Hardware atrocities



  • OK, what's the worse thing you ever did  in order to get a broken computer going again?

     Since a significant part of my C4I duties involve (unfortunately) the military equivalent of managing a help desk and performing hardware and software repair and maintenance on a wide range of heterogeneous machines, I've seen and done quite a few crudities on the job.

     What to I mean by "wide range of heterogeneous machines" ? Well...I had already mentioned it in an older comment of mine but basically the machines on my corp's network can be anything from a Pentium MMX with 64 MB of RAM, upwards to a Core 2 with 2 Gigs, with the majority being machines in the Pentium 3 - Pentium 4 class (with Socket 478 Celerons being the most common). And let's not even digress on the types of RAM modules and hard disks used... ;-)

    Anyway, if something is reasonably modern and the only problem is an aging installation of windows, after excluding hardware problems like bad memory modules, bad hard disks and PSUs, the usual procedure is doind a repair installation of Windows (since in most cases a complete reformatting is not an option) and getting hold of all the latest available drivers for every given mobo, graphics card, or whatever.

     If something however is crap-tastic, like e.g. a Pentium II with 128 MB of RAM running windows XP from a 2 GB hard disk, then we found a way to speed it up a bit by adding an obsolete 500 MB hard disk as a "swap drive". Voila', no more trashing, and a use for an otherwise unusable hard disk.

     Sounds tragic?

     Well...what to say about the bearing-less cooling fans from old nVidias TNT2 I had to discard because they turned into a rusty, self-locking mass? Now more than one TNT2 still works...with no fan. It was just a dead weight anyway, since it didn't even spin.

     Or, the fact of installing Windows XP SP3 on a Pentium I with 64 MB of RAM, which surprisingly worked quite well after trimming some services down (could be used for editing a single doc file in Office XP, and browsing a couple of intranet pages). Others did such experiments for fun, I had to do them due to service necessities.

     Put me to the test, post your own tragic hardware and help desk stories ;-)

     

     



  • I was running Visual Studio 98 (VC 6) on a 128 MB XP machine pretty well.

     



  • @alegr said:

    I was running Visual Studio 98 (VC 6) on a 128 MB XP machine pretty well.

     

     

    Try running JBuilder (any version) with less than 256 MB and then we're talking.



  •  I remember Eclipse trashing wildly when I had 512 MB.



  • @Juifeng said:

     I remember Eclipse trashing wildly when I had 512 MB.

     

     Throw a secondary 512 MB hard disk for swap space, and j00r b0x w1ll b3 t3h p4wn 4dn t3h h4x.



  • @C4I_Officer said:

    Or, the fact of installing Windows XP SP3 on a Pentium I with 64 MB of RAM, which surprisingly worked quite well after trimming some services down (could be used for editing a single doc file in Office XP, and browsing a couple of intranet pages). Others did such experiments for fun, I had to do them due to service necessities.
    I had to do upgrades on a laptop with similar specs a few weeks ago (though IIRC, the CPU was Celeron - it was some ThinkPad with 64MB RAM running XP Pro and Office XP). It had a single USB port, but I couldn't insert my USB pendrive to it because it was placed in such way that only plain USB cables could be inserted. And to add insult to the injury, the only network connection went through a 10Mb hub that limited actual network speed to around 200kB/s.



  • I'm pretty sure a mixture of modern-ish Dell boxes and late-80s dumb terminals in the same office qualify, the former with emulator software of some description installed to run the delivery schedule database that the company had presumably been using since the terminals were cutting-edge. Though in fact TRWTF wasn't the workstations or even the software, which was actually still perfectly serviceable, but that's another thread.



  • Mine isn't up there with your stories, but I am currently running a secondhand ATX motherboard and and PSU out of an AT case. The AT case's power button can only be pressed and held down, so I wired up a simple pushbutton to cross the motherboard's start leads. The pushbutton protrudes from the back of the case from a small hole near the case fan. Works like a charm.



  •  Well, not that this is all that uncommon:

    Some years ago I was a part-time tech for a computer store. We had a customer bring in a Packard Bell which I believe she got at a yard sale. It was blue screening and needed a complete reload. After a memory upgrade, it installed XP Home. Then the cx brought in MS Works (v7 I believe) and insisted we install it. The techs were placing bets whether the PB would choke or not. Damned if the old machine didn't pull it off and run it. Of course, it took a full 30 minutes to open Word, but it did it!

    Oddly, the cx made us uninstall it all later. Can't please some folks I guess.



  • My first computer was an AMD K6-2 333 Mhz with 96 MB of memory (64+32). Ran Windows 2000 perfectly. Used about 40-45 MB of memory with some services disabled. It was a 3.2GB hard drive though.

    So it's really nothing special... operating systems can work on some very old computers if configured the right way.



  • I used to run Windows 2000 under Virtual PC on my 1 GHz G4 and would cross debug from a different Mac to the virtual machine. I was running Discreet Combustion on it, which was a fairly heavy duty application on a real machine. It was slow, but it saved me from having to buy a real PC for a year or two, and still allowed me to produce Windows software written and compiled on my Mac. Fun times! Slow, but fun. ;)



  • I continue to own a Mac SE and an IBM PS/2 XT. Both of them still boot. Not that I use them for anything... but once, for no real reason, I tried installing MS-DOS 2.10 on the XT's hard drive. It doesn't actually know what a hard drive is. It will cheerfully install from the floppy onto the hard disk, but when you try to boot off the C drive, it panics and misbehaves in amusing ways.



  •  The case of the dumb terminals could just be because of the use of legacy software. It is a WTF, but of a different category, much like flight reservation systems. But using an os with an update pack (XP SP3) that came more than a decade after the hardware it is called to run on...and be used in a productive setting, that is. Well, as productive the work of military office staff can be, anyway.



  • @Welbog said:

    Mine isn't up there with your stories, but I am currently running a secondhand ATX motherboard and and PSU out of an AT case. The AT case's power button can only be pressed and held down, so I wired up a simple pushbutton to cross the motherboard's start leads. The pushbutton protrudes from the back of the case from a small hole near the case fan. Works like a charm.
    Your AT case didn't have a reset button? Anyway, I did the button at the back of case once with my home server, since it was in such location that the front wasn't easily accessible.

    And speaking of cases, this used to be my workstation for a few years.



  • @ender said:

    And speaking of cases, this used to be my workstation for a few years.

    I don't think that site likes hotlinking; it redirected me to something called 'hotlinkr.png' that crashed Firefox. I did manage to pull them off with Free Download Manager, though. I think I just found a new DIY project. 



  • @Welbog said:

    Mine isn't up there with your stories, but I am currently running a secondhand ATX motherboard and and PSU out of an AT case. The AT case's power button can only be pressed and held down, so I wired up a simple pushbutton to cross the motherboard's start leads. The pushbutton protrudes from the back of the case from a small hole near the case fan. Works like a charm.

    No duck tape in Canada, eh? 



  • @morbiuswilters said:

    @Welbog said:

    Mine isn't up there with your stories, but I am currently running a secondhand ATX motherboard and and PSU out of an AT case. The AT case's power button can only be pressed and held down, so I wired up a simple pushbutton to cross the motherboard's start leads. The pushbutton protrudes from the back of the case from a small hole near the case fan. Works like a charm.

    No duck tape in Canada, eh? 

    Nope.  They have geese, but nowhere to stick 'em.



  • @C4I_Officer said:

    OK, what's the worse thing you ever did  in order to get a broken computer going again?
     

    i think it was trying to run my Athlon 1.2Ghz without fan so I could touch it to find out whether it's warm (and therefore there's power going through it)...

    I was young and stupid... well.. kinda more than I'm now, and I didn't realise that if there weren't no power going to CPU, the computer wouldn't start at all.

    So, it was warm indeed. Very warm. More like hot.

    Burned in 6 seconds.

    (Yup, another stupid thing - when i noticed that it's getting dangerously warm (:-D) i tried to turn the computer off with its power switch. You know, that one you have to press and hold for 5 seconds before it turns off...)



  • @SEMI-HYBRID code said:

    @C4I_Officer said:

    OK, what's the worse thing you ever did  in order to get a broken computer going again?
     

    i think it was trying to run my Athlon 1.2Ghz without fan so I could touch it to find out whether it's warm (and therefore there's power going through it)...

    I was young and stupid... well.. kinda more than I'm now, and I didn't realise that if there weren't no power going to CPU, the computer wouldn't start at all.

    So, it was warm indeed. Very warm. More like hot.

    Burned in 6 seconds.

    (Yup, another stupid thing - when i noticed that it's getting dangerously warm (:-D) i tried to turn the computer off with its power switch. You know, that one you have to press and hold for 5 seconds before it turns off...)

    No correctly functioning CPU should burn out in 6 seconds just because it was running without a fan.  The thing must actually have had a power rail short and been just dumping supply current straight through the silicon to do that.



  • @ender said:

    @Welbog said:
    Mine isn't up there with your stories, but I am currently running a secondhand ATX motherboard and and PSU out of an AT case. The AT case's power button can only be pressed and held down, so I wired up a simple pushbutton to cross the motherboard's start leads. The pushbutton protrudes from the back of the case from a small hole near the case fan. Works like a charm.
    Your AT case didn't have a reset button? Anyway, I did the button at the back of case once with my home server, since it was in such location that the front wasn't easily accessible.

    I just recently soldered a couple of really long flying leads onto a spare pin header and pushed the reset switch plug onto that header instead of the pcb header, then I routed the long flying leads out the back of my PC so that I could reset it when I wanted to by sparking them together with a screwdriver.




  • @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds just because it was running without a fan.  The thing must actually have had a power rail short and been just dumping supply current straight through the silicon to do that.
    32bit Athlons didn't have any thermal protection, and if you tried running them without heatsink, they let out the magic smoke really fast. That said, I remember when I bought my first Athlon (800MHz model), the booklet had a very visible warning to not ever run the CPU without heatsink, even if it's just for a few seconds (speaking of this CPU, I later gave it to my sister, and for some reason it wouldn't be stable on her motherboard unless overclocked to 887MHz - never found out what caused that).
    Oh, and one more hardware atrocity :)



  • I regularly ran Windows 2000 and Visual Studio 6 on a Pentium I 233MHz laptop with 64MB RAM and a 2GB hard drive.  Worked pretty well.  Amusingly, WinME refused to install on the laptop (saying a minimum of 250MHz was required) but Windows 2000 didn't care...

    Yes, yes, I know, TRWTF is that I tried to install WinME on it.



  • @alegr said:

    I was running Visual Studio 98 (VC 6) on a 128 MB XP machine pretty well.

     

    I used to run Microsoft C++ 7 on a 486/33 with 4MB of RAM. Worked just fine.

    On the hardware front, my webserver is running on a Pentium MMX with either 16MB or 48MB of RAM, depending on if the CPU fan was spinning during power-on or not.

  • Discourse touched me in a no-no place

     Recently worked on a project to replace an old terminal based logistics system with a shiny new modern system.

     The customer, however, wanted to keep their room full of dinosaur-era heavy iron for whatever reason, and wanted to keep the database running on them. The problem: The application was completely custom and we weren't about to write any fucking code to run on those. Since the new app is essentially a direct port,  plus some additional features that are really just taking data from one part of the old system and shoehorning it into another, we opted for a disgusting solution. Into the Dinosaur Pit went a rack stuffed with modern servers, switches, power units, all the fixin's - every one of those servers came specially equipped with a pair of 16-port serial interface cards for 32 ports each. Those serial cards were then connected to the old (and now vacant) dumb-terminal ports on the iron. If you can't guess already, our app presents to the user and then talks to the database by way of sending keystrokes to the vaxen over the serial lines and scraping the data out of the ASCII it sends back. It's actually particularly clever, because if it needs more than one screen of data, it'll go fetch it simultaneously via another serial port - and it does this intelligently based on server load - i.e. if the server has a bunch of connected users, it'll just do it all over a single serial connection.

     By the way, at the speeds we're running the serial connections we get about .3mbits of theoretical bandwidth between each server and the dinos. But remember, the dinosaurs are still spitting out all the old terminal formatting characters and whitespace and etc. required to make the app look nice on character screens - so there's a CRAPTON of overhead.

     

    Yes, the software team did in fact make DAMN CERTAIN that the entire data access layer is abstracted away - in fact, we did most of our testing with a SQL Server data access layer - which we can activate for the customer in the future by swapping out a DLL, installing SQL Server (Express, given the customer's needs), and doing a data migration - which we also already coded... to run over the serial ports). So in essence, they've gotten two products - the shitty hack they're using, and the Real Thing, which they SHOULD be using if it weren't for their unwillingness to let dinosaurs die.



  • @DaveK said:

    No correctly functioning CPU should burn
    out in 6 seconds just because it was running without a fan.  The thing
    must actually have had a power rail short and been just dumping supply
    current straight through the silicon to do that.
     

    Heh, it's obvious you never watched the infamous video by Tom's Hardware.  Just sit back, enjoy the cool Project X music, and watch what happens ;-)

    For the record, I did power two boards once with no heatsink on because although they appeared to come alive, there was no POST or keyboard activity to indicate a living CPU. So I just powered it up with no heatsink and...well...the CPU stayed stone-cold in one of the boards. Nada. The other time the board didn't support a certain FSB on the CPU (but I didn't know that) and it partially booted up, again with no POST or error beeps. Excluding memory etc. faults, I tried to see if at least the CPU worked...and yes, it did :-D I left it on for maybe 2 seconds and got hot really fast, but I was ready to unplug everything with my other hand.

    On a related note, I once left a Pentium MMX @ 166 MHz powered on with the fan stuck...it got awful hot but survived, and as soon as I restored it, it just cooled of nicely. And it's still working :-p These things sure could work with much higher temperature extremes....





  • @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds just because it was running without a fan.  The thing must actually have had a power rail short and been just dumping supply current straight through the silicon to do that.

    Your continued commitment to ignorance is a source of inspiration for us all. 



  • @C4I_Officer said:

    @DaveK said:

    No correctly functioning CPU should burn
    out in 6 seconds just because it was running without a fan.  The thing
    must actually have had a power rail short and been just dumping supply
    current straight through the silicon to do that.
     

    Heh, it's obvious you never watched the infamous video by Tom's Hardware.  Just sit back, enjoy the cool Project X music, and watch what happens ;-)

      I did.  What I saw could best be described as "an obvious stitch-up" :-) 

    Besides, there's a lot of difference between the load on a CPU running a 3d-heavy FPS under a full operating system, as those were doing, and the load on a CPU that has just switched on and is busy doing POST stuff and probably hasn't even fired up the clock multiplier yet.

    @C4I_Officer said:

    For the record, I did power two boards once with no heatsink on because although they appeared to come alive, there was no POST or keyboard activity to indicate a living CPU. So I just powered it up with no heatsink and...well...the CPU stayed stone-cold in one of the boards. Nada. The other time the board didn't support a certain FSB on the CPU (but I didn't know that) and it partially booted up, again with no POST or error beeps. Excluding memory etc. faults, I tried to see if at least the CPU worked...and yes, it did :-D I left it on for maybe 2 seconds and got hot really fast, but I was ready to unplug everything with my other hand.

    On a related note, I once left a Pentium MMX @ 166 MHz powered on with the fan stuck...it got awful hot but survived, and as soon as I restored it, it just cooled of nicely. And it's still working :-p These things sure could work with much higher temperature extremes....

    Also for the record, I did this just the other day with a laptop motherboard that was giving me trouble.  After thirty seconds or so of sitting there playing POST beeps at me, the cpu was just barely warm to the touch.  I suspect that the reason your other CPU got hot so quickly was because it had already developed or was already developing a fault; did it go on to have a long and happy life after your experiment?



  • @morbiuswilters said:

    @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds just because it was running without a fan.  The thing must actually have had a power rail short and been just dumping supply current straight through the silicon to do that.

    Your continued commitment to ignorance is a source of inspiration for us all. 

    Well, since I've actually done what we're talking about here within the past fortnight, and you're just spouting armchair theoretical knowledge, I'd say that your continued commitment to your career as a lame warrior is a source of lulz and fail for us all!

    Wait, that's probably a bit beyond your reading level.  Would you like me to illustrate it with pretty pictures?  ;-) 



  • @DaveK said:

    @morbiuswilters said:

    @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds just because it was running without a fan.  The thing must actually have had a power rail short and been just dumping supply current straight through the silicon to do that.

    Your continued commitment to ignorance is a source of inspiration for us all. 

    Well, since I've actually done what we're talking about here within the past fortnight, and you're just spouting armchair theoretical knowledge, I'd say that your continued commitment to your career as a lame warrior is a source of lulz and fail for us all!

    Wait, that's probably a bit beyond your reading level.  Would you like me to illustrate it with pretty pictures?  ;-) 

    As someone who worked extensively with the 1.2 Ghz Athlons, I can tell you that they will burn up quite rapidly if booted without a heatsink.  Without a fan (but still with a massive heatsink) the CPU would lock up 10 to 20 seconds into booting Windows with a heatsink surface temperature of 180 F or so.  In their day, those were some of the most powerful desktop CPUs available and they overclocked quite well (a 1.2 could easily reach 1.7 with good air cooling) but keeping them cool was quite difficult.  I really wish you would educate yourself and stop spreading ignorance or outright lies.



  • @morbiuswilters said:

    @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds

    Without a fan (but still with a massive heatsink) the CPU would lock up 10 to 20 seconds into booting Windows 

    You need to learn to read for comprehension, not for a pretext to flame someone.  HTH.

     



  • @Heron said:

    I regularly ran Windows 2000 and Visual Studio 6 on a Pentium I 233MHz laptop with 64MB RAM and a 2GB hard drive.
     

    i had similar laptop, exept it ran WinXP. And Also Dreamweaver, Flash and Fireworks MX.

    it was LOTS of fun (but required LOTS of patience).



  • @morbiuswilters said:

    Your continued commitment to ignorance is a source of inspiration for us all.
    Your continued commitment to gratuitous flaming is a source of irritation for us all.



  • @DaveK said:

    @C4I_Officer said:

    Heh, it's obvious you never watched the infamous video by Tom's Hardware.  Just sit back, enjoy the cool Project X music, and watch what happens ;-)

      I did.  What I saw could best be described as "an obvious stitch-up" :-)

     Where's the stitch-up? They were tested under the exact same circumstances, with the same 3D-heavy game, and the reaction was quite different. One (the Pentium 4) obviousy slowed/idled itself down to safe enough levels and restored full power as soon as cooling was restored, while the Athlon XP barfed with no hint of self-protecting itself.It just proved the point that Athlon XP don't have sufficient (if any, as if, at all) thermal protection.

    @DaveK said:

    Also for the record, I did this just the other day with a laptop motherboard that was giving me trouble.  After thirty seconds or so of sitting there playing POST beeps at me, the cpu was just barely warm to the touch.  I suspect that the reason your other CPU got hot so quickly was because it had already developed or was already developing a fault; did it go on to have a long and happy life after your experiment?
     

     Yup, it's still working fine. It was a Celeron 1.7 GHz , which normally has a core temp of about 40 C when idle with the heatsink installed. Anything above 50 however feels red-hot to the bare skin (especially a metallic/ceramic surface like the CPU's packaging). It could be that if left on long enough, it would self-idle itself like the Pentium 4 on tom's hardware and actually get cooler to the touch, but for the first few seconds it was just firing itself at full power and the temperature differential was very obvious.



  •  In any case, the full story about the THG test can be found here, and lo' and behold, even AMD promptly admitted there was a major fuck-up with that  processor line, and that unless the motherboard also integrates certain protection functions and the heat-sink accident doesn't happen during boot, there's a very real possibility of sending the CPU up in smoke.

     On the converse:

    [b]THG said:[/b]

    The Pentium 4 core comes equipped with a thermal monitoring unit
    that permanently checks the temperature
    . As soon as the core
    temperature has reached a certain trigger value, the thermal unit
    throttles down the clock of Pentium 4 until a safe temperature has been
    reached.

    This solution is clearly commendable and proves that Intel's idea of equipping Pentium 4 with clock-throttling was far from a bad idea, as some sources want to make us believe. It is pretty much impossible to 'fry' a Pentium 4 processor. Additionally, Pentium 4 remains operational even once the thermal catastrophe took place and the heat sink fell off.

     

    While for the Athlon XP...

     

    The removal of the heat sink proves to be fatal. In less than a second Athlon 1400 dies the heat death. It doesn't take long and the core reaches a temperature of extremely hefty 370 degrees Celsius / 698 degrees Fahrenheit. If the user of the Athlon system doesn't turn off his box immediately, the motherboard will be destroyed too. There's even the risk of a fire.

    AMD did not bless the Thunderbird core with ANY thermal protection whatsoever. If the heat sink should come off, the owner is facing a significant financial loss.

     

    Quod erat demonstrandum.

     

     



  • @DaveK said:

    @morbiuswilters said:

    @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds

    Without a fan (but still with a massive heatsink) the CPU would lock up 10 to 20 seconds into booting Windows 

    You need to learn to read for comprehension, not for a pretext to flame someone.  HTH.

    So in Fuckwitville (Population: You), "6 seconds without a heatsink or fan" is equivalent to "10 to 20 seconds with only a heatsink" and "burn out" is equivalent to "lock up".  Please stop this sad little theater and admit you were wrong. 



  • @morbiuswilters said:

    So in Fuckwitville (Population: You), "6 seconds without a heatsink or fan" is equivalent to "10 to 20 seconds with only a heatsink" and "burn out" is equivalent to "lock up".  Please stop this sad little theater and admit you were wrong. 

     Well...don't expect him to bow down touching the ground samurai-style and humbly saying "Morbiuswilters-san, gomenasai!" anytime soon.



  • @DaveK said:

    @C4I_Officer said:

    @DaveK said:

    No correctly functioning CPU should burn
    out in 6 seconds just because it was running without a fan.  The thing
    must actually have had a power rail short and been just dumping supply
    current straight through the silicon to do that.
     

    Heh, it's obvious you never watched the infamous video by Tom's Hardware.  Just sit back, enjoy the cool Project X music, and watch what happens ;-)

      I did.  What I saw could best be described as "an obvious stitch-up" :-) 

    Besides, there's a lot of difference between the load on a CPU running a 3d-heavy FPS under a full operating system, as those were doing, and the load on a CPU that has just switched on and is busy doing POST stuff and probably hasn't even fired up the clock multiplier yet.

    A local computer magazine here in Finland tested booting without a heat sink and reached very similar temperatures.  I recall the Athlon going off the scale (which ended at 200 °C) in under 10 seconds.

    @DaveK said:

    @C4I_Officer said:

    For the record, I did power two boards once with no heatsink on because although they appeared to come alive, there was no POST or keyboard activity to indicate a living CPU. So I just powered it up with no heatsink and...well...the CPU stayed stone-cold in one of the boards. Nada. The other time the board didn't support a certain FSB on the CPU (but I didn't know that) and it partially booted up, again with no POST or error beeps. Excluding memory etc. faults, I tried to see if at least the CPU worked...and yes, it did :-D I left it on for maybe 2 seconds and got hot really fast, but I was ready to unplug everything with my other hand.

    On a related note, I once left a Pentium MMX @ 166 MHz powered on with the fan stuck...it got awful hot but survived, and as soon as I restored it, it just cooled of nicely. And it's still working :-p These things sure could work with much higher temperature extremes....

    Also for the record, I did this just the other day with a laptop motherboard that was giving me trouble.  After thirty seconds or so of sitting there playing POST beeps at me, the cpu was just barely warm to the touch.  I suspect that the reason your other CPU got hot so quickly was because it had already developed or was already developing a fault; did it go on to have a long and happy life after your experiment?

    You do realize that there's a large difference between desktop and laptop CPUs, right?  Athlon XP processors have a thermal design power of around 70 W.  Pentium M processors have a TDP of only 10-20 W.  Also, being designed for low-power devices, they can completely shut down parts of themselves, so the power consumption during boot is likely to be even lower.  Desktop processors have only recently gained that ability.

    Being a scientific person, I did a bit of analysis and simulation on the topic:

    Silicon has a specific heat capacity of 0.705 J/g K and a density of 2.329 g/cm³.  This gives a CPU die with an area of 150 mm² and thickness of 2 mm a heat capacity of 0.49 J.  The full 70 W of power dissipation heats the die at over 140 K/s.

    The main way of getting heat away from an uncooled die is conduction through the piece of PCB it sits on.  Thermal radiation at temperatures survivable by the die is a mere 0.2 W and I think convection is rather insignificant too with such a small heat exchange area.  Assuming 300 °C to be the equilibrium temperature, the thermal conductivity must be 0.25 W/°C.  A standard cooler is somewhere in the 3 W/°C ballpark, so I think that's a pretty optimistic estimate.

    And how fast will the chip be destroyed?  Pretty fast.  I wrote a small simulation loop, according to which a temperature of 100 °C was reached in 0.65 s, 150 °C in 1.22 s and 200 °C in 2.01s.  Even after halving the heating power to 35 W, 100 °C was reached in 1.65 s and 150 °C in 5.15 s.

    Now, this experiment was definitely oversimplified.  I had to guesstimate the efficiency of the PCB as a heatsink, and I honestly don't know how much heat is carried away by convection.  But unless I made some horrible mistake, I believe it does show that a CPU being destroyed by running without a heatsink is plausible. 



  • [ folding two replies into one ]

    @C4I_Officer said:

    @DaveK said:

    @C4I_Officer said:

    Heh, it's obvious you never watched the infamous video by Tom's Hardware.  Just sit back, enjoy the cool Project X music, and watch what happens ;-)

      I did.  What I saw could best be described as "an obvious stitch-up" :-)

     Where's
    the stitch-up?

    Hence the smiley.  It wasn't a serious accusation, but it did look like that Athlon was already damaged before the cooler came off. 

     @C4I_Officer said:

     In any case, the full story about the THG test can be found here, and lo' and behold, even AMD promptly admitted there was a major fuck-up with that  processor line,

    Yep, I had an early athlon (850) myself and I do remember how hot the damn thing always ran.  In the end I ripped the side off the case and shoved a floor-standing air blower bodily into it, and even then it used to get hot enough on a summer day after being running heavily loaded for a while to trigger the BIOS overheat alarm (on the lowest setting).

    The only thing I'll add to this discussion is that I don't call something with a massive design flaw "correctly functioning".  I call it "unable to function correctly".  YMMV.



  • @morbiuswilters said:

    @DaveK said:

    @morbiuswilters said:

    @DaveK said:

    No correctly functioning CPU should burn out in 6 seconds

    Without a fan (but still with a massive heatsink) the CPU would lock up 10 to 20 seconds into booting Windows 

    You need to learn to read for comprehension, not for a pretext to flame someone.  HTH.

    So in Fuckwitville (Population: You), "6 seconds without a heatsink or fan" is equivalent to "10 to 20 seconds with only a heatsink" and "burn out" is equivalent to "lock up".  Please stop this sad little theater and admit you were wrong. 

    You're starting to get so angry you can't read straight!  My point was that those two are NOT the same thing.

    I really am going to have to resort to cave paintings to communicate with you at this rate :-)



  • @DaveK said:

    Yep, I had an early athlon (850) myself and I do remember how hot the damn thing always ran.  In the end I ripped the side off the case and shoved a floor-standing air blower bodily into it, and even then it used to get hot enough on a summer day after being running heavily loaded for a while to trigger the BIOS overheat alarm (on the lowest setting).

    The only thing I'll add to this discussion is that I don't call something with a massive design flaw "correctly functioning".  I call it "unable to function correctly".  YMMV.

    What in the fuck are you talking about?  The early Athlons ran hot, but that's not a "massive design flaw".  It wasn't that difficult to cool them properly and if you were having that much trouble you either had a defective CPU or you don't know how to cool a fucking computer or house.  None of this changes the fact that an Athlon without a heatsink would fry in under 10 seconds, which was the original point that was made.  You jumped in like an ignorant asshat and claimed the only way that could happen is if the CPU was defective, which has been refuted by several people already.  In spite of your pathetic attempts to "prove" your point by ignoring every legitimate argument and rambling like a autistic dipshit, you are clearly wrong. 



  • Don't bother,morbiuswilters. You are practically debating the shotgun argument with someone who is not willing to yield, for whatever reasons.

    If you continue, you will just catch yourself in a useless philosophical wordplay over what constitutes a "correctly functioning" CPU.

     



  • @C4I_Officer said:

    If you continue, you will just catch yourself in a useless philosophical wordplay over what constitutes a "correctly functioning" CPU.
    A correctly functioning CPU should clean the house, cook you dinner, and not steal the covers in bed.  Or is that a wife?



  • @bstorer said:

    A correctly functioning CPU should clean the house, cook you dinner, and not steal the covers in bed.  Or is that a wife?
     

    My wife thinks I should do those things.



  • @Heron said:

    @bstorer said:

    A correctly functioning CPU should clean the house, cook you dinner, and not steal the covers in bed.  Or is that a wife?
     

    My wife thinks I should do those things.

    Clearly you have not domestically violated her enough with domestic violence. 



  • @Heron said:

    @bstorer said:

    A correctly functioning CPU should clean the house, cook you dinner, and not steal the covers in bed.  Or is that a wife?
     

    My wife thinks I should do those things.

    Sounds like it's time for an upgrade.



  • @bstorer said:

    Sounds like it's time for an upgrade.
     

    Our baby girl was born on Monday.  Does that count?



  • @Heron said:

    @bstorer said:

    Sounds like it's time for an upgrade.
     

    Our baby girl was born on Monday.  Does that count?

    inb4pedophelia


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.