Liquid Cooled PowerMacs!

Posted:
in Future Apple Hardware edited January 2014
Apple relies on a PPC update cycle that is somewhat less than ideal. Smaller fab process and SOI 'might' provide a significant initial boost. G5 'might leapfrog x86, but inevitably x86 will reel them in and stomp on them before they get going on the updates and subsequent generations, so here's a little solution to all your PPC megahertz myth induced woes:



LIQUID COOLED POWERMAC.



Now, before you roll your eyes too much, lets think about what Apple's strengths are. Industrial design and efficiency. Currently they do a whole lotta work getting a PPC to run cool in a VERY small space, like iMac. What if they used some of that case design/heat exchanger talent to get a PPC to run HOTTER (well, FASTER) than spec in an actively cooled case?



Currently, some chips are certified for different speeds based on the environment (and ambient temp) at which they'll be used. Thusly, a chip certified for use at 500Mhz in a relatively cool office environment, might only be certified for use at a much slower rating in a very hot industrial environment.



The Powermac sounds like a turbine by many accounts, but what if a chassis could be supplied taht didn't upset the lines of a case (unlike those coolermaster PC beasts) that used a sealed liquid cooling setup to cool the machine? Something like a huge cooling plate on the back of the MoBo? Instead of the PPC card facing up, it faces down. When you swing open the case, you don't actually see the PPC, you see the front of the MoBo, and the BACK of the daughtercard because now the surface of the PPC chip (or chips) is pressed against a sealed liquid cooled heat exchanger.



The machine powers up and the cooling system gets going while the procs run at a slower speed, a few seconds after the machine is booted, a logic board reads CPU activity and adjusts temp control accordingly. To keep things quiet, the cooling system isn't quite as aggressive as some of the PC mod jobs, and the case itself is made of sound deadening material. It probably couldn't look like the current case, but it should keep the easy access swing out MoBo access, at least for the portion of the MoBo that contains the PCI slots.



Given a good reliable design, Apple could then use the case to constantly offer powermacs with a reliable 30-40 more MHZ than the air-cooled PPC ratings.



I don't think it'll ever happen, but it would sure make for an interesting Pro machine modding scene.

Comments

  • Reply 1 of 20
    spartspart Posts: 2,060member
    The G4 is already SOI...and .13 will be here at MWNY...



    Me: Yea i need to go to the store.

    Friend: Why?

    Me: Gotta fix my computer.

    Friend: And youre going to the store?!

    Me: yea i need some antifreeze



    <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
  • Reply 2 of 20
    matsumatsu Posts: 6,558member
    <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />



    <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />



    But it needn't be like that. A sealed system that isn't too agressive could be small, quiet, and efficient enough. Though, at that point, it might not provide significantly more cooling than a good fan.



    Still, I have a freezer/fridge in my house that is almost 25 years old. It has never needed a recharge, or given any problems. I don't think any macs sold today will still be in use 25 years from now.



    [ 04-14-2002: Message edited by: Matsu ]</p>
  • Reply 3 of 20
    Is that anything like on the screensavers, Leo pouring liquid nitrogen into a cooling cup ontop of a P4 to overclock it and using a hairdryer to keep the condensation off the board just so they can say they have the fastest or highest framed Quake III system?
  • Reply 4 of 20
    cdhostagecdhostage Posts: 1,038member
    A good refrigeration sstem would do it just fine. Would be noisy though.
  • Reply 5 of 20
    Liquid cooling is simply too complex, it has too many moving parts. The nice thing about fans is that they last virtually forever and the rest of the computer is solid state electronics. But once a liquid cooling system is added, Powermacs would be going in for repairs much more often.



    I see 10 year old Macs up and running all the time...do you think a liquid cooled Powermac would still run in 10 years, without any maintenance? I doubt it.
  • Reply 6 of 20
    serranoserrano Posts: 1,806member
    too complicated?!?!?



    don't be ridiculous, all it would be is a damned radiator, hardly an impossible task for apple engineers.



    <a href="http://www.research.ibm.com/thinkresearch/pages/2001/20010808_cooling.shtml"; target="_blank">IBM is doing it.</a>
  • Reply 7 of 20
    bartobarto Posts: 2,246member
    Water cooling could be done. All it is is a pump pumping water through a waterblock to a radiator (which you then cool). Remember, you still have to cool the radiator with fans. I've seen custom powermac cooling with a fan cooled water cooled peltier cooled G4 533 dual (forget where).



    The thing is Apple would have to get Motorola to have 2 lines of G4s, a standard line and a line which has its clock tested using water cooling, with cooling specs saying you HAVE to water cool it. Imagine if Sonnet started selling 466MHz tested CPUs as 500MHz CPUs simply because Apple cooling was good enough to cope with that! Oh wait...



    Barto
  • Reply 8 of 20
    xmogerxmoger Posts: 242member
    There seems to be some confusion on the different methods here. One option, phase change cooling(refrigerators), is too loud, bulky, expensive, and sucks up a decent amount of power. These are sold by <a href="http://www.kryotech.com"; target="_blank">kryotech</a>. What I think is the best method is water cooling, little pumps, water blocks, and radiators. These are sold by <a href="http://www.koolance.com"; target="_blank">koolance</a>. They are both quieter and cool more effectively than just fans. Unless there is some radical change in processor or cooling tech, some OEM is going to offer these to the general public at some point in the future.



    As a side note that liquid immersion technique that they showed on the screensavers is one of the most extreme, and cool, types of cooling. A couple years ago, I looked into doing it myself. Some people use mineral oil, but 3m makes some better stuff used for cleaning electrical components. I called up a sales rep, when they left a message on my machine with a price quote, my parents thought I was attempting to 'make drugs' with it. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" /> It was a few hundred dollars a gallon at the time.
  • Reply 9 of 20
    matsumatsu Posts: 6,558member
    If I'm not mistaken, Toshiba was also looking into water cooling for it's laptops. :eek:
  • Reply 10 of 20
    baumanbauman Posts: 1,248member
    I have a friend whose office is right next to a freezer room. I keep trying to convince him to throw his PC in the fridge and drill some holes for his cables. But, I guess it'd be a pain to walk into the fridge every time it has a hard freeze (no pun intended).
  • Reply 11 of 20
    [quote]Originally posted by janitor:

    <strong>too complicated?!?!?



    don't be ridiculous, all it would be is a damned radiator, hardly an impossible task for apple engineers.



    <a href="http://www.research.ibm.com/thinkresearch/pages/2001/20010808_cooling.shtml"; target="_blank">IBM is doing it.</a></strong><hr></blockquote>



    Do you actually READ posts before you respond, or do you just see "key words" and fire off a knee-jerk response?



    I didn't say that a liquid cooling system would be impossible, I said it would be too problematic from a durability perspective. I'll try to make it clear for you: fans=don't break easily, last for a decade without repairs. "radiator"=needs repairs often, doesn't last very long. A liquid cooling system that was significantly better than fans would be expensive and unreliable compared to fans. It's simply not a good idea for a consumer computer that should last for many years without any hardware maintenance whatsoever.
  • Reply 12 of 20
    matsumatsu Posts: 6,558member
    PowerMacs are already well beyond a consumer price point. Any costs of a built in cooling solution would be amortized over a minumum 4 year design lifecycle for the case.



    Also, third party cooling systems are quite reliable, but they're ugly, if not too noisy. Reliable technology exists, I'm suggesting Apple re-define it to a new level of aesthetic integration, and quietness. One avenue with which to achieve this is to be less agressive with the cooling system -- smaller, less pressure, less noise (albeit with a little less maximum cooling potential). Another is to sound-proof the walls of the case, giving you a little leeway for more agressive cooling.



    Expense is an issue, but reliability is moot. Small time manufacturers with VERY small R&D have been delivering RELIABILITY (at the expense of BEAUTY) for a couple of years now. Your computer breaks down more often than your fridge. You might own 3-4 refridgerators during your life-time, you'll probably own 5-10 times as many computers (20 if your name is Murbot).
  • Reply 13 of 20
    prestonpreston Posts: 219member
    you have a liquid cooled system



    you have your computer in sleep mode (relatively cool)



    you wake it up, start multitasking FCP3, PS7, AI10 (processor warms up)



    water condenses inside your case (g4 fried)



    yeah yeah yeah use the spongey material that absorbs the water..... oops there goes your easy-access powermac case design
  • Reply 14 of 20
    bartobarto Posts: 2,246member
    That can happen with refidgeration, but not plain water cooling.



    The water cools components (CPU, vid card, IC etc) selectively, and the water is room temp at best, so condensation is not an issue.



    Barto
  • Reply 15 of 20
    This I like.



    But I would go with a passive cooling design, i.e., radiator vs. refrigeration.



    Just think of the unintended computing consequences. "Souping up" a computer could have a completely new meaning. Instead of just changing jumpers around, we'd be buying bigger radiators, pumps, etc. If Apple could add un upgradeable header in there somewhere, I'd be in hog heaven.



    Just a note: liquid-cooled intercoolers for turbos are generally more efficient than air-based ones, but adds complexity as well as cost and weight to the system. But who cares if you're going faster?



    This would go well with my leather-stitched mouse cover (sorry, Steve) and the fuzzy dice on my monitor.
  • Reply 16 of 20
    tsukuritetsukurite Posts: 192member
    It's interesting to note (well, it is to me at least) that the first Xeon based servers from HP used hollow, monster heatsinks that were filled with alcohol. There were still a sh*t pot of fans, but it was pretty cool all the same, from a geeky perspective. Some thing like that might be doable if you had a way for natural convection flows to be used to better effect. Like the cube's design. How much sense does it really make to cover the top of the machine when that's where the heat wants to go?



    &lt;mimics engineer&gt; Oo, let's add some fans! &lt;/mimics&gt;



    Gad, it's still late. sorry.
  • Reply 17 of 20
    prestonpreston Posts: 219member
    i wont buy a loud computer. Us creatives don't like distracting fans. The quicksilvers are loud enough... i dont want a barfridge on my desk
  • Reply 18 of 20
    matsumatsu Posts: 6,558member
    I don't doubt that it would be a quite heavy beast, compared to other tower cases, but it needn't be any louder than the current QS case. Sound deadening materials could be sandwiched between the steel inner case and the external polycarb panels. Building the inside out of thicker aluminum plate would also help dampen a lot of the higher frequency sound. It would hum (if done right, just at the threshold of hearing) but the whine would largely disapear. Just look at car exhausts. A steel muffler is ussually raspier than an aluminum unit, which is often a bit bassier. The noises coming out of your computer are mostly higher freq noises, a thicker aluminum plate construction with a layer of sound absorbing material between it and the external plastic would make a much quiter case than the current design.



    [ 04-17-2002: Message edited by: Matsu ]</p>
  • Reply 19 of 20
    [quote]Originally posted by Matsu:

    I don't doubt that it would be a quite heavy beast, compared to other tower cases, but it needn't be any louder than the current QS case. Sound deadening materials could be sandwiched between the steel inner case and the external polycarb panels. Building the inside out of thicker aluminum plate would also help dampen a lot of the higher frequency sound. It would hum (if done right, just at the threshold of hearing) but the whine would largely disapear. Just look at car exhausts. A steel muffler is ussually raspier than an aluminum unit, which is often a bit bassier. The noises coming out of your computer are mostly higher freq noises, a thicker aluminum plate construction with a layer of sound absorbing material between it and the external plastic would make a much quiter case than the current design.

    <hr></blockquote>



    ARrghh...



    Sorry-- I don't like sound deadening in cars. The muffler is just one aspect, but there could literally be hundreds of pounds of sound deadening in a car (passenger compartment) that seriously saps a cars performance.



    To me this bandaid engineering is endemic of poor design-- a loud motor, wind/road noise, etc., needs to be masked, so you add deadening material.



    Why not make a quiter motor?



    I know this doesn't apply to the principle of this thread, which is, we have a quite motor-- what happens when you soup it up?



    More heat.



    The real answer is to get a better (i.e., faster) chip that can run at today's power dissipation level. Intel is actually doing this-- not in the desktop, but in the laptops. Mobile P4s after faster than the desktop G4. Think about that for a minute.
  • Reply 20 of 20
    matsumatsu Posts: 6,558member
    You may not like sound deadening materials, but within the limits of cost conscious manufacturing, they are something we have to live with. Take a very competent (and quiet) sports sedan, maybe a BMW or a Mercedes (Steve likes that comparo) and strip out every last bit of sound deadening material. (the wheel arches, the headliner, door seals, carpet and under carpet materials, everything) Dynamically, it's every bit as good as before, actually much better -- a few pounds lighter -- but it is a heck of a lot noisier than before, even with those nice smooth engines.



    Now we're talking about a computer, it doesn't have to change lanes or manouver faster, it just has to sit on your desk. Sound deadening will make it more pleasant in that situation, and it is an acceptable and cost effective way of quieting ANY design even further, especially for something that doesn't incur any real penalty from being on the porky side (unlike a car).



    Of course, at the base of this discussion, you're right about the real answer WHICH IS A FASTER, QUIETER, MORE EFFICIENT ENGINE -- whether that be a CPU or a V8.



    However, there isn't any real reason why Apple industrial design can't apply such a cooling scheme to faster/more efficient chips and make ANY chip it uses run that much better.



    [ 04-19-2002: Message edited by: Matsu ]</p>
Sign In or Register to comment.