Slashdot Mirror


Skylake Breaks 7GHz In Intel Overclocking World Record (hothardware.com)

MojoKid writes: Intel's latest generation of processors built on the Skylake architecture are efficient as well as seriously fast. The flagship, Core i7-6700K, is an interesting chip as it's clocked at a base 4GHz, and can peak at 4.2GHz with Turbo Boost. Of course, as fast as the 6700K is, overclocking can always help take things to the next level, or at least temporarily explore future potential. In Chi-Kui Lam's case, he did just that, and managed to break a world record for Intel processors along the way. Equipped with an ASRock motherboard, G.SKILL memory, and a beefy 1.3KW Antec power supply — not to mention liquid nitrogen — Lam managed to break through the 7GHz barrier to settle in at 7025.66MHz. A CPU-Z screenshot shows us that all cores but one were disabled — something traditionally done to improve the chances of reaching such high clock speeds.

85 comments

  1. About 4 times less performance than without OCing by Nikademus · · Score: 5, Insightful

    This is a very fun experiment as it lets you play with liquid nitrogen.
    However, the CPU has now only 1 core instead of 8 and only about 1.6 times the clock frequency. This means a huge decrease in performance...

    --
    I gave up with the idea of an useful sig...
  2. Nope. by chris200x9 · · Score: 1

    The skylake broke the record for OCing a skylake CPU, big whoop.

    1. Re:Nope. by chris200x9 · · Score: 3, Informative

      Forgot link http://www.engadget.com/2007/0... so TFA is wrong, it doesn't even beat records for and intel processor it literally is just a new skylake record.

    2. Re:Nope. by Anonymous Coward · · Score: 0

      I know right? Let's blame timothy for posting shit again.

    3. Re:Nope. by Anonymous Coward · · Score: 0

      "for and"?

    4. Re:Nope. by Anonymous Coward · · Score: 0

      timothy just came out of the closet as an intel fanboi writing that headline.. this is the "intel only single core skylake" record.. an amd chip beats it by far on clockspeed *and* number of cores enabled at that speed.. but does tfs mention that at all?

    5. Re:Nope. by JazzLad · · Score: 1

      Timothy is just a bot (running on Intel archetecture, I'm sure) and has been for a long time.

      --
      "If you have nothing to hide, you have nothing to fear." - Every fascist, ever
    6. Re:Nope. by requerdanos · · Score: 1

      Well played, sir.

  3. Re:About 4 times less performance than without OCi by Anonymous Coward · · Score: 0

    Awesome for single-thread performance. As long as the other LLC slices are still caching for that last core running.

  4. Twice the speed... by Anonymous Coward · · Score: 0

    ... and one eighth of the total performance. Nice try.

    1. Re:Twice the speed... by Sloppy · · Score: 1

      Maybe it's a dedicated Dwarf Fortress machine.

      --
      As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
    2. Re:Twice the speed... by binarylarry · · Score: 2

      Fastest python box in the world.

      --
      Mod me down, my New Earth Global Warmingist friends!
  5. the true purpose by Anonymous Coward · · Score: 0

    What a useless waste of time.

    Admit it Chi-Kui, the only reason you did this is so you could play with liquid nitrogen. **SAY IT!!**

  6. Barrier? by AmiMoJo · · Score: 2

    What is this 7GHz barrier? Is it like the sound barrier or something for CPUs?

    --
    const int one = 65536; (Silvermoon, Texture.cs)
    SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    1. Re:Barrier? by JargonScott · · Score: 1

      It's like deja vu. The calculations travel back in time, so they have to do it all over again. It's like deja vu. The calculations travel back in time.

      --
      Nuke Gay Whales for Jesus.
    2. Re:Barrier? by Joe_Dragon · · Score: 3, Funny

      Need to get it to 88Ghz

    3. Re:Barrier? by tibit · · Score: 1

      That's the clock multiplier limit on that chip. You can't seem to be able to set it any higher...

      --
      A successful API design takes a mixture of software design and pedagogy.
    4. Re:Barrier? by __aaclcg7560 · · Score: 1

      A practical limit for silicon-based CPUs. I've been told that military uses a different semiconductor material to run CPUs at 100GHz at a much higher temperatures.

    5. Re:Barrier? by Anonymous Coward · · Score: 1

      If only there was a power supply that could provide 1.21 gigawatts.

    6. Re:Barrier? by U2xhc2hkb3QgU3Vja3M · · Score: 1

      It's the same as the 90 mph wind barrier for buses.

    7. Re:Barrier? by Applehu+Akbar · · Score: 1

      We do seem to have been stuck at a max of 4G processor speed for a number of years. Has Moore's law been declared dead?

    8. Re:Barrier? by fnj · · Score: 2

      What the hell does Moore's law have to do with GHz? All it is, is an observation on the growth over time of the number of transistors possible to build into a dense IC. As far as I can tell, it is still operative.

    9. Re:Barrier? by MobileTatsu-NJG · · Score: 1

      Need to get it to 88Ghz

      Jiga, not Giga!

      --

      "I like to lick butts!" by MobileTatsu-NJG (#32700246) (Score:5, Informative)

    10. Re:Barrier? by TeknoHog · · Score: 2

      A practical limit for silicon-based CPUs. I've been told that military uses a different semiconductor material to run CPUs at 100GHz at a much higher temperatures.

      I'm not sure they are CPUs in the same sense. You can easily find simpler circuits that operate at such frequencies, e.g. microwave amplifiers, but a modern CPU involves much more than the raw switching speed of transistors. Keeping the core in sync with itself will be harder with a wavelength of 3 mm (This would be for 100 GHz in vacuum, in a solid it would be even less).

      --
      Escher was the first MC and Giger invented the HR department.
    11. Re:Barrier? by buchner.johannes · · Score: 1

      https://www.comsol.com/blogs/h...

      Essentially the answer is transistor noise, then transistor count, and heat. (Not the Speed of Light).

      Because power increases with (clock speed)^3, but executing speed only increases by (clock speed)^1, people go for parallel processors.

      --
      NB: The message above might reflect my opinion right now, but not necessarily tomorrow or next year.
    12. Re:Barrier? by bobbied · · Score: 1

      We do seem to have been stuck at a max of 4G processor speed for a number of years. Has Moore's law been declared dead?

      No, and neither is Moore... He's 87. I don't think he will live long enough for his law to be declared dead..

      --
      "File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
    13. Re:Barrier? by Anonymous Coward · · Score: 0

      Jigga-Jigga Waaaattt

    14. Re:Barrier? by Anonymous Coward · · Score: 0

      Barrier? AMD did 8.429Ghz years ago.

    15. Re:Barrier? by michaelmalak · · Score: 1

      0.3mm signal path per cycle (assuming 5:1 settling time).

  7. Not the overclocking record by hajile · · Score: 5, Informative

    I believe the official Guinness record is 8.429GHz on an AMD pre-release bulldozer in 2011. Another record was set at 8.723GHz on an AMD FX-8370 in 2013, but I don't recall it being "official".

    http://www.tomshardware.com/news/bulldozer-amd-overclock-guinness-record,13431.html

    http://www.theregister.co.uk/2014/09/02/amd_fx_series/

    1. Re:Not the overclocking record by Anonymous Coward · · Score: 0

      Net f-ing title says it is about Intel records...

    2. Re:Not the overclocking record by Anonymous Coward · · Score: 0

      But a stock i3 is more powerful...blah blah blah. This is what Intel fanboys actually believe.

    3. Re:Not the overclocking record by Xinef+Jyinaer · · Score: 1

      It does say 'Intel Overclocking World Record'. Which is just silly. One can assume that Intel will always hold the Intel Overclocking World Record.

      --
      Some days I just get bored and Troll post all the memes I can think of...
  8. 4V Core Voltage by Virtucon · · Score: 1

    Yeah that's feasible for everyday use with all but one core disabled. Not. This is a plateau of unusable.

    I'm curious to see how it performs with all cores enabled at that multiplier/voltage?

    --
    Harrison's Postulate - "For every action there is an equal and opposite criticism"
    1. Re:4V Core Voltage by Anonymous Coward · · Score: 0

      Yeah that's feasible for everyday use with all but one core disabled. Not. This is a plateau of unusable.

      Uhm, don't you think the bigger problem is that liquid nitrogen isn't feasible for everyday use...

      I'm curious to see how it performs with all cores enabled at that multiplier/voltage?

      It most likely doesn't run. Otherwise they would have enabled all cores.

    2. Re:4V Core Voltage by hippo · · Score: 4, Funny

      Very brightly

  9. Apparently AMD still holds the record though by The-Ixian · · Score: 3, Informative

    As impressive as 7GHz is, it's not the highest frequency we've ever seen. Over at HWBOT, we can see that the overall world record belongs to The Stilt, who pushed an AMD FX-8370 to a ridiculous 8,722.78MHz

    --
    My eyes reflect the stars and a smile lights up my face.
    1. Re:Apparently AMD still holds the record though by Anonymous Coward · · Score: 0

      Except that this GHz race is completely bogus. It doesn't make sense to compare AMD chips to Intel based on clock speed alone since each chip is able to do different amounts of work per clock. Here's an old article about it, though I'd wager it's still at least partially true.

      AMD vs AMD, Intel vs Intel. Unless they can demonstrate some amount of work being done, clock speed is meaningless.

    2. Re:Apparently AMD still holds the record though by slashmydots · · Score: 1

      I was about to quote the exact same thing. What world record? What 7GHz barrier? AMD CRUSHED that a long time ago.

    3. Re:Apparently AMD still holds the record though by Anonymous Coward · · Score: 0

      Title: ... In Intel Overclocking World Record

  10. Windozs by Anonymous Coward · · Score: 0

    Windows 10 still took 20 minutes to boot and send it boot time telemetry.

    CAP === 'irrigate'

    1. Re: Windozs by Anonymous Coward · · Score: 0

      Systemd boots systems 2x faster. Systemd, it's what Linux has and Windows computers crave.

    2. Re: Windozs by sexconker · · Score: 1

      systemd has electrolytes?

    3. Re: Windozs by Anonymous Coward · · Score: 0

      systemd has electrolytes?

      It's got what processors crave!

  11. The Intel 10GHz prediction by Gr8Apes · · Score: 4, Interesting

    In 2000, Intel predicted they'd have 10GHz chips by 2011. Here it is, 2016, and they've been going backwards since around 2006 with a peak roughly 3.8GHz production chip.

    --
    The cesspool just got a check and balance.
    1. Re:The Intel 10GHz prediction by Anonymous Coward · · Score: 0

      They predicted that only in a desperate attempt to validate NetBurst as the architecture of the future, and at a time when Ghz was considered a proxy to speed. AMD showed them a better way, intel followed and ~ 4G has been the peak since.

    2. Re:The Intel 10GHz prediction by Anonymous Coward · · Score: 0

      Back then Intel assumed they could move to Gallium-Arsenic chips such as the military has been using for 20 years. Turns out that it is costly, dangerous to the workers, and not particularly power efficient. However, special military users have overclocked CPU chips for the past 10 years by immersing them in a bath of liquid nitrogen. There is even a liquid nitrogen extractor device, about the size of a milk carton, that can be inserted in the machine with the CPU chip on the bottom. Takes 30 minutes to create a bath of liquid over the chip. Talk about a heat sink!

    3. Re:The Intel 10GHz prediction by Anonymous Coward · · Score: 0

      Yes, because GHz is what's important. Hey everyone, it's 1994 again and Gr8Apes is hosting a totally relevant conversation about today.. in 1994!

    4. Re:The Intel 10GHz prediction by Anonymous Coward · · Score: 0

      What was the OP again?

