Date: Sun, 2 Sep 2001 01:37:26 -0700 From: "Ted Mittelstaedt" <tedm@toybox.placo.com> To: "Mike Porter" <mupi@mknet.org>, "Sean Chittenden" <sean@chittenden.org>, "Bsd Newbie" <bsdneophyte@yahoo.com> Cc: <freebsd-questions@freebsd.org> Subject: RE: overclocking and FreeBSD stablity... Message-ID: <001301c1338a$7e753b00$1401a8c0@tedm.placo.com> In-Reply-To: <200109020302.f8232pl07186@c1828785-a.saltlk1.ut.home.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>-----Original Message----- >From: Mike Porter [mailto:mupi@mknet.org] >Sent: Saturday, September 01, 2001 8:03 PM >To: Ted Mittelstaedt; Sean Chittenden; Bsd Newbie >Cc: freebsd-questions@freebsd.org >Subject: Re: overclocking and FreeBSD stablity... > > >On Friday 31 August 2001 10:22 pm, Ted Mittelstaedt wrote: >> >-----Original Message----- >> >> From: owner-freebsd-questions@FreeBSD.ORG >> >> >[mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of Sean Chittenden >> > >> >Slowaris wasn't meant to be a performance system and probably chokes >> >when it runs at speeds above 400Mhz. >> >> Solaris runs fine on our Compaq 550Mhz system. >> >> My $0.02 is that the base of the troubles is the machine code that the >> compiler produces. I suspect that when a CPU is overclocked that unless >> the parts are good that the CPU is unable to execute SOME of it's opcodes, >> opcodes that produce certain electrical patterns inside of the CPU that >> may ring and generate electrical wave colissions. While I'm not an EE >> I do know that lengths of traces and such inside of a CPU are held to >> precise tolerances in order to deal with clock propagations and such. It's >> not just the cooling but when you overclock the CPU you can have signals >> arriving at internal parts of the CPU earlier than the designer intended. >> >What you fail to realize in this is two things. First, the designers of >processors measure everything in terms of clock cycles rather than other, >more objective, standards. So for a signal to arrive at its destination >"earlier than intended" the various parts of the CPU have to be operating at >different clock speeds. No no no that's not what I meant. It takes time for a signal to go from one side of the die to the other. While you have a point in that everything is relative to the master clock, it's very easy to have a long and a short trace inside of a chip where a pulse is applied to the the ends and at the other ends it arrives at different times. While it may be negligable because the designers compensate by holding the clock hi or low long enough for a hi or low bit to register, I do know that the chip designers are already having to deal with problems like this caused by the higher and higher clock speeds. This applies to die sizes much smaller than those >currently in use and to clock speeds much higher than those >currently in use. Um, not exactly because a square wave (which is what a clock is) generates lots of nasty harmonics at the transition (they call it the leading and the trailing edge) some of which are much higher frequency than the frequency of the square wave itself. That has to be taken into account as well in the design of the chip. > Eventually, yes, that will be a problem, but not until frequencies with >wavelengths smaller than the internal pathways of the chips (hint: we ain't >there yet...you'll fry your chip (or turn it into something resembling the >Vegas strip, at least) before you'll reach that threshold). > That's a different issue. Once the frequencies are so high that the wavelengths are shorter than the traces in the chip, then we simply won't be able to make chips run past this point using current technology. That's why there's been so much interest in optical chips and such as the hope is that they can be made smaller. But before that time we will run into a different issue - that is that you may be able to get the signal propagated properly, but you must hold the signal high or low a certain amount of time to get the semiconductor receiving it to actually register a logic high or low. The faster the clock the less time you have to hold the signal at a given state. >Second, and this is the primary reason people overclock, is the >*method* used >to determine what clock speed a chip is capable of. (and this varies by >product and manufacturer, of course, but we'll stick with Intel for the time >being). When intel makes a chip, it is part of a wafer, which has several >(as many as 10-12, depending on what they are building) chips. This entire >wafer is tested to determine the maximum clock speed every chip in the wafer >will run reliably at. This is based on a number of factors, including the >maximum clock speed they are currently building for that product line, and >other stuff. If EVERY chip in the wafer passes at the maximum clock speed, >then the entrie wafer is packaged as that clock speed. If one (or more) of >the chips FAILS, however, they step down to the next clock speed, and try >again. If every chip on the wafer passes at that clock speed, they mark the >WAFER as that clock speed. But you have a one-in-twleve (if there are 12 >chips on a wafer) chance of a CPU that is really capable of the >fastest clock >speed for that chip design. This process continues until they reach a clock >speed at which all of the chips pass, or they reach a "bottom" threshold >where the cost of producing the chips has exceeded the revenue they >will get, >and they throw the wafer away. So for any given clock speed marked >below the >maximum clock speed for that family, you have pretty decent chance of having >a chip which can run significantly faster than the marked speed, up to the >maximum spped for which they are marking chips. (of course, you may be able >to go faster than that, but in that case, you really are taking a chance). I've heard this story before, usually from people wanting to overclock, and I really question that this actually happens. There's no doubt that Intel tests the CPU's before packaging them, as the packaging costs money and it's stupid to package a failed chip. But, beyond that, I'm not so sure. Every time I've asked anyone who actually WORKS in process control at Intel (and keep in mind that one of the plants that Intel makes Pentiums at is located about 3 miles from where I work) I've been told that this story is an urban legend and a pile of baloney. The problem is that if Intel sets up a production line to make, for example, 500Mhz CPU's, then if the chips don't pass the test then there's too much liability to retest at a lower speed, hoping to get a chip to pass. Instead they throw away the CPU. Consider that if they DID pass the CPU at a lower speed then they would in effect be selling a CPU that's a failed part. Remember that Intel's chips go into a lot of other gear besides consumer PC's that have no warranty, there's military gear and medical gear a bunch of other stuff where they are held to a lot higher warranty standard. Consider that if someone dies on the operating table as a result of a failed Intel CPU then if their lawyers find out that the CPU had failed the test that it was _supposed_ to be produced for, there would be hell to pay. I think that the reason that overclocking works is that it's standard practice to heavily derate electronic parts. This is really a cost issue. For example, it costs a fraction of a cent more to manufacture a diode that passes 1 amps before burning up rather than one that burns up at 1/2 amp, so as a manufacturer your really stupid to not make 1 amp diodes and mark them as 1/2 amp, 2 amp diodes and mark them as 1 amp, and so forth. In some instances, deration is built into the law itself - try getting a licensed electrician to install a 15 amp circuit that's intended to feed a device that draws EXACTLY 15 amps, for example. With chips, they probably design the production line to make 600Mhz CPU's then start the line and mark every single part coming off the line as a 500Mhz CPU. Consider that these CPU's are going into really grungy motherboards, have you ever measured the speed of a typical motherboard with a test instrument? Even without deliberate overclocking it's not uncommon for garbage-grade motherboards to vary 5% or more on both CPU clock and voltage. Doing it this way protects Intel (or whomever) because in a given run you may have parts that will work up to 700Mhz and some parts that won't work better than 525Mhz. >The other wrinkle in this scheme is that Intel is completely free, if the >demand is there, to remark their OWN chips to a LOWER speed. So if demand >spikes for a 300Mhz Celeron, and they have a pile of 450Mhz Celerons sitting >on the shelf, there is nothing illegal, immoral, or fattening about calling >them 300Mhz Celerons and pricing them accordingly. (after all, they >passed as >450Mhz celerons..and don't forget, any given chip in the lot of 450's has a >one-in-ten or so chance of being capable of much faster speeds than even >450Mhz). > Once more I've been told this is a crock, because Intel cannot afford to have a "pile" of CPU's sitting around, the economics of chip prices are such that every day that a completed CPU sits there and isn't bought, the company loses money on it. Now, that's not to say that they don't do this with _old_ CPU designs, I understand for example that there's still 8088 family chips being used in embedded systems. Undoubtedly the tolerances on an 8088 being manufactured today are lightyears better than they were 15 years ago, so you would probably be able to overclock one. It would really be interesting if someone who actually WORKS in process at Intel would be willing to officially confirm or deny what the real story is. > >The other wrinkle is that your motherboard may not be able to handle >overclocking correctly: if they follow Intel's instructions properly, for >example, without considerable effort, you can't change the multiplier. This >means that to overclock, you must also run the motherboard faster than >intended. And the distances involved on a motherboard *are* longer than a >100Mhz wavelength, which can cause all sorts of problems, if your >motherboard >will even allow you to. Then all of your peripherals have to support the >higher clock speeds, becuase all the motherboard does is count 3 100Mhz >clocks and produce a 33Mhz clock for your PCI bus....but if you are counting >3 109Mhz clocks, suddenly you get a 36Mhz clock, and THOSE components may or >may not support running that fast.. ..if your network card, for example, >relies on a 33Mhz PCI clock to generate the 20Mhz 10Base-T carrier, and your >33Mhz clock is off....you might not be able to talk on the network. >(fortuneately, most network cards don't do this, they have their own 20Mhz >crystal for that; it's more reliable...but...if your network card heats up >more than normal becuase it is running faster than normal (more clocks=more >work=more heat) then that will screw up the crystal, too, and might make you >unable to talk on the network...or worse, abnle to talk on the network when >you fire up your computer in the morning, but not when you come back from >lunch in the afternoon, until you shut your computer off overnight, and it >cools down, and starts talking again.....try troubleshooting THAT one!) > Oh yes - this problem first appeared when smart-asses decided it would be cool to manufacture motherboards that would allow you to jack up the speed of the ISA bus past 8Mhz, the standard. There's lots of people that had unreliable 486's that could have been stabilized simply by setting the BIOS to the industry standard speed. > >The moral of the story: if you can't afford to replace your >processor, don't >overclock. If you have money to burn, then its your business, but I might >suggest either 1) buying a faster processor to start with and/or 2) >contributing to the freeBSD project <(}; > Actually, I do think that overclocking DOES have it's place - and that is experimental computer designs. For example I've seen one pictures of one motherboard that a guy rigged up with liquid nitrogen to supercool the CPU that he claimed overclocked a 500Mhz part to 2Ghz without any trouble. Now, in that instance the neat thing about that deal is that supercooling wasn't necessary to extract enough heat to keep the CPU from burning up. Instead, the supercooling changed the electrical properties of the semiconductors to permit the thing to run much faster. One of the problems that I think we have with the computer companies like Intel and AMD is that they know damn well that if they would just change the computer BOX design, we could get a whole lot of performance a hell of a lot cheaper. But instead they seem locked into this idea that no matter how fast the CPU runs, it absolutely MUST do it at whatever temperature you can obtain with a big blob of aluminum stuck to it and a fan, with room-temperature air blowing across it. There seems to be zero interest in investigating changing the operating temperature of the computer itself - at least for consumer designs that is. But the issue is that at some point the cost of the cooling apparatus (whether that be a traditional Freon compressor or something else) will become cheaper than whatever gryrations and exotic materials they will need to enable the current 75 degrees norm to be used. Right now you can get small refrigerators with freezer compartments in them for less than the cost of a new hard disk - I fail to see why the computer case of the future couldn't include an icemaker as an add-on peripheral. :-) At least the overclockers recognize this! Ted Mittelstaedt tedm@toybox.placo.com Author of: The FreeBSD Corporate Networker's Guide Book website: http://www.freebsd-corp-net-guide.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?001301c1338a$7e753b00$1401a8c0>