anything or anything of consequence? nacl back* rather HEMI: It was my understanding that the issue was just getting power consumption controlled, not so much technical inability heh, massively parallel game consoles and lazy game devs....lot of bugs gonna be on their way sleep: from what I understood it was a technical issue that they couldn't overcome...Basically, the power consumption levels they had to go to was too high and was burning up the CPUs. Intel predicted that they could lower the "turn-on" voltage for their logic. They haven't been able to do that. If they could, then the threshhold voltage would go down, overall CPU voltage would drop, etc. As it stands, they haven't been able to lower that turn-on voltage. poor poor Intel LIke there isn't enough differential between ON and OFF (due to leakage) to be stable at lower voltages? So if there was a nice breakthrough in insulator tech, Netburst could be competitive again. No...Like the power required to turn a transistor on is too high. There's consequences for packing in transistors so close and having the voltage be a certain level. netburst did what it was supposed to, it made intel money and held being the fastest up until the opteron the mitsake was in not preparing something to fight the opteron off sleep: yeah, that was the original idea. when Intel made their 10GHz prediction they were expecting a breakthrough in insulator tech that would drop the turn-on voltage. but Netburst was supposed to scale to like 8GHz 10GHz. Ah. well, it was a gamble as it turns out Yeah. Strained silicon and silicon-on-insulator are methods to help lower the turn-on voltage, but they didn't work well enough to get the turn-on voltage down to what Intel needed it to be. i see what the solution is the processor should ship already turned on not funny and then it never turns off, MS just makes the screen black when you shut down The way I had it explained is that cramming transistors in too closely and running a high voltage would amount to lightning across the top of the CPU...Breakdown of the silicon, etc. oh well, time for photonic processors. it wouldnt be a problem if we made our transistors out of lego CMOS and used diamond heatsinks not just diamond, but one-way conductive diamonds. diamonds are great conductors of heat carbon nanotubes they dont exactly dissipate the stuff AssKoala: I know. That's why I said it. ahem dissipate conduct different No shit diodemonds? :) sherlock erm, isn't dissipation as far as processors go just conducting the heat off to air or water or something? They should just build a TEC on the top of each core --> Epic| (~awd@dialup-4.240.111.58.Dial1.Phoenix1.Level3.net) has joined #2cpu it's what happens to the heat after it gets to the air or water. in place of the nickel-coated copper IHS slug sleep: that'd be a pretty powerful TEC for each core and 2 of them in 1 tiny area, that'd be a pretty huge heatload It won't be too long until they have to do something like that, I bet. well so if diamonds are good at conducting, why are they bad at conducting to air or water? because u can transfer the heat but that doesnt mean that you are dissipating it aluminum is better at heat dissipation than copper, that is, its better at cooling off from my understanding, this makes a difference apparently the sides of the diamond tend to want to keep the heat to itself. yes thats why its a good conductor, itlltake the stuff in itself and, well, conduct afaik cooling off is just a matter of giving your heat to something else i guess i could see how a material might be good at accepting heat, but not giving it up yeah, the diamond passes it off to something else saying "good conductor" implies symmetry there though tho to air it may not no, it doesnt it does to me maybe i'm a noob basically it is, you just want to push the heat to an area that has greater thermal capacity. toughness and strength dont mean the same thing well its not my fault if someone misuses a word i.e. use toughness to describe strength http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/heatra.html Yes it is. All your fault. think about diamonds like a PVC pipe with hot water in it. then compare that to a copper pipe with hot water in it. conduction versus convection thats what it is the copper pipe is more "lossy" than the pvc pipe. Mmmm, PVC... that the difference between diamond and copper. convection is a whole other process than conduction Although for fireworks tubes you want to use fiberglass. no shit diamonds are good conductors when i said dissipation, i should have said convection ok. i never pictured heatsinks doing convection, i pictured them as conducting to the air. and then you push the air out of the case or whatever Heh. no, the arctic silver between the heatsink and the proc conducts copper is better at conduction than aluminum but aluminum is better at convection, i think Diamond = Teh win thats why a lot of good heatsinks are a mix of copper and alum that doesn't even make sense to me but there are problems of course in doing such things that cancel out the efficiency Zenith: until we can change graphite into diamond, it'll always be expensive though :P all the alphas are copper cores and aluminium fins. aluminum in my hsf doesn't flow around my room changing densities what Pinski - They can grow diamond semi-conductors... it just contacts air and conducts heat to it, then a fan nearby blows the hot air away the swift tech coolers are copper bases with aluminium pins for fins. Japanese and German scientists have done it. no it transfers heat to the air through convection Zenith: at what cost thougH? the heat is passed from cpu to hsf through conduction Pinski - Whatever it costs to take carbon to 5,000c at. :) Expensive. Zenith: soo it's just like Fusion! and the heat u get when u walk outside is heat through radiation you mean it transfers heat to the air that is far away by heating nearby air and relying on it to flow away? Pinski - Hot shit, jah. it exists, but it's not plausable i don't think you are using convection right. I don't believe your vocabulary. now all we need is somebod to create Cold Fusion! Thermaltakes Si-93/XP90/XP120 rock using nickel plated copper heatpipes, then the fins are all aluminum. They work, WELL. xwred1_ - then read the fucking website i linked VERY WELL, with the heatpipes taking almost all the heat away from the processor. Im not a thermodynamics master ok mattsteg might be tho My Si-97(Whoops) has no fins near the proc at all. ... in my physics class, i didn't get the impression of convection as how you are using it then read the page I linked it i am read it what's all the hubbub? heat conduction versus convection what about them? first sentence "Convection is heat transfer by mass motion of a fluid such as air or water when the heated fluid is caused to move away from the source of heat, carrying energy with it." how the fuck does the heat from the hsf get into the air? you get one guess and a material being better at one than the other uh air is a fluid convection == fluid carries the heat away with it, conduction == heat diffuses independently of matter physics and convection?! that's chemistry not physics well, that was the answer soim not sure what the problem is you still conduct heat into the air so? he's saying the hsf "convects" its heat into the air and it's not that simple xwred1_ no, thats not what I meant heatpipes move heat via convection well that's how I read it. my apologies. convection = transfer by air i believe convection is the flow of a fluid due to density change by heat conduction = direct contact theres a whole friggin class I gotta take on thermodynamics, so its fairly complicated...I would be taking it now but it got cancelled..fuckers NeoSadist - not air, fluid NeoSadist: not that simple lol * HEMI wants to move heat via gravy train. ah, fluid conduction occurs in fluid too HEMI: the gravy train! and what mattsteg said HEMI: LOL xwred1_: not true convection is any fluid motion regardless of cause well ok how do you get heat into the fluid like a pot of water on a stove Conduction is me running to the bathroom on fire...I'm moving heat with me. :) it was always my understanding that convection was how the heat moved and conduction was how well it moved. heh Black_Baron: nope the water on the bottom of the pot gets heat conducted to it from the pot, then it convects through the rest of the water, no? HEMI: naw, you're convection heat err convecting I thought convection was the flow of heat through matter. *shrug* it is matter, but more of a fluid matter * mattsteg assumed HEMI was made of matter The matter moving around because of that heat was a different---but related---system. fluids act different than solids I think we should all take a class in thermodynamics and not cause mattsteg to pull his hair out... fluids move, solids don't move much (they vibrate, etc, quantum shit, but not like a fluid) * NeoSadist took biology :( look. the immediate heat transfer from the surface of the hsf to the air particles that are actually contacting it is conduction, correct? xwred1_: mostly yes, but the rate of conduction is affected by the rate of convection away from the surface sure sure i can understand that also, there's radiative heat transfer that in some situations matters 2 a : the circulatory motion that occurs in a fluid at a nonuniform temperature owing to the variation of its density and the action of gravity b : the transfer of heat by convection -- compare CONDUCTION, RADIATION 2 a : transmission through or by means of a conductor; also : the transfer of heat through matter by communication of kinetic energy from particle to particle with no net displacement of the particles -- compare CONVECTION, RADIATION b : but at that immediate snapshot, the heat is moving by conduction. the air flowing away wouldn't have much to do with the hsf material, beyond the material just conducting the heat really fast so the air heats up and flows elsewhere more quickly xwred1_: no, because the atoms move around in a fluid, and therefore hot molecules move rather than staying and transferring heat directly. in a fluid, i belive the atoms are individually heated and move. like if you've ever seen hot water move when it's boiling hot, it has small water currents (as in currents = moving water, not = electricity) i believe that's how it works, but i could be wrong www.google.com conduction == through solids, then? xwred1_: the air has a *lot* to do with things HEMI: sorta In space, no one can cool their CPUs. *screams* conduction is like diffusion LOL in space, noone needs to cool their CPU's its fairly cold there I hear mattsteg, sure, the air does. but the hsf material shouldn't have too much bearing on the air flow should it? copper vs. al vs. lead vs. diamond vs. whatever hmmm, how do i make a folder on a place i have hosting for open to the public to look at? depends on what side of the sun it's facing, iirc like if you put dye with teh same density as water in still water i have ssh access along with ftp, i just don't know how to make the folder open to the public lol ok fine i'll solve this * NeoSadist gives himself a lobotomy Uki: what kind of hosting? xwred1_: different materials have different conduction and convection rates. web hosting on a linux machine HEMI, see, i don't even comprehend the idea of a convection rate xwred1_: the hsf material determines how hot the surface of the hsf is k and u make your web dir and put stuff in it HEMI, the closest thing to me seems to be the idea of how it might flow around Uki: chmod 755 directory and it's on the web mmk, thanks You might go 744, even... mattsteg, yes thats what i thought But if you don't make the directory executable no one will be able to cd in to it and see the information in there without an absolute path. xwred1_: think of molasses - it's not going to move much so any convective effects are going to be minor i did that before, and i wanted the whole folder to be viewable rather then giving out the link to each file Uki: don't put a index.html file heat's mostly going to transfer by conduction, until it melts of course and it should have a directory listing i didn't, and it still didn't work unless that is disabled i dunno otherwise you're going to have to make your own directory listing no mod_autoindex? mattsteg, earlier AssKoala pretty much said diamonds are good conductors, but bad convectors. which doesn't make sense to me. xwred1_: have you ever melted a diamond? well almost any solid is a bad convector yes except maybe like butter or cheese or something how high a temp is needed to melt a diamond which aren't solids Convection (basically) happens in fluids, it seems. diamonds, uh, aren't solids? no, he means butter and cheese oh * NeoSadist slaps himself mattsteg, how does diamond not flowing around much in a molten state have anything to do with its performance as an hsf though? that seems irrelevant. I would think an hsf's performance is mainly determined by its conducting ability. because precisely what it does is conduct heat to the air contacting it. Arctic Silver LARD... some people don't reference their statements enough for poor old me HEMI: LMFAO xwred1_: if it doesn't flow, it doesn't convect xwred1_: ability to retain heat? holy shit suse is taking forever the whole convection deal was a red herring though i should've fucking downloaded the entire fucking thing first yeah, i get an error 403 mattsteg, well thats what I thought. convection of the hsf material is irrelevant in an hsf discussion isn't it? HEMI - I love the dansdata.com thing where the jelly and toothpaste gave better temps than the arctic silver ;) and its setup with chmod 744 of course, it didnt last a second HEMI, heat retention is a function of conductivity... a bad conductor retains it due to not conducting it away to whatever touches it xwred1_: well, heatpipe designs incorporate convection internally, and all have external convection effects i'm thinking of just your standard chunk of metal + fan setup the external convection is gonna be pretty much the same whether your hsf is a block of cu, al, or diamond. sorta, and that's really beyond what we're discussing that turns more towards aerodynamics usual geek nitpicking I guess well, we were starting with heatsink materials and the sort of flow you establish so things like surface roughness can have an effect i jumped in when people started talking about hsfs out of diamonds, someone else said "oh they are good conductors but bad dissipators" that's silly thats what I thought. my understanding was dissipation from my hsf was conducting to the air in contact with the surface if you can conduct well you can dissipate well unless you have difficulty establishing an interface as in like mechanically keeping air around, or nearby or whatever? it'd be tough to make fins out of diamonds, for example so you can't get as much surface area easily yeah, and it would probably be too expensive for Joe User well, heat transfer is proportional to surface area and temperature difference just the fact that they do crazy stuff for surface area always seemed to reinforce the conduction argument for me "cnduction argument"? mattsteg: so what's all this i've read about copper conducting heat better than aluminum, but aluminum releasing it to the air better (and not just on stupid OC sites, also on science ones) umm, that's BS the argument that a hsf conducts heat to the air contacting it thats what ive read as well, the BS must come in that aluminum is easier to machine I suppose aluminum may have a lower thermal density, but in steady operation that doesn't matter for shit so you can get more surface area mattsteg, btw, is conduction symmetric both ways? yes yeah so that cu vs. al thing just mentioned is easily bs then it's completely isotropic in normal materials well, the point was still right, you wouldnt make the entire heatsink out of diamond since you cant quite make the fins the reason was wrong