Thread: PC still shutting off randomly

  1. #16
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by cpjust View Post
    I have a similar problem with my computer, except instead of shutting down, it freezes and the video gets all screwed up.
    That sounds like bad memory to me.

    I finally decided to buy a little fan and power cable (soldered them together) and glued the fan onto the heatsink of the chip.
    It still locks up every once in a while, but not quite as often as before.
    Now I have a frankenstein machine with a loud fan...
    Why wouldn't you just use a proper Chipset fan and take a tap off your power supply for it?

    Newegg.com - Computer Hardware,Fans & Heatsinks,Case Fans,40mm

  2. #17
    Registered User kryptkat's Avatar
    Join Date
    Dec 2002
    Posts
    638
    good point i did assume that the latest versions of the os had software that would monitor the thermister too.

  3. #18
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by kryptkat View Post
    good point i did assume that the latest versions of the os had software that would monitor the thermister too.
    Ahhh... well if you want to peek at temperatures...

    CPUID - System & hardware benchmark, monitoring, reporting

    But, at least now you know why they don't just turn things off...

  4. #19
    Master Apprentice phantomotap's Avatar
    Join Date
    Jan 2008
    Posts
    5,108
    Intel CPUs, for example, don't shut down... they slow down until they cool off.
    AMD machines do the opposite, they try to stay cool at all times, running at low speed and staying cool until demand increases.
    But, at least now you know why they don't just turn things off...
    O_o

    Seriously? Okay, you need to stop speaking about the heat issue. You obviously have no real world experience with this stuff.

    Go to any technical support forum and you will find dozens to tens of thousands of examples thrashing such ridiculous statements.

    Soma
    Last edited by phantomotap; 01-02-2011 at 06:11 PM.

  5. #20
    and the hat of sweating
    Join Date
    Aug 2007
    Location
    Toronto, ON
    Posts
    3,545
    Quote Originally Posted by CommonTater View Post
    Why wouldn't you just use a proper Chipset fan and take a tap off your power supply for it?

    Newegg.com - Computer Hardware,Fans & Heatsinks,Case Fans,40mm
    Sure, now you tell me. I was looking for a 40mm fan with the proper power cable, but I couldn't find one. So I bought one from The Source, but it just had bare wires...

    Maybe I'll try the memory again, but I'm pretty sure it's the G41 chip that's the problem because it's much easier to reproduce the lock-up when there's a very high level of disk I/O which is handled by that chip.
    "I am probably the laziest programmer on the planet, a fact with which anyone who has ever seen my code will agree." - esbo, 11/15/2008

    "the internet is a scary place to be thats why i dont use it much." - billet, 03/17/2010

  6. #21
    Internet Superhero
    Join Date
    Sep 2006
    Location
    Denmark
    Posts
    964
    Quote Originally Posted by phantomotap View Post
    O_o

    Seriously? Okay, you need to stop speaking about the heat issue. You obviously have no real world experience with this stuff.
    The part about Intel CPUs slowing down is quite true, it's called power throttling, my old E8500 would get unusably slow when above 100 degrees celsius. I haven't owned an AMD CPU since the Athlon 2000+ days so i wouldn't know about that.
    How I need a drink, alcoholic in nature, after the heavy lectures involving quantum mechanics.

  7. #22
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Ireland
    Posts
    8,446
    Yup. Collectively, these technologies are known as dynamic frequency scaling. AMD uses it too across its entire line of modern processors. CommonTater is quite right. I would just clarify that the AMD CPU is throttled when idle and the Intel CPU is throttled when overheated.

    Quote Originally Posted by phantomotap View Post
    Seriously? Okay, you need to stop speaking about the heat issue. You obviously have no real world experience with this stuff.
    It's alright. Everyone eventually have to put their foot in their mouth. This is your turn.
    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

  8. #23
    Master Apprentice phantomotap's Avatar
    Join Date
    Jan 2008
    Posts
    5,108
    The part about Intel CPUs slowing down is quite true, it's called power throttling, my old E8500 would get unusably slow when above 100 degrees celsius.
    It's alright. Everyone eventually have to put their foot in their mouth. This is your turn.
    O_o

    So you two honestly believe that neither Intel or AMD powered boxes shutdown due to overheating!?

    Wow. Will you two please promise me to stay well away from any server environment?

    Soma

  9. #24
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by phantomotap View Post
    O_o

    Seriously? Okay, you need to stop speaking about the heat issue. You obviously have no real world experience with this stuff.

    Go to any technical support forum and you will find dozens to tens of thousands of examples thrashing such ridiculous statements.

    Soma
    Seriously? Okay...

    Check out AMD "Cool and Quiet"... Yes, it really is demand based CPU throttling with the exact intention of keeping things as cool as possible. AMD introduced this with the Athlon II processor and continues it in all supsequent designs.

    Also check into Intel's "Speed Step" technlogy which is... well lookie there... more CPU throttling. Intel introduced this with the Pentium IV processor and continues it to present.

    For example: I'm currently on an AMD X2 system. The CPU is presently running at about 1.2ghz. If I do anything more demanding than typing it's going to ramp up it's speed incrementally until it hits 2.8ghz under about 80% load. If I use CPUZ I can actually see this happening on it's report screen.

    CPUID - System & hardware benchmark, monitoring, reporting

    Seriously? Okay...

  10. #25
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by phantomotap View Post
    O_o

    So you two honestly believe that neither Intel or AMD powered boxes shutdown due to overheating!?

    Wow. Will you two please promise me to stay well away from any server environment?

    Soma
    As I pointed out earlier... by design they do not simply shut off. This is to allow the system to cool and restore itself at least far enough to allow operators to save their work before shutting their systems down. If you've ever worked in an environment where data loss of even a couple of minutes is a disaster you would understand why these safeguards are in place.

    If you have machines shutting off, without warning, because of heat related problems... I guarantee you are doing something horridly wrong.

  11. #26
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by cpjust View Post
    Sure, now you tell me. I was looking for a 40mm fan with the proper power cable, but I couldn't find one. So I bought one from The Source, but it just had bare wires...

    Maybe I'll try the memory again, but I'm pretty sure it's the G41 chip that's the problem because it's much easier to reproduce the lock-up when there's a very high level of disk I/O which is handled by that chip.
    If your BIOS gives you control over chipset voltages, try reducing the voltage a step or two and see what happens....

    By way of a general guideline, you should use the minimum voltage that gives stable operation.

  12. #27
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Ireland
    Posts
    8,446
    Quote Originally Posted by phantomotap View Post
    Wow. Will you two please promise me to stay well away from any server environment?
    Don't worry. I'm not on MIS. But your ad hominem attempts keep being a complete miss since he was quite right. He wasn't saying anything ridiculous as you made a point of saying.

    Meanwhile, I can assure you nobody in here will think that once all failsafe mechanisms give out, a processor would keep operating as if nothing was happening. But I guess pointing out the existence of such mechanisms is reason enough for you to appeal to their ignorance. Guess what...
    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

  13. #28
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by Mario F. View Post
    Don't worry. I'm not on MIS. But your ad hominem attempts keep being a complete miss since he was quite right. He wasn't saying anything ridiculous as you made a point of saying.

    Meanwhile, I can assure you nobody in here will think that once all failsafe mechanisms give out, a processor would keep operating as if nothing was happening. But I guess pointing out the existence of such mechanisms is reason enough for you to appeal to their ignorance. Guess what...
    If you must know the history behind all this... It started with a single video produced by Toms Hardware (the reviews BBS with attitude) that proved AMD processors were not safe for business use.... Intel, had licked the overheat problem by incorporating logic on their chips that would change the multipliers at various stages of temperature... AMD had no such precautions.

    YouTube - What happens when the CPU cooler is removed?

    Shortly after that video released, it was proven AMD's cooling solution was mechanically unstable in tragic ways, since they relied upon 4 compressible rubber "feet" on the chip that behaved very badly when the heatsink became hot. Several people demonstrated that forces as low as 5lb applied at 90 degree angles across the heatsink could --and did-- cause loss of cooling.

    AMD then did some very fast re-engineering and only a few months later, when their 64 bit processors debuted they suddenly had chip-top heat spreaders, thermal throttling and far better "stock" cooling solutions.

    Lesson learned....

    The first line of defense is a properly mounted heatsink and fan. Then comes thermal throttling. Finally comes forced standby until the CPU cools. None of the current generation machines will simply shut off if the CPU overheats... and once again it is because of the risk of losing crucial data.

    Think about it... a typical bank processes tens of thousands of transactions a minute... how disasterous would it be for them to lose even 2 minutes of data?

  14. #29
    Internet Superhero
    Join Date
    Sep 2006
    Location
    Denmark
    Posts
    964
    Quote Originally Posted by phantomotap View Post
    O_o

    So you two honestly believe that neither Intel or AMD powered boxes shutdown due to overheating!?

    Wow. Will you two please promise me to stay well away from any server environment?

    Soma
    I have never said they didn't, if run without a cooler for a sufficient period of time, or given enough vCore, any Intel CPU will power off (Yes i have done both things numerous times). It is however not a heat protection feature, but a failsafe that steps in after a triple fault in the CPU, which CAN be caused by heat, but also a host of other things, or so i believe, anyone is free to correct me on that one actually.

    Edit:

    By way of a general guideline, you should use the minimum voltage that gives stable operation.
    While i don't disagree with this statement, i'd just like to add that most modern motherboards have an auto-setting for the vCore, which is almost guaranteed to give more stability than any manual setting one can come up with, so if stable is what you want, just pick auto instead of fooling around with it yourself. It's very hard, if not impossible to be certain that you have a stable vCore setting. It might run for months without problems and then suddenly BSOD in your face, so don't try to lower the vCore manually even though it seems to be working, atleast if stability is your goal :-)
    Last edited by Neo1; 01-03-2011 at 10:38 PM.
    How I need a drink, alcoholic in nature, after the heavy lectures involving quantum mechanics.

  15. #30
    Banned
    Join Date
    Aug 2010
    Location
    Ontario Canada
    Posts
    9,547
    Quote Originally Posted by Neo1 View Post
    While i don't disagree with this statement, i'd just like to add that most modern motherboards have an auto-setting for the vCore, which is almost guaranteed to give more stability than any manual setting one can come up with, so if stable is what you want, just pick auto instead of fooling around with it yourself. It's very hard, if not impossible to be certain that you have a stable vCore setting. It might run for months without problems and then suddenly BSOD in your face, so don't try to lower the vCore manually even though it seems to be working, atleast if stability is your goal :-)
    Please note this thread seems to be two interlaced, but separate threads at the moment. I did not advise turning down VCore to the CPU... Auto is best for that. I was talking about lowering the voltage on a CHIPSET that is already overheating, possibly because it was mistakenly set to a voltage that's too high -- either manually or automatically.

    I can't give you a brand by brand breakdown but I do know that some motherboards do set these "auto" voltages too high. We had a whole rash of Gigabyte boards at one time that were setting the wrong values on chipsets and causing borderline overheating. The company ended up replacing about 150 of them and then hounding their suppliers for replacements... quite the mess.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Ubisoft loses $54 million
    By VirtualAce in forum General Discussions
    Replies: 19
    Last Post: 05-25-2010, 03:13 PM
  2. Nintendo 3DS
    By Cheeze-It in forum General Discussions
    Replies: 30
    Last Post: 03-27-2010, 02:32 AM
  3. Replies: 1
    Last Post: 10-27-2006, 01:21 PM
  4. System randomly shutting down
    By jverkoey in forum Tech Board
    Replies: 17
    Last Post: 06-10-2005, 06:00 AM
  5. Post you pc pics (mods)
    By biosninja in forum A Brief History of Cprogramming.com
    Replies: 33
    Last Post: 03-15-2004, 04:15 PM