The Sandia Cooler

in the video it says they sold a license to a cpu cooler manufacturer. i wonder which one.
 
Yeah, it's been out there for a while, nothing ever really came to market yet though.
 
watching that video.... I am forced to wonder if it hasn't showed up on the market due to possible safety / liability concerns with a relatively large hunk of metal with (sharp?) fins spinning about @ 2K rpms inside the case...???


it would only take one genius getting their hand mangled messing with it while it was running
 
also it sounds pretty god damn loud in their demo video ... well not when they shut the motor off...but you know it will really need the motor to run realistically and it is freaking loud with the motor going.
 
I think there were concerns about mounting it in directions other than flat to the ground. Any sort of vertical mounting or even if the case was jostled too much caused problems, I think I remember seeing once.

I love the idea for these kind of out-side-the-box cooling (remember when heat-pipes were a novelty?) but for some reason this one never made it to production.
 
Looks like all this information was posted 6/19 of this year, so I'm not sure where the comments saying it's been out for awhile and nothing came of it are coming from. Does someone have more information indicating this is older than that?
 
I first read about it while I worked in a different office than I do now. I left that office in August of 2011. So, information (that YouTube video) was available on it at least as early as Aug2011. I don't know how late I was to the party, so it may have been out even earlier than that.
 
I wrote a magazine article about the Sandia Cooler and its inventor last year. Having seen the technology first-hand in prototype form, I can (I hope) disabuse some misconceptions about what I saw and the retail technology I hope comes to pass. First, while the prototype fan may have been attitude-sensitive (I didn't get a chance to invert it) the production cooler will not be sensitive to attitude or orientation. Using the permanent magnets in the rotor, the impeller assembly can effectively be "caged" such that it won't run off the air bearing. In a production model, there is no reason it couldn't be run sideways or upside down. Concerning noiselessness, most of the noise comes from the electric motor itself, and again, production-scale optimization should reduce this. The impeller/fan itself impinges on the air quite silently, and is much quieter per heat-watt than the equivalent conventional fan + heat-sink combination.

There should be no more liability surrounding this product than any other fan. One would literally have to stick one's finger into the annulus of the impeller to be injured, and it would likely be equivalent to sticking a finger into any other fan. If lawyers get worked up over this, putting a cage around the assembly would be trivial.

While the inventor has been pretty quiet about progress towards commercialization, in the past year a few significant developments occurred. The technology was successfully licensed to a couple of vendors. Though the original patent was granted in 2009, the Department of Energy's licensing process is onerous and slow, and licensure for this product was complicated by the fraudulent use of Sandia's intellectual property by CoolChip Technologies to win the MIT $200K Clean Energy prize. You can read about this scandal here: http://chronicle.com/article/Who-Deserves-MITs-200000/128810/. As you can imagine, the controversy needlessly alarmed potential partners who assumed CoolChip's use of the technology was sanctioned by the lab. This complicated licensing negotiations.

MIT has investigated itself, and cleared itself of any wrongdoing.

I believe that with licensing behind us, the only remaining delays have to do with A) optimization and B) design for manufacture. Neither process is trivial, and once a design is finalized, it will take some serious tooling to make this technology affordable. In other words, these are physical widgets, not electrons, so it's reasonable to expect a longer development time than you're used to seeing for software.

Having seen the prototype, I can tell you that it is real, and the inventor has staked his excellent reputation on some extraordinary claims about its performance, which have so far been borne out by all critical reviews that I've seen. I believe this is indeed a breakthrough in thermal management, and I am eagerly anticipating its release to the market, because I believe it will be the new standard in air cooling. Simply put, it will be cheaper, quieter, more dust-resistant, and longer-lived than any conventional cooling solution.
 
The biggest problem I have is no where do they show actual heat transfer and dissipation data. Why didn't they install the prototype onto a CPU and see how well it actually works. I am guessing that their device works well in theory, and not so well in practice.
 
You can review the basic science and the prototype technology here:

http://prod.sandia.gov/techlib/access-control.cgi/2010/100258.pdf

The prototype was not on a chip because it was mounted to testing equipment. The prototype's purpose was not to demonstrate commercial applicability; rather, to prove the underlying physical principle. So certain corners were cut. For example, the production model will have to produce its own air bearing hydrodynamically (i.e., by "spinning up" -- a well-trodden mechanical engineering path) but the prototype used a valved source of pressurized dry nitrogen gas to establish the air bearing. This is not a "cheat", but a simple experimental convenience to enable the experimenter to control the gap distance of the air bearing to find optimal settings. Basic science. With these parameters defined on the prototype and the technology proven, the design has to be optimized for effectiveness and cheap mass manufacture, which are the licensee's job, not Sandia's. And just to nail this down, those tasks are not trivial, because you are dealing with two hairy problem sets: thermodynamics and fluid dynamics. Change one little thing, and some baroque-ass equations have to be completely re-run. So I am looking forward to this product reaching the market, but I am not surprised at the pace things are moving.
 
wabernat said:
You can review the basic science and the prototype technology here:

http://prod.sandia.gov/techlib/access-control.cgi/2010/100258.pdf

The prototype was not on a chip because it was mounted to testing equipment. The prototype's purpose was not to demonstