Slashdot Mirror
Silent Pump for Water-Cooled PCs
Wycliffe writes "New Scientist has an article about a silent pump for water-cooled PCs.
The system, developed by a Californian start-up company, aims to silently solve the problem that the faster chips get, the hotter they become."
That's really cool.
I've often wondered whether we couldn't attach a Stirling engine directly
to the surface of the processor and recover some of the energy being lost
as heat and turn it into electricity. If we could do that then I could have
a cooler laptop that has a longer lasting battery since processor heat is
being used to power the processor.
Anything that stops my laptop burning my private parts and makes the
battery last longer would be very nice. If at the same time it makes
no noise (especially compared to the lawnmower like noise my Dell laptop's
two fans make) it would be even better.
John.
"Even if all the technical details are ironed out, I think it will be five years at least before fans are replaced. They are still the cheapest option."
Seems to be a way off. I wonder if in 5 years we will have different processors, where this will not be effective. Think little bacteria or DNA or something organic as a CPU.
Rosco: "If brains were gunpowder, Enos couldn't blow his nose."
Really, we don't need better cooling. Every heat problem we look at revolves around better ways of pumping heat away from the CPU. As soon as we come up with something cool (pardon the pun), the chip manufacturers have implicit carte blanche to produce hotter chips.
We Need Cooler Chips. We need CPUs and GPUs that consume maybe 10W instead of 80+W, and then we can go back to heatsinks perhaps with small fans. Looking back on my first x86 machine (a 486), I discover that it was the last processor Intel certified for use without a heat sink (or maybe without a fan--something like that). Now we have BIG copper heatsinks, monster CPU fans as well as extra case fans and dual-fan power supplies, and the companies are starting to look at liquid cooling as a mainstream "solution."
When will it end? At this rate, we'll actually be maxing out 500W power supplies in a few years. Half a kilowatt is too much power to be drawing for a computer! (and consider that it doesn't even include the monitor or peripherals.)
Let's start leaning on Intel and AMD, and get them to reduce the power consumption rather than giving us meaningless MHz increments.
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
before the I get modded to oblivian, I've looked everywhere for the answer to this question... all I find is overclockers bragging about their 15C AMD1600+ OC'd to 3.5gHz, which doesnt help me...
I run an AMD 2600+, and it sits about 53-57C (its on a chaintch zenith 7jns board)... is that too hot?
I was thinkig of getting a water cooling system, but I dont know if its too hot or if I am just not used to a high powered AMD processor/mobile heating station...
This is my sig. Its pathetic.
They've been doing that for a while. They also do chipsets. Now you're also starting to see watercooled hard drives.
But I still remember back when I was in college and I got to take a tour of Cray headquaters. I can still remember seeing one of the first Cray-2's there, in its clear shell. It was about 3 feet tall and 3 or 4 feet in diameter, and the whole computer sat inside completely submerged in fluid (not water). You could see little bubbles rising up through the densely packed circuit boards and wires. And nearby was a really cool lighted fountain. So cool. Of course the engineer there said it created so much heat that there was an entire building out back which was just the heat exchanger for the fluid.
In The Hunt For Red October they were using Magneto Hydrodynamic drive, where the magnetic asymetry of the water molecule is exploited by using an intense magnetic field to align the water within the drive tube (much like what is done in a MRI), and then moving/pulsing the magnetic fields along the drive tube to push the water through it.
There is no screw. (except as a backup.)
If you've ever seen pictures of the levitating frog (being held up by magnetic fields), it is the same principle.
What is the difference between a small revolutionary change and a large evolutionary change?
If I recall correctly, it's a simple plastic tube, with a motor driving a pair of rollers to maintain flow, leaving the system completely closed (usually these kinds of pumps are used in heart/lung machines, dialysis machines, etc). Also, having seen such a pump in action before, they're incredibly quiet.
e dotasp/PagEnd/QStr/TargetInventoryID/Eql/36054/Pag e.htm
The closest reference I could find is at http://www.appraisalmedical.com/PagBgn/ProductPag
There should be a way to manufacture a similar pump for far cheaper, since you wouldn't need as accurate a pressure control, or have to excessive monitoring.
The benefits are:
(1) Virtually silent, no impeller noise.
(2) Pump/Motor are completely isolated from the coolant fluid, leading to a lower chance of failure due to pump contamination/oxidation/short circuits.
(3) The pump speed and thus the fluid moved can be controlled with a potentiometer. As such, with a bit of creative work, you could conceivably design a failsafe mechanism to increase coolant flow when temperatures appear critical.
(4) Replacement parts *should* be cheaper, you can replace a roller, the hose the rollers ride over, or the motor individually, rather than the entire unit.
Just because you can mod me down, doesn't mean you're right. Shoes for industry!
Interesting to think about this in terms of Moore's Law. ML states that costs of costs of chips will come down (and/or transistor density rises.) Regarding costs, what good does it do to make chips cheaper if the costs of electricity negate the benefit? Some CPUs are exceeding 100W consumption. Over the course of a 3 year lifetime, at 15 cents per kWh, we get a cost of: 100 W * 3 years * 365 days / year * 24 hours / day * $0.15 / kW - hour * (1 kW / 1000 W) = $394.20 So that Pentium 4 that cost you $300 actually is going to cost you that again just for electricity!! Not everyone leaves their machines on all the time, and YMMV, but this is only going to get worse over time. It will be interesting to see how CPU manufacturers repond with more efficiency, rather than just more transistors.
A 1GHZ C3 isn't even close to the capability needed for audio production, either soft synths and effects or multitrack recording, let alone doing both at the same time.
The processor in a Nehemiah board performs about like a K6-300, in my experience with them. It does have the silent factor solved. I've already tried the mini-itx approach and have been thorougly disappointed with the performance of the C3 processors.
-fb Everything not expressly forbidden is now mandatory.
Dear Slashdot community,
This product sounds fantastic, but it's not currently available on the market. How about some links to water-cooling kits that are available for purchase now? Which ones are recommended? Which ones not?
As a bit of backstory, we have a couple of new AMD Athlon computers here in the office with excessively loud fans. They're so loud that most of us (myself included) choose to do our work on old laptops, only using the desk computers when we need to calculate tabulations for our research studies. It's a real shame to see good computers wasted, so I'm thinking that watercooling may be the way to see that our investment in hardware actually gets used. Interesting/informative watercooling kit links would be appreciated.
Kind regards,
Michael Judge
SurveyComplete
Well said, well said.
/. we have a crowd of 98% hobbyist, do-it-yourself types who keep a close eye on technology and how the guts work.
I see the geek reaction to water cooling as very similar to the hobbyist reaction to everything useful in recent history. Let me elaborate.
Think back about 100 years to the start of aviation. The really dedicated aeronauts built their own planes - in fact, the Wright brothers were basically hobbyists compared to Mr. Langley, who was well-financed, especially by the government. What was the reaction among the hobbyist crowd to mass-produced planes?
Think back about 120 years to the start of the practical automobile. The really dedicated enthusiasts built their own cars. What was the reaction among that crowd to the Model T? "Junk. Never sell well. Impractical." Let's take that one a step further. "Computer controlled ignition and fuel injection systems? Preposterous. How can I tweak it?" Well, where are we today? When's the last time you drove a car with real points or a carburetor?
Think back about 50 years to the early days of home audio. The really dedicated geeks built their own systems - remember the Heathkit stuff (or am I too old for this crowd)? Build your own TV, you could. Same thing for stereos, etc. What was the typical audiophile reaction to the idea of a mass-produced stereo? "It'll be junk. Never work. Sounds lousy. No control over the details." Sorry to mess with that world view, but walk into any WalMart and you can take home a stereo system that sounds far better than many of those hobbyist's systems, and costs about 1/10 as much.
Think back about 20 years to when PCs were solely the domain of either the goverment/high-end research facilities, or hobbyists. What was the reaction to mass-produced PCs? Remember IBM's reaction to Bill Gates? Even more to the point, how about the hobbyist reaction to Bill Gates?
