim Allens (tooltime) of CS, in my mind. Overclocking "just because" does seem to me to be a testosterone-fueled thing, and bragging rights are part of it. But there's nothing wrong with that. That's fun and drives the tech forward in new ways. However, I also agree with the guy who "compiles distros all day"; when your test cycles are 20 minutes between each test, every extra CPU cycle/second counts and you probably don't care where you get it: from the core or from more simultaneous threads.

    1. Re:sorta off-topic, but still kinda related by BLKMGK · · Score: 1

      Well except that today's 3.4GHZ CPU can run nearly 5ghz with proper cooling while still running all 4 cores+. The increase in cores IS nice but sadly compiling with -j4 on my Atom machines I run into issues with the code almost as often as not and have to compile single thread - it sucks! That's compiling XMBC BTW and it takes awhile. Now, encoding video? Hell's yes the more cores the better! I have seen huge increases in FPS overclocking when encoding video, it can save hours and hours. I wish x.264 was using the GPU....

      --
      Build it, Drive it, Improve it! Hybridz.org
  70. Re:Most people don't understand that it's a bad id by Khyber · · Score: 1

    "200W is really hard (and expensive) to cool"

    No, not really. I've got 300w LED arrays in a 30mm x 30mm package that are kept cool with nothing more complex than a copper heat sink that's been pressure-blasted with mesophase carbon pitch. Junction temps never get past +10C room temp..

    --
    Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  71. I still OC'd my i7 by cod3r_ · · Score: 0

    3.8ghz.. Water cooled and runs like a dream.. It will never die to that extent. Next i7 that is 4.0ghz will be OCable to 5ghz.. and that's what we'll do. Maybe at some points the speeds will be so fast we can't possibly notice the difference in performance, but we'll still do it. long live OCing

  72. It has long been a matter of personal style by MichaelCrawford · · Score: 1

    You would probably gripe at the fact that I prefer top-posting over bottom posting, as it lets me read the latest contribution to the thread.

    I also like to Capitalize Some but not all words for emphasis. I also use Bold Face and italics for the same reason.

    • If I want to drive home a particular point, I create a bulleted list with just one entry

    Or Maybe I Use A Blockquote In Which All The Words Are Capitalized.

    For some reason that completely escapes me, my colleagues at Kuro5hin all regard my writing style as some manner of symptom of mental illness. While I am without a doubt mentally ill, that's not reflected in my writing style. That has more to do with having read a lot of archaic writing during my education, and having come to the conclusion at all the text decoration is more expressive than the dull tedium that passes for typography these days.

    I Am Absolutely Serious.

    --
    Request your free CD of my piano music.
  73. The adventure just isn't there... by sco_robinso · · Score: 1

    I remember back in the hay-days of overclocking (which I would define as around 2001-2004), there was a huge sense of adventure about it all. It was still seen as somewhat dangerous, and there was a certain amount of lust about it all (geek lust, that is). Buying better coolers, tinkering around with multipliers and bus speeds, sometimes even having to make physical modifications to the chips and motherboards was the norm.

    Now, unless you're a hardcore overclocker (which even back in the hay days was a small percentage), it's a matter fo clicking a check box and maybe changing a setting in the BIOS (and by changing setting, I don't mean changing multipliers, I mean setting the overclock mode). There's really not much fun, lust, or adventure about it.

  74. No by flanders123 · · Score: 1

    I think a lot of the allure of overclocking is the challenge to see if it can be done.

    CAR ANALOGY ALERT!!!

    Its like squeezing an extra 2 hp out of your muscle car by modifying the exhaust or reprogramming the chip. Can you really notice the difference between 410 and 412 hp, and isn't 410 enough? No and Yes, but people cannot resist.

    / CAR ANALOGY

    People like to tinker and push tech to the limits, which probably leads to progress in said tech. Overclocking of any sort will never die.

  75. Mineral Oil Is An Effective Insulator by MichaelCrawford · · Score: 1

    Unlike water cooling, you don't need to isolate it from the circuitry, you could just immerse your whole PC in a tank of it, say a fish tank.

    It's also quite an effective feline laxative. That is, if you catch my drift.

    Sorry, it has been a long time. I don't have a link.

    --
    Request your free CD of my piano music.
    1. Re:Mineral Oil Is An Effective Insulator by rubycodez · · Score: 1

      and watch the oil creep up the cables over time and make a mess on the floor, watch the oil get dirty over time.....screw that mineral oil nonsense

  76. Duron 1.3Ghz by C_Kode · · Score: 2

    I haven't overclocked since I bought an 1.3Ghz AMD Duron around 2002. It didn't have much headroom though to OC. I think I could only get it to around 1.4Ghz and change.

    Lately I don't see much use in OCing. Chips are plenty fast enough for almost anything you throw at them today and you can buy faster chips relatively cheap in the next year or so when your current chip is becoming insufficient.

    Both the Intel i5-2500k and AMD Phenom II X6 1100T sit around $200. They aren't the fastest on the market, but at around $200 they are cheap any easily fast enough to handle anything you throw at them.

  77. Overclock to save cost by na1led · · Score: 1

    It use to be that Overclocking was done to save cost of upgrading your PC. I remember owning a Celeron 300 processor and was able to overlock to 450Mhz with little effort. This was a huge cost saving versus buy a Pentium II 450mhz. Today, it makes no sense to overclock since you can't get much more speed versus buying a new cpu or motherboard. A lot of software today is dependant on multithreading, GPU, and memory versus actual clock speeds. My recent cpu hack was unclocking a Phenom III (3 core) to (4 cores). This saved my some money and provided me the same product as purchasing a quadcore.

    --
    -- By all means let's be open-minded, but not so open-minded that our brains drop out.
  78. Re:Overlocking was only ever a dick waving contest by ewhenn · · Score: 2

    Not sure what you are on about spending $1000 for 25% more performance. I have a cheap $22 CM212+ cooler, that's a pretty far cry from $1000. The gains are absolutely worth it, I have a 2500K that has a stock speed of 3.3 GHz, it's overclocked to 4.5 GHz, or about a 36% increase.

    $22 for 36% more performance is absolutely worth it, maybe not for gaming now, but it's definitely useful for other tasks.

    Personally, I do a decent amount of encoding video files, and the speed increase is absolutely time saving. I encode roughly 5-6 ~10 min. 1080P videos into H264 a day. The overclocking saves me about 45-50 minutes a day in encoding time. That's nothing to sneeze at.

  79. There isn't much value in it by brainzach · · Score: 1

    Instead of buying expensive memory to overclock my RAM, I realized that I get more performance by buying more of the cheap stuff.

    Overclocking CPUs require buying expensive power supplies, coolers and cases which produces a negligible improvement in my day to day task. Modern CPUs are good enough for what I need the computer for. I am fine sacrificing the 1000 watt power supply and noisy fans to play games at medium quality settings.

    Fast CPU's and RAM are overrated anyways. A SSD hard drive has a much greater impact on my performance than overclocking ever will.

  80. Re:Overlocking was only ever a dick waving contest by Viol8 · · Score: 1

    That tonal color is essentially just a low pass filter effect. Turn down the treble on your amp a notch and you'll get the same effect with a CD.

  81. It's all about the percentiles. by Anonymous Coward · · Score: 0

    Before, gaining those last few MHz was a much greater percent than now. On top of that, most CPUs these days are already powerful enough to do almost all user uses.
    The only real reason to do it now is just to do it, which is a good enough reason if I've ever heard one.

  82. I laugh... by g0bshiTe · · Score: 1

    mostly when people say they are getting a Core i7 and are going to overclock it. I ask them why. Mind you if you are trying to squeeze every bit of life out of hardware before you absolutely have to upgrade then by all means, the worst case is you will have to upgrade a bit sooner. BTW I do have my current CPU oc'd it's only a Core 2 Duo E8400 currently oc'd to 4 ghz on air and very stable.

    --
    I am Bennett Haselton! I am Bennett Haselton!
    1. Re:I laugh... by BLKMGK · · Score: 1

      Video encoding
      Crypto cracking
      Compiling
      Rendering

      Just a few of the things that overclocking is good for. What might suit you doesn't necessarily suit everyone else.... Now imagine your current machine with 4cores, 4 more hyperthread pseudo cores, and a higher clock. A clock that BTW isn't 1:1 with what you have now so even a slower clocked i7 might be faster than what you have. Still so sure it makes no difference? If so then yeah you're right web surfing and reading email aren't boosted by overclocking.

      --
      Build it, Drive it, Improve it! Hybridz.org
  83. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    Inside a computer case, with other stuff creating heat as well and only so many places for the heat to go.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  84. Re:Overlocking was only ever a dick waving contest by Anonymous Coward · · Score: 0

    I disagree, at least the absolute statement you're making.

    10 years ago it was possible to overlock a 300 mhz celeron to 450 mhz with just a normal cooling unit and 100% stability. The performance gains were very substantial, saving hundreds of dollars. If you didn't have a lot of cash, this was a Big Deal. I'm not into the overclocking scene anymore, but I think those days of massive gains for little investment are over. So I agree with you in the sense that overclocking to save money died long ago.

  85. Re:Overlocking was only ever a dick waving contest by Megane · · Score: 2

    I not even close to being an audiophile but even I can certainly tell the difference between CD and DVD audio of the same thing. (CD being significantly worse).

    Say what? CD is 44Ksamples/sec 16-bit PCM stereo, or about 1500Kbits/sec. DVD stereo audio is usually a Dolby Digital bitstream of 192Kbits/sec, or maybe 384Kbits/sec for 5.1 audio, and usually from a 44Ksamples/sec source. And you say CD sounds worse? Perhaps what you are hearing is the extra channels from 5.1 audio. Or maybe your amp is processing the CD audio differently, like not mixing a center channel, or not doing Dolby Stereo, or maybe even trying to do Dolby Stereo when the source material wasn't recorded with it.

    --
    #naabhaprzrag, #sverubfr-000, #agi-fcbafberq, negvpyr[pynff*=' negvpyr-ary-'] { qvfcynl: abar !vzcbegnag; }
  86. Re:Overlocking was only ever a dick waving contest by Megane · · Score: 1

    I have a simple fix for you. Just turn the volume way up and wait for it to kill the rest of your hearing.

    --
    #naabhaprzrag, #sverubfr-000, #agi-fcbafberq, negvpyr[pynff*=' negvpyr-ary-'] { qvfcynl: abar !vzcbegnag; }
  87. Give CLang and LLVM a try by MichaelCrawford · · Score: 2

    I don't know how it compares to Visual Studio but they complete builds quicker and generates faster machine code than does GCC.

    The CLang command-line is mostly GCC compatible. The parser is larger Visual Studio-compatible.

    It is also Open Source under a BSD-style license.

    A friend gave a talk at Microsoft one day. Upon his return he told me why Windows was so slow. It turns out that all the OEMs - Dell, Gateway, HP and the like - donate hardware to Microsoft's coders so they can be certain that the next version of Windows will run cleanly on them.

    To encourage these coders to actually use the donated machines, they donate the very fastest hardware they make.

    I'm very strongly of the opinion that coders should use, as their regular desktop machine, the very slowest hardware that can possibly get the job done. To this day I write most of my code on an Early 2006 MacBook Pro. It has a Core Duo, not a Core 2 Duo, and so is not even 64-bit. It only has two gig of RAM and a tiny cache compared to the MacBook Pros that are available today.

    This has the effect on me that I can easily tell when my own code is slow. I don't need to use a profiler to know that.

    --
    Request your free CD of my piano music.
    1. Re:Give CLang and LLVM a try by oojah · · Score: 2

      I agree. I used to do most of my home development on a headless 600MHz VIA C3. It broke at the start of December so I'm on a 1.6GHz dual core Atom instead now, it's zippy! :)

      --
      Do you have any better hostages?
    2. Re:Give CLang and LLVM a try by jones_supa · · Score: 1

      Not a bad replacement. With both cores utilized, I'd say the Atom has 10x the speed of your previous CPU.

  88. Stopped long ago by SimplyGeek · · Score: 1

    I stopped overclocking years ago. Today's CPUs are just fine compared to what we had in the 90's and early 00's.

    The attitude nowadays is that we have shit to do, and if the hardware is fast enough, we move onto what it is we actually need the PC for. That's how it is with office desktops. You don't even need a high end machine to run Win 7, Office 2010, and a few proprietary business apps.

  89. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 2, Insightful

    Are you not from earth or some shit?

    He's from earth and knows what he's talking about. You obviously don't know what you're talking about. TDP is the OVERRIDING reason you don't see higher stock clocks out of intel products. 980Xs are expensive because Dell fucks up cooling? REALLY?!?! Go back to nursing your bong and quit posting crap about things you know nothing about.

  90. Things are so cheap, why bother? by Anonymous Coward · · Score: 0

    High end components are just so damn cheap these days. Overclocking was great when you could take a $200 CPU and make it perform like a high end $800 CPU. Nowadays the $800 CPU costs $250. Anyone concerned with performance is just going to spend the extra $50. Discounting inflated hard drive prices, a top of the line PC can be had for around $600 these days.

    Of course there are always some who want to push the limits, but the gain is simply not as great as it used to be.

  91. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 0

    200W is really hard (and expensive) to cool.

    When you dont know what you are talking about, dont talk! Seriously...

    There is a reason that gaming machines have 600W and higher power supplies.. and its not because the machine isnt drawing way the fuck over 200W, or has "hard" and "expensive" cooling solutions..

    Fuck man.. video cards pull as much as 300W these days... amazingly you think that Intel cant offer a cooler for a $1000 CPU?? Intels problem is that regardless of what heat sink and fan they throw on, the OEM's stick it in a shitty case with poor ventilation. Period. Thats the fucking problem.. dont imagine problems that don't exist.. people are clocking the latest CPU's at 4.5ghz on the stock coolers.. but not in shitty dell cases.

    --
    "His name was James Damore."
  92. it as always over by markhahn · · Score: 1

    overclocking was always a niche thing - it had some support from game-the-system types, but was mostly tweakers (yes, akin to meth.) it had an aspect of public self-abuse, sort of like getting a gang tatoo on your neck or face. and now, it's totally mainstream and vapid - the elements of display like leds on your fans and clear side panels are no more than a sort of meth-trailer-trash-chic.

  93. not over for me by Anonymous Coward · · Score: 0

    as long as the system runs stable under full load (and over, say 48h) i still benefit pretty much of an overclocked cpu when used for 3d- or videorendering

  94. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 0

    Plenty of people have 300W video cards. Those go inside computer cases... with other stuff creating heat...

    Please stop advertising your ignorance.

    --
    "His name was James Damore."
  95. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 0

    Inside a computer case, with other stuff creating heat as well and only so many places for the heat to go.

    There is only one place for the heat to go.. in that tiny little space called THE ENTIRE UNIVERSE OUTSIDE THE CASE.

    Plenty of people have 300W video cards.. those go inside the case... with other stuff creating heat... somehow the facts have eluded you. Talk about things that you dont know about much?

    --
    "His name was James Damore."
  96. Re:yes, there is - RAID by Anonymous Coward · · Score: 0

    Unless you actually logged out, your mods were undone. Simply selecting "Post Anonymously" is insufficient to protect the mods.

    This misfeature should be better documented.

    Not a misfeature. In total keeping with the spirit of the rule.

    I blame Theodore Roosevelt.

  97. Re:Most people don't understand that it's a bad id by Rockoon · · Score: 0

    980Xs are expensive because Dell fucks up cooling? REALLY?!?!

    Do you know the warranty request rates when they are in Dell machines? No? Thanks.

    --
    "His name was James Damore."
  98. Re:Overlocking was only ever a dick waving contest by VinylPusher · · Score: 1

    Pardon?

  99. No, there's more overclocking than ever! by RanceJustice · · Score: 1

    On the contrary, overclocking is rising. On the standard PC front, it used to require highest-grade components alone, difficult-to-balance cooling (I got my "hardcore" overclocking start during the 486 era - there were some heatsinks that were so heavy that they could actually damage the motherboard), a deep knowledge of settings, and the will to do something that could very well limit your stability or cause your system damage if you fouled up.

    Today many items come with a minor factory overclock, and many mid+ level components come with software (Asus is notably good here) that will allow you to overclock easily from within the OS with nearly no chance of damaging your system. Even if you go to the old-fashioned way for big gains, with most motherboards, especially OC-centric boards like Asus' Republic of Gamers (debatedly the best motherboards in existence for OCing and the enthusiast feature set) line and their Sabertooth (which comes in right under RoG boards) boards, have tons of settings to help you overclock. RoG boards have specially BIOS settings if you're going to be using LN2! AMD versions even come with tools to unlock a Phenom II or Athlon processor with limited cores - with the notable exception of Intel themselves, most hardware companies now encourage you to overclock safely and would rather provide value-added features to do so. Today if you foul up with your TIM contact and your temp raises higher, unless you've gone to a level of safety-removal that isn't necessary even for the highest "home" OCs, your board/chip will shut down way before any actual damage occurs.

    There are fewer "You must buy new hardware or overclock significantly to even play X" titles out, which is a good thing, but that doesn't mean people aren't overclocking. Enthusiasts who do it for the fun of it will always do so; they like to see big numbers on benchmarks and distributed-computing apps. For those who game or want more performance, they'll continue to do so because its simply a good value. In today's economic climate especially, overclocking is favored over buying ultra-expensive computing parts which have minimal games. Intel's recent pricing and performance line up for the new Sandy Bridge-E Socket 2011 processors is very telling in this regard, and I believe proves that overclocking is quite popular.

    3-4 years back, at the launch of the "Core iX series" Intel diverged with their socketing and created a two-tiered system on the desktop. Socket 1156 and the associated P55/H55 chipsets were the Mainstream platform, offering Core i3, i5 and the lowest i7-8xx series. This platform had dual channel RAM and a handful of PCI-E lanes etc... The Enthusiast platform however came on Socket 1366 and X58 chipset, holding an assortment of hyperthreaded Core i7 processors, the QuickPathInterconnect (QPI), TRI-channel RAM, more PCI-E 2.0 lanes, native support for both SLI and CrossFireX multiGPU solutions (which was a big deal as during the Core 2 era and prior, you typically had a board that could do one or the other, not necessarily both) and other great features. Power users gravitated towards 1366/X58 greatly because of overclocking potential and great value. At launch, the Core i7 920 basically had all the same cores/cache/HT as the $1000+ Extreme Edition 965/975; only the speed was different. The Core i7-920 overclocked phenomenally, reaching in excess of 3.6ghz for most enthusiasts with an after-market air cooler from their 2.6ghz start point; all without having to disable hyperthreading or any other processor features! I'm personally typing this on a i7-920 OCed to 4.0ghz on air - those with lucky chips or liquid-cooling systems could reach 4.3 stable with a minor voltage increase.

    This OCing performance and value made Socket 1366 popular far beyond Intel's original predictions that it would be limited primarily to workstation use; especially as prices dropped for tri-channel RAM kits and high quality X58 motherboards. Its performance overclocked easily eclipsed the mainstream

  100. Re:No, it simply doesn't provide the extras it use by tompaulco · · Score: 1

    I bought the i2600k CPU because it could be overclocked. I also bought the Asus Maximus IV Extreme because it makes overclocking easy and has a lot of built-in safety features to prevent stuff from blowing up. All that said, I have never actually overclocked the system. However, the fact that I could is worth the extra money to me. To me it is important that my system is not locked down or walled in. If I want to mess with it, I can, and if I don't want to, I don't have to. I don't like the vendor making that decision for me.

    --
    If you are not allowed to question your government then the government has answered your question.
  101. Hmmm by lightknight · · Score: 1

    These incredibly high speeds are wonderful, and have their place.

    My interests, though, are for sustainable high speeds. I want the equipment to overclock my CPU / GPU / RAM, and be able to keep it running at a high speed indefinitely. And that's part of the problem: these high speed trials require liquid nitrogen or helium, which can be rather expensive.

    --
    I am John Hurt.
    1. Re:Hmmm by BLKMGK · · Score: 1

      Agree, I see these guys doing max speed runs that aren't sustainable and I shrug. When you can hold that speed for days on end then I'm more interested. I'll take a near 50% overclock that's stable any day of the week and keep my RAM and peripheral buss speeds down where they can be stable.

      --
      Build it, Drive it, Improve it! Hybridz.org
  102. Yes, most of the enthusiasm is gone by Sloppy · · Score: 1

    13 years ago you were probably (though by no means certain) using a lower level language, so there was a vague feeling that from a software perspective, you had already done about all you could do, so tweaking the hardware was what remained.

    And the scalar speed of the hardware was pretty much the one thing to change.

    Nowdays you're always thinking about the software, and the hardware is presented to you as something with overflowing abundance that you're not taking full advantage of. You go to "I need threads there" or "where can I pipe/fork this?" or even less nerdy things like trying out a new compiler, before you worry about clock speed. It's not that clock speed doesn't matter sometimes, but it's just one of a great many things you have access to and some control over, so it's naturally going to be de-emphasized compared to the older days.

    Then from the implementation angle of overclocking, it's less work/thinking/hacking. You simply change some very well-labeled BIOS settings, which weren't always there with your 1990s computers. And you can literally buy water cooling systems, mass produced and already fitting your processor's package, off the shelf. Your case already has (or if it doesn't, you go buy one) holes and fans in the right places. It's all completely mainstream now, where you flick the switch someone else made, rather than having to come up with your own way to get it to work.

    It's not "hacking" anymore.

    --
    As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
  103. Re:we wouldn't need over clocking if our code were by Anonymous Coward · · Score: 0

    the only code that ever touches the metal, after all, is machine code.

    And, surprise surprise, it turns out that compilers write much better machine code than humans do--in most cases, they even optimize better than humans do.

    consider initializing a 2d array in a nested loop. if you increment columns in the inner loop, the memory cache will speed you up. but a dumb mistake could increment rows instead. in that case the cache actually slows your code down dramatically.

    ...for a given hardware representation of a 2-dimensional array (the one used by ANSI-C, so the most popular, but not the only one possible). For some representations, your "optimization" would actually cause the problem that you're trying to avoid.

    I'm sorry, but every time I see one of these complaints I feel like ripping my hair out. Premature optimization causes far more problems than it solves. There's a reason that all of the reasonable software development professionals tell you to optimize code for reading first--it turns out that only a tiny portion of your code actually matters for performance, and very few people (you're almost certainly not one of them) can tell what that code will be before they measure.

    Given a limited number of hours to spend working on a project, if I can spend X hours making my code more "optimal" with no visible benefit to the end user, or spend X hours providing a cool, innovative new feature, guess which one I'll choose? If you use the software, which one do you want me to choose--something you'll never know about, or something that will make your life better?

    Don't get me wrong, I think learning assembler and hardware architecture were important classes, but this nonsense about how programmers used to be real men, and real men optimize their code is getting in the way of producing more and better software engineers.

  104. Overclock VS core unlock by phorm · · Score: 1

    I've see a lot more traffic around core unlocking as opposed to overclocking. Most people I know don't need an extra 200Mhz, especially when most of the power for media comes from a GPU, but an extra core is oftimes still quite useful.

    As with the old days where vendors sold faster chips underclocked to meet demand, many sell chips with disabled cores. In some case it's a quad-core chip that had a less-than-perfect core, and thus is a three-core chip, but in others it's just that it was cheaper to make quad-core chips and disable a core for the three-core market...

  105. Re:Overclocking CPUs = overclocking GPUs? by Anonymous Coward · · Score: 0

    CPUs are no longer really interesting for number crunching, having been replaced by GPUs.

    GPU's are only good for data parallel number crunching.

  106. Overclocking is Jersey Shore vanity by Lime+Green+Bowler · · Score: 2

    My personal opinion is that overclocking does not buy you much, other than bragging rights. Sure you can get a few more FPS, or a few hundred extra MHz out of your CPU. But does that translate into anything usable? A false edge for gaming; false because there are so many other factors that can nullify that edge such as your connection parameters. Perhaps I'm too pragmatic, but then I don't watch Jersey Shore or Kardassians or fauxlebrity shows either.

    1. Re:Overclocking is Jersey Shore vanity by BLKMGK · · Score: 1

      How about hours saved on a video encode or 30 minutes on a large code compile? Render any animation? People who do this are often not just doing it to play games anymore, there's plenty of CPU intensive tasks out there other than games.

      --
      Build it, Drive it, Improve it! Hybridz.org
  107. Re:Most people don't understand that it's a bad id by tompaulco · · Score: 1

    200W is really hard (and expensive) to cool.
    Maybe so, but once you get up to 600W it gets much easier, apparently. I use the stock fans that came with my case and the CPU stays in the 60s Celsius running full tilt with a 6990 graphics card in the same case.

    --
    If you are not allowed to question your government then the government has answered your question.
  108. Overclocking is Out, Low-Power is In by eepok · · Score: 2

    I overclocked first computers (2000-2004). I bought a budget system in parts, put it together, got online, and learned that I could make my computer even faster with a little risk and careful effort.

    But then the prices of components began to fall and I stopped overclocking new rigs 2004. Why? Because a normal $30 heatsink was barely enough to keep some of the hotter processors cool without overclocking... and I was not willing to risk losing my processor for a few more FPS in Counter-Strike or whatever I was playing that month.

    Fast-forward to now, I still leave my main computer on 24/7, but as a career-person, I need to save more (house, retirement, vacations to placate the lady) and spend less on utilities. I also have less time to clean the dust out of computer cases that effectively had hoovers for cooling. So where I used to go for a balance of cost, heat, and overclockability, I now look at cost, heat, and power consumption. I now take pride in being able to comfortably play modern games (though not at max settings) on a rig supported by a 260 watt power supply. I have no guilt leaving that on overnight.

    Note: I never got into water-cooling. I never had the space or disposable income to mess around with the kit or the risk.

    1. Re:Overclocking is Out, Low-Power is In by Anonymous Coward · · Score: 0

      Revolutionary idea: How about you stop leaving the computer on overnight? That's the worst thing you can do if you're really worried about power consumption.

  109. Underclocking is big by Animats · · Score: 2

    Many industrial PCs are underclocked. They have more CPU power than they need, and they need more reliability and temperature range than the consumer manufacturers provide.

    The end of overclocking is coming anyway, because speed of light lag across the chip, rather than transistor switch time, is becoming the bottleneck. No amount of cooling will help with speed of light lag.

    1. Re:Underclocking is big by Anonymous Coward · · Score: 0

      No amount of cooling will help with speed of light lag.

      That's why everyone's switching to neutrinos these days.

  110. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 0

    Do you know the warranty request rates when they are in Dell machines? No? Thanks.

    WTF? The price of a 980X CPU has absolutely nothing to do with Dell warranty request rates.

    It's extremely rare for a modern CPU to fail because of cooling problems, whether it's in a Dell or some other OEM. Modern CPUs - including the 980X - throttle down/shut off when overheated; this occurs at a threshold well below what it would take to actually damage the chip. You can run an i7 without any heatsink at all in an unventilated box and still not damage the chip. Other bad things will certainly occur, but the CPU will be fine (unless done repeatedly over a long period of time).

    If Dell's shitty cooling causes a 980X to throttle, and the owner of the system returns it to Dell, the CPU doesn't get returned to intel because there's nothing wrong with it. But don't let this information stop you from continuing to grind your axe. Dell SUX!!!

  111. How about a youtube video on the subject by Anonymous Coward · · Score: 0

    http://www.youtube.com/watch?v=om7O0MFkmpw

  112. Low-Power is In by Sloppy · · Score: 2

    Here here. Low power is the new stupid (half kidding) thing to obsess over. My most-used home computer with a UI is an Atom. In 2010 when people were drooling over how great Sandy Bridge might be, and how much kickass-per-$ the X6 Phenoms offered, I was looking for an Athlon II 240e for my server to downgrade to (eventually finding, to my joy, a 610e for sale, so that I could finally pay $130(?) for the downgrade), just so I could say I had a 45W-TPD-but-still-reasonably-powerful-for-transcoding CPU. Not that doing such things makes any sense at all, from a "green" or money-savings perspective; it's all about low wattage being the new dicksize ruler.

    [Pompous English accent] "At your next dinner party, impress the all the wives by being the man with the smallest power supply."

    And that's the relatively non-nerdy approach. Serious dudes are getting all excited over stuff like Raspberry Pi, etc

    --
    As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
  113. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 0

    In years past those gains had real world significance, now its plain ego. I think that is the core of this discussion. Overclocking has moved almost completely into the realm of ego only.

  114. I still overclock! by BLKMGK · · Score: 2

    I still overclock and nearly every PC I've ever owned has been overclocked to include an 8088 clocked up with a radio crystal back in the day (not a great idea). I was playing with water cooling and Peltiers before you could buy ANY hardware for that off the shelf too. Cut down heatsinks, PVC caps, fountain pumps, and overseas sourced Peltiers made for some really quick computers for their time! Games were fast, looked great, and I ran RC5 cracking programs to use up idle cycles for years.

    Fast forward to the present. I still game but I am not quite into the really crazy high end stuff. I still use a PC for gaming almost exclusively. I no longer run programs in the background to eat up spare cycles and the cooling of my room thanks me for it. I AM running a water cooled CPU though using mostly off the shelf stuff that doesn't leak, my CPU is rock stable and not quite pushed to the edge. I upgraded my computer in the not so distant past for more speed and I'm pondering doing it again to the later SandyBridge architecture from my older i7 920 (4.1ghz). I'm also looking at the new 6core CPUs that have come out but they strip H.264 instructions apparently. :-(

    Why? Well it certainly isn't gaming since right now games seems woefully poor at using multiple cores! Now I have another "hobby" and that is compressing video. I buy BluRay, rip them, and put them on my personal server for viewing on efficient Atom powered STBs (overclocked though lol). When I was doing this with a C2D running in the mid 3-4ghz range some movies like Watchmen took 8 hours or more to encode with my high settings. Now I can do a movie in 2 hours or less while still having CPU available to do other things. If I move to the more efficient CPUs produced now, and especially if x.264 supports their ENcoding instructions one day, my times will drop again as I should be able to hit close to 5ghz. At that point I'll either encode with higher settings or just enjoy that it's as fast as it's going to get. I boot from an SSD so that's quick enough. My video card is a fairly pedestrian GTX275 which might get a bump too, I'm not sure.

    I have tinkered with using the GPU to encode as well. Right now my CPU alone can keep up with encoding on my GPU alone but mixing my GPU and CPU together (I found ONE package doing that and it wasn't x.264) was noticeably faster but severely limited my encoding options so I've stuck to CPU brute force. I'm hoping that with CUDA being open sourced more programs will begin using the GPU too.

    I've processed A LOT of video and I do video for friends too sometimes. Being able to run tons of apps, lots of browser windows, and generally not care too much about what is and isn't running is a side benefit. I may try BF3 out but doubt it'll be so much better than UT2K4 that I'll be sold as it will likely be exponentially more difficult to play. I'd love to find more things to do with the CPU power I have and I do try to use power wisely. My server(s) are actually underclocked and sleep their drives when not being accessed, my video front-ends draw less than 15watts apiece, the PSU in this box is Silver rated and under 650watts. Rendering or compiling code would be fun but I am neither developer nor artist. Those of us who are could certainly find value in overclocking! I know a certain Apple guy who was pretty butt hurt his 8 core powerhouse costing quite a bit more than my computer couldn't encode video as quickly when he challenged me :-)

    P.S. Yeah, I tinker with cars too, it's fun. I also laugh at those who talk about "shortened CPU lives" - get a clue. I have had exactly ONE CPU die and that was within the first 24 hours - warranty replaced. I've had overclocked CPU go for 5 years or more being passed down with no issue. This 920 has seen temps as high as 90C under full load for hours at a time when I had voltages too high and it's still ticking fine. My current peak is recorded as 75C. If you REALLY want to drop some heat water cool the video card, sadly these water blocks tend to be pretty custom and I don't do it since an upgrade on the video means a costly new block.

    --
    Build it, Drive it, Improve it! Hybridz.org
  115. Distinctly I remember it was in the bleak December by silverspell · · Score: 2

    Checking the usernames on this subthread, it would appear there is a conspiracy of Ravens in here!

    Yes, I've been keeping count too. This is a real statistical outlier, even for December.

    (Mind you, I've seen fewer ravens on one thread -- but never more.)

  116. A rule of thumb. by Anonymous Coward · · Score: 0

    When I was making a decision to upgrade my computer I came upon this observation. It goes "to really notice any performance you have to upgrade your system to be a minimum of twice as powerful." The assumption is that a lot of the performance increase is gobbled up by the system. You see, user input is put near the bottom of the CPU time hierarchy. If this is true, then going from 2.66 MHz to 3.15 Mhz (an 18% increase) is not enough of an increase to be noticeable.

    1. Re:A rule of thumb. by Anonymous Coward · · Score: 0

      Measuring your CPU speed in MHz is not going to get you anything meaningful comparing any two CPUs made in the past 5-10 years. Increasing clocks speeds was a relatively easy way to improve performance until around 3GHz when, among other issues, the speed of light limits clock speeds from increasing much more. Chip designers have been focusing on other ways to improve performance, so the megahertz myth is even less true today than it was when we actually commonly measured chip speeds in megahertz.

  117. Worst UI ever by Anonymous Coward · · Score: 0

    "You're behaving like you have no motor control. Would you like me to help you, spazz?"

    Fucking stupidest thing ever.

  118. Maybe it never was? by pepty · · Score: 1
    The historic way people benchmarked their performance for gaming was the maximum/average frame rate for the game. This was great for hardware blogs and the overclocking hobby: new hardware and tweaks happily produced longer bars in the barcharts (see "overcock" comment upthread), which in turn created advertising dollars for the blogs and buzz for the overclocking community. Who can argue that 144 isn't better than 130 FPS?

    But the metric that matters for gaming is a lot closer to the minimum framerate, or even better "the number of seconds that the frame rate was below X". These are measures that come a lot closer to telling you whether a setup will be annoyingly or unplayably slow for a given game. They also tend to flatten the differences seen between all but the most drastic changes in hardware and overclocking, which is not good news for hardware blogs or the overclocking community - who cares if your setup has a 10% difference in average FPS if still bogs down just as often? Those metrics (finally!) become common over the past few years, but were hard or impossible to find ten or fifteen years ago. How much of the benefits seen during the "golden age" of overclocking would disappear if appropriate metrics were used instead?

  119. Re:Distinctly I remember it was in the bleak Decem by Anonymous Coward · · Score: 0

    This is a real statistical outlier, even for December.

    True... For December, I would expect at least one white raven. Winter is coming.

  120. Re:No, it simply doesn't provide the extras it use by SteeldrivingJon · · Score: 1

    Maybe people should get into retro-overclocking. How fast can you get that NeXTStation to run? Or an SGI Indy?

    --
    September 2011: Looking for Cocoa/iOS work in Boston area Cocoa Programmer Quincy, MA
  121. Oh no... by Simulant · · Score: 1

    What's a high testosterone geek to do now?

    I tried overclocking but did not fined the performance gain worth the effort.

  122. Re:Overlocking was only ever a dick waving contest by bemymonkey · · Score: 1

    http://en.wikipedia.org/wiki/DVD-Audio

    Not the same thing as a video DVD.

  123. CPU not important, video unstable by bky1701 · · Score: 1

    Most games today loaf on CPU, and really, games are the only place most people care about speed that intently. The real benefits come from overclocking the video card, but those are clocked very close to their maximum safe speeds these days. When it's not an uncommon thing for factory clocked cards to begin overheating, there just isn't much room to work in.

  124. Phones? Heresy I say by triffid_98 · · Score: 1

    First off, for mobile devices this is a bit of a mixed bag, so lets talk about desktops. We all still have them, if you're a gamer you have a pretty nice one. Power efficiency is not something you are at all concerned about, the PSU in your rig could comfortably power a submarine.

    Some of us would like to buy a $100 CPU and clock it up to compete with a $300 CPU. This used to be quite easy to do, and at that point Intel had no competition, so it was an $800 CPU, and games were quite often CPU bound.

    Modern games are generally not CPU bound, they are GPU/memory bound. The good news is that you can overclock GPUs just like you can CPUs. The bad news is that you're probably only looking at a 15% overclock without losing stability and you still won't have all of the extra pipelines/shaders of the fancier card.

    So...is it still practical? I still do it because I'm a tight bastard and I still enjoy gaming/fast compiles. It helps, but the rewards for doing so aren't nearly what they used to be, and sane people are far more concerned with a reliable overclock than a +2Ghz e-peen that crashes after 5 minutes.

  125. Overclocking for performance makes no sense by pclminion · · Score: 0

    I won't make the mistake of criticizing those who overclock. I've done it myself, with the famous "Celery" processors, SMP dual Celerons clocked to 450 megahertz. It was fun!

    But if your goal is overall performance improvement, it makes hardly any sense at all. You're taking basically the fastest component in the system, and making it faster. Anybody who knows how to optimize anything should know that you optimize by speeding up the SLOWEST component first. For most machines, that's going to be the disks. Once the disks are as fast as possible, the next slowest component is the RAM. Increase the amount of it, increase the speed of it, or both. Then I'd start looking at the motherboard. Only after all that was taken care of would I start worrying about the CPU. And for that, cache is going to be far more important than clock rate.

    An overclocked CPU just spends more cycles and power, waiting for the same amount of time it would have before, for data from the glacially slow RAM.

  126. Over-clocking is only for incompetents by gweihir · · Score: 0

    Over-clocking removes safety-margins and reduces durability. It is basically for people with big egos and small minds. And it does not even give noticeable speed advantages in most cases.

    The basic misunderstanding of the over-clockers is that they think they get something equal but faster for the same price. They do not. What they get is a crappy imitation of the faster version, where they had a solid version of the normal speed device before. Stupid.

    --
    Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
  127. Re:Most people don't understand that it's a bad id by Khyber · · Score: 1

    That carbon pitch is so efficient at dissipating heat that even gentle airflow provides enough heat removal.

    My 300w array is in something MUCH smaller than a computer case, again with only so many places for heat to go and other stuff creating heat (power drivers.)

    --
    Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  128. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    Och aye? You might want to look for a job at Intel. They couldn't find a good way to deal with 150W or so from the late Pentium 4s, so your genius should find a well-paying job there.

    Or, you know, you might be full of shit.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  129. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    You seem to have a serious case of the gays for me, so let's save you some time: I'm happily heterosexual.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  130. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    I'll tell you the same as the other guy, but with less snark since you're not being a dick: if you can do a better job at cooling ~150W than Intel, by all means apply for a job there.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  131. Re:Most people don't understand that it's a bad id by Nimey · · Score: 1

    "full tilt"

    Don't know about you, but I don't think most people want to hear fans running at full tilt.

    You too are invited to get a job at Intel. Or, you know, maybe it's not so simple as you think to keep a high-wattage CPU cool.

    --
    Hail Eris, full of mischief...

    E pluribus sanguinem
  132. Re:Most people don't understand that it's a bad id by adolf · · Score: 1

    Here's what I know:

    If I'm Dell, and I buy a lot of 1,000 CPUs, I take them in OEM packaging without a warranty from the manufacturer. Why? Simply because it's cheaper that way, from start to finish.

    Meanwhile, Intel has a lovely little FAQ about their processor warranties that you can amuse yourself with once get done imagining things.

  133. No by Anonymous Coward · · Score: 0

    No

  134. Manufacturers got on board. by Mal-2 · · Score: 1

    Once upon a time, manufacturers didn't care if you overclocked unless it was a warranty issue. Then they started to care as their chips were passed off as a higher grade and failed early, which they rightfully saw as damaging their reputations, so they started locking multipliers. Later, they saw a market at the top end for enthusiasts who weren't doing it to scam anyone, just get the most out of what they bought -- which resulted in things like the AMD "Black Edition" and Intel's "Extreme". All these really are is chips that haven't been speed binned because there isn't enough of a target market to justify the expense of testing them that hard.

    THIS is why old-school overclocking on the desktop is dead. That, and the fact that laptops are usually designed with barely adequate cooling, so you really don't want to throw any more heat at them than they are already emitting.

    --
    How is the Riemann zeta function like Trump rallies? Both have an endless number of trivial zeros.
  135. It's just physics... by Anonymous Coward · · Score: 0

    Numerous issues with over clocking arise from the physics perspective. These relate to the standard methods for over clocking being both over clocking, but also "over voltaging".

    1. Most current processors are operating on their edge of their lifetimes for current room temperature accelerated life. Current lifetimes for MOSFETs are targeting 10-15 years right now. This is in radical contrast to lifetimes at RT that were typical 30 years ago of 1000-10,000 years, depending. This all comes down to shrinking per Moore's Law: lifetimes monotonically decline with decreasing size. We will likely hit device lifetimes as short as 1-5 years in the not too distant future (10 years of more Moore's Law if we get that far). The point of all of this is that over voltaging now pops the transistors right into the accelerating stress regime typically used for rapid accelerated life testing - the margins of stress and lifetimes are very small now.

    2. There are physical limits to how much you can over clock based on device physics. Most of the "challenges" of Moore's Law today come from bumping into these limits. Each recent radical process change in ICs in the last decade have all been about overcoming these limits that simply scaling dimension, voltage or clock rates can not fix. This includes Low-k inter metal dielectrics (transmission line capacitances), Copper metal interconnects (transmission line and gate resistances), High-k gates stacks (scaling induced gate tunneling), and High-mu/HEMT on silicon (channel mobility limits), the "Next Big Thing" after high-k. Because of these, just bumping up clocks or voltage in typical over clock fashion doesn't accomplish much in terms of performance enhancement.

  136. Better ways to boost performance by Anonymous Coward · · Score: 0

    In general, home computers far surpassed what most people need in terms of computing power a long time ago without any sort of overclocking. So why risk making one's system unstable or shorten its lifespan to get more of something that you don't need?

    In terms of games, which are the primary application that stress a modern computer at all, they are generally long on graphics processing (and typically short on content, imho) and hence stress the graphics card far more than the CPU. Even there, it doesn't cost much to get a graphics card these days that will work just fine with modern games without overclocking.

    The other interesting development in recent years has been that as chip designs hit thermal and power limits, the designs changed from vertical (higher frequencies) to horizontal (more cores). The end result has been that applications realize more performance by being design to exploit parallelism. Sure, overclocking by 10% yields benefits, but going from 4 to 6 cores is a much larger benefit.

    One other observation - for most modern computers and operating systems, beefing up the storage system often does more to improve performance. Replace that HDD if your laptop with a SSD and suddenly the system is significantly better. If you are nuts (as I am) get a higher end RAID controller with multiple SATA3 HDDs and a SATA3 SSD that is used as a cache for the primary array, and you are also off to the races in terms of everyday performance.

  137. Re:Distinctly I remember it was in the bleak Decem by c++0xFF · · Score: 1

    Mind you, I've seen fewer ravens on one thread -- but never more.

    Oh, how I hope some mods with a sense of humor read this!

  138. Quantum Carelessness by hicksw · · Score: 1

    What is the ground state of caring?

  139. Re:Most people don't understand that it's a bad id by tompaulco · · Score: 1

    Don't know about you, but I don't think most people want to hear fans running at full tilt.
    The fans aren't running full tilt, the CPU and the GPU are. But the fans are able to keep them sufficiently cool while not running full tilt. My fans are temperature controlled and the environment never gets hot enough for them to spool all the way up, even when I am running bitcoin mining or Flight Simulator.
    In my system, the GPU fan is the noisiest item. The case fans are relatively quiet (because they are bigger). GPU cooling is severely limited by case design. Basically all they can do is a squirrelcage type approach. it would probably be better to just have a big fan on the back plane blowing across the GPUs, but it is hard to suddenly change case design, and until that changes GPU designers have to assume there is no external backplane cooling.

    --
    If you are not allowed to question your government then the government has answered your question.
  140. I disagree by alexo · · Score: 1

    Development should be done on the most powerful hardware possible.
    Testing should be done on the least powerful that's still practical.

  141. gf and i used to overclock, by Anonymous Coward · · Score: 0

    but all we did was produce suds. now we focus on quality

  142. Re:Most people don't understand that it's a bad id by Khyber · · Score: 1

    I have access to materials with over 1,000 w/mK thermal conductivity. Intel does not.

    And I'm quite comfortable making about $15,000 USD daily. I'd rather not work for Intel. They rejected my stuff before, they pay the price for such rejection by being disallowed access to my technology.

    --
    Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  143. I had the world's fastest by uninformedLuddite · · Score: 1

    TNT2-M64 for a few days. I don't even bother to try and OC anything anymore as i don't need the performance increases.

    --
    The new right fascists are bilingual. They speak English and Bullshit.
  144. Re:Most people don't understand that it's a bad id by Anonymous Coward · · Score: 0

    So you have access to glass, then? I think Intel has plenty.

  145. Re:Most people don't understand that it's a bad id by Khyber · · Score: 1

    Glass has poor thermal conductivity. I don't think you even know what you're talking about.

    --
    Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  146. Re:Overlocking was only ever a dick waving contest by Megane · · Score: 1

    FWIW, "DVD audio" != "DVD-Audio"

    --
    #naabhaprzrag, #sverubfr-000, #agi-fcbafberq, negvpyr[pynff*=' negvpyr-ary-'] { qvfcynl: abar !vzcbegnag; }