    5. Re:The Intel 10GHz prediction by Gr8Apes · · Score: 1

      AMD did indeed have a better way, but at that point and time, there were numerous other better architectures that were faster and more efficient.

      --
      The cesspool just got a check and balance.
    6. Re:The Intel 10GHz prediction by Raenex · · Score: 1

      AMD showed them a better way, intel followed and ~ 4G has been the peak since.

      If CPU speeds had kept on doubling along the lines of Moore's Law like they used to, it would have been much preferred and a glorious thing. Alas, the party came to an end, and both Intel and AMD were forced to go down the multicore path.

  12. In other news... by RogueWarrior65 · · Score: 1

    Overclocked processor pops bag of popcorn in next room.

    1. Re:In other news... by CaptnCrud · · Score: 1

      I read that as "poops" bag of popcorn, which I initially found much more humorous.

  13. Overclocking too far... by Anonymous Coward · · Score: 0

    Back when I had a Celeron single core, I used to overclock it up to 3Ghz. The only problem is that anything above 10% OC would kill the audio. So what use is this if it kills something else on your MB? It might be fine for Boinc projects, but for everyday use nothing gained, a lot to loose.

  14. Re:About 4 times less performance than without OCi by kimvette · · Score: 1

    "4 times less performance" than what?
    I HATE that wording on tech and science sites.
    1/4 the performance is the correct wording, unless comparing to a difference.

    --
    The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
  15. Re:About 4 times less performance than without OCi by Junta · · Score: 1

    It's actually only 4 cores, so the super fast overclocking isn't as bad as you say, it's almost half as fast as pre-overclock.

    --
    XML is like violence. If it doesn't solve the problem, use more.
  16. 6700K by Anonymous Coward · · Score: 0

    Are Intel chips named after the temperature in kelvins that they reach without a heat sink?

    1. Re:6700K by Anonymous Coward · · Score: 0

      No, silly, at 6700K the processor would vaporize.

      However, 6700K *is* the temperature of the blackbody radiation emitted by the heat glow of Intel CPUs.

  17. Anybody overclock the IBM Power 8 cpu? by Anonymous Coward · · Score: 0

    Wonder how high it would go. The base clock operates at 5GHz... and it can be overclocked.

  18. What we need is doable overclocking that is by Anonymous Coward · · Score: 0

    worth doing. Give me a cooler than can give me 1Ghz more for about $100 and I am good.

  19. They had a 10GHz ALU back in 2002 by zerofoo · · Score: 2

    http://www.anandtech.com/show/...

    Intel figured out that high clock speeds bring all kinds of other problems like clock propagation delays across the chip and high power consumption.

    The mobile computing revolution simply did not warrant that kind of clock speed - so here we are at around 4GHz.

    1. Re:They had a 10GHz ALU back in 2002 by Gr8Apes · · Score: 1

      Intel had to be beaten in the x86 arena for that to matter. There were all sorts of better designs around at that time. It kind of goes back to Intel not really being all that good of a chip nor innovator. They were taking lessons from their pal in announcing vapor to keep people from jumping ship. Sadly, it worked again.

      --
      The cesspool just got a check and balance.
  20. Re:About 4 times less performance than without OCi by Anonymous Coward · · Score: 1

    4 times less... times is multiplication, less is subtraction.
    So without overclocking, its performance is a base of "100%".
    After overclocking, its performance is 100% - (4 * 100%), making it -300% now.

  21. Re:About 4 times less performance than without OCi by kimvette · · Score: 1

    75% less would be correct.

    --
    The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
  22. Re:About 4 times less performance than without OCi by Kjella · · Score: 1

    However, the CPU has now only 1 core instead of 8 and only about 1.6 times the clock frequency. This means a huge decrease in performance...

    Amdahl's law says that depends on what you're doing. Also it has 4 cores/8 threads but yeah. This is obviously just for doing it. As someone who started with a 0.985 MHz C64 and got a 1.2GHz Athlon not so long into the new millennium I'm quite underwhelmed though, despite the IPC improvements.

    --
    Live today, because you never know what tomorrow brings
  23. 9 years since a Pentium broke 8 GHz by Anonymous Coward · · Score: 0

    People point out AMD but the title of TFA is is breakig "Intel World Record".

    But how is this a world record, even for Intel? There were several 8 GHz results for Intel processors for several previous generation, starting at least in 2007, and even TFA mentions a Celeron record from 2013... So please, someone explain what record is this?

    1. Re:9 years since a Pentium broke 8 GHz by Anonymous Coward · · Score: 0

      I too remember the 8 GHz Celeron OC:
      http://www.tomshardware.com/news/Overclock-Intel-Celeron-TiN-CPU,9509.html

  24. Funny thing is by Anonymous Coward · · Score: 0

    ...Adobe Flash would still eat 100% of that usage.

  25. Bullshit by Anonymous Coward · · Score: 0

    Nobody has ever been able to make an ASRock motherboard do anything other than cause a fit of rage while you get an RMA.

  26. At full tilt ? by justthinkit · · Score: 1

    I'm curious also if that one core was even running at full tilt. Processors these days sleep by default. So is this test one of "boot up and run some cpu measuring utility" or did it at least run at 100% utilization for a while?

    --
    I come here for the love
  27. Re:About 4 times less performance than without OCi by Anonymous Coward · · Score: 0

    Yeah but think about how amazingly fast javascript would run.

  28. Re: About 4 times less performance than without OC by Anonymous Coward · · Score: 0

    Yeeeeessssss, and 100-400 = -300

  29. Asrock?! by slashmydots · · Score: 1

    They probably would have hit 10GHz if they got 7GHz on some garbage Asrock board. If you're not familiar, it's a budget spin-off of ASUS and I've had nothing but problems with defective Asrock motherboards at my computer repair shop.

    1. Re:Asrock?! by Banana+Slamma · · Score: 1

      Asrock's products are crap at the lower end, but at least from what I've _heard_ their high end stuff is the same as everything else.

    2. Re:Asrock?! by toddestan · · Score: 1

      Considering the problems the genuine Asus boards have given me over the years, I can only imagine how terrible Asrock must be. I'd take my changes with ECS first.

  30. Re:About 4 times less performance than without OCi by thisisauniqueid · · Score: 1

    You're making the ridiculously optimistic assumption that users are running software that can actually make use of more than one core. (It's 2016; why do all mainstream languages still generate single-threaded code by default?)

  31. Re:About 4 times less performance than without OCi by dsmatthews9379 · · Score: 0

    Circuits using liquefied gases for cooling passed 20 GHz over 40 years ago, so this fun experiment is the equivalent of strapping a JATO onto a bicycle.

  32. Obligatory Car Analogy by GrahamCox · · Score: 1

    It be like: strip a car down to its bare chassis, removing all extraneous weight, seats, anything that's there for comfort or convenience, then keep cutting out more metal - panels, half the floor, roof. Bolt on a huge turbo and nitrous and watch it do a ¼ mile in 8 seconds. Great, but it ain't going to get you to the shops and back.

  33. I'd underclock... by unixisc · · Score: 1

    Run it at something like 4-5GHz and save on the power consumption. Hope that the multi-core helps here

  34. Pointless by Anonymous Coward · · Score: 0

    This is single thread 7GHz, it's entirely useless today.

  35. Re: About 4 times less performance than without OC by radiumsoup · · Score: 1

    circuits != processors

  36. It seems that /. is actually slower than newpapers by Anonymous Coward · · Score: 0

    One benefit I find to online news is that it happens faster than the newspaper variety.

    I read about this on Flipboard FIVE days ago.

  37. Faster and Faster by djnforce9 · · Score: 1

    I remember back when even 4Ghz required liquid cooling. This was when people were doing it to Pentium IV chips. It's surreal knowing that 4Ghz is now the "stock speed" while 7Ghz is the new threshold to reach.