What I'm getting at is this: the folks who do the most whining and complaining about anything going mainstream is the hobbyist crowd. Why? I believe it's because the hobbyist perceives it as a threat to his control over his hobby. It's inconceivable to him that something could be mainstreamed successfully. Certainly, here on
Okay, but what's the reality of the situation? Simply put, does your grandma care how her PC works, if she even uses one? No, she simply cares that she can turn it on, use it, and turn it off again. The chance that she'll EVER see the guts of a home PC approaches zero. Same thing with a stereo, or a car, or a plane. Most non-geeks not only don't KNOW how most things work, they emphatically don't WANT to know.
So I see water cooling the same way. It WILL work, it WILL be accepted, and it WILL be part of many systems soon, regardless of how many hobbyists think it's a mistake. Because OUR desire for control over the innards of the PC is completely irrelevant to the mass market. PCs are simply not built for us anymore - they are built for Joe Sixpack and your grandma.
As proof of this, take the average notebook PC. They're accepted widely, but almost completely impossible to upgrade, or hack, or tweak. Well, water cooling is going the same way. The manufacturers don't care if water cooling is good for the computer hobbyist, as long as they can make it work in a mainstream, sealed-box PC. And that's where the average home PC is headed - a sealed PC that is effectively non-upgradable.
--Brandon / Split Infinity Music
I don't mention the Pentium IV because I truely believe it is a step backwards, and that eventually it will be replaced by the Pentium M ( Banias ).
What I was trying to say is the most efficient processors we have today are much more efficient than a 486 ( partially attributable to die shrinks, but also due to advances like pipelining, branch prediction and OOOE to reduce wasted cycles, and dynamic power across the die on moden chips like the Pentium M ). If manufacturers could actually sell on efficiency instead of pure power, we'd see more of these architectures. Unfortunately, "cheap performance" rules an early market.
Think about it this way: back in the early 1900s Ford sought to give everyone a car with the Model T. Eventually, they saturated the market and nobody wanted the rough but cheap Model T anymore. People wanted stylish cars, more areodynamic cars, larget engines, more efficiency etc...
I believe the market is only beginning to realize that the desktop market is saturated, and is now replacement-only. The only reason we havn't seen a wide pickup of purchases for more efficient computers or more stylish computers is because manufacturers are ignoring the desktop market saturation and bleeding production over to laptops.
Once the laptop market crashes in a few years, I expect manufacturers to finally 'get it' just nlike car makers in the 1920s did, and offer a wider range than just "cheap performance".
Man is the animal that laughs.
And occasionally whores for Karma.
Great. Now your house is full of toxic gas.
The maximum permissible continuous exposure limit to ozone in air is 0.1 ppm averaged over an eight-hour work shift. Bear in mind, it is possible to detect ozone by smell at levels as low as 0.01 ppm.
I recommend full faceplate breathing apparatus (NIOSH/OSHA), if you have to stay over eight hours in the contaminated area.
This water pump with no moving parts seems great. I'll be the first to buy one once the price drops from the $1000 that it's likely to start at down to like $50. It doesn't solve the problem of creating a quiet pc though.
Historically water cooling has been used more for overclocking than for quiet PCs. There have been some experiments in passive, convection based, water cooling. One experiment was particularly interesting, but I can't find the link. One trick is to use very large diameter tubing. The larger the diameter of the tube the slower the water can move and still cool the CPU.
Unless you attempt some kind of open system that gets water directly from the house plumbing or groundwater, you still need a radiator and fan to cool the water. The advantage is that you can go to a larger fan or even two or more fans and a much larger heatsink (radiator) than is possible directly on the CPU.
With an open system, you could put that waste heat to work as a water heater with the addition of some kind of storage tank.
Quite an experience to live in fear, isn't it? That's what it is to be a slave.
does this firm include a sump pump for the times the system 'leaks'? if they asked me, i would use another inert liquid/gas/refrigerant as a coolant, instead of a conductive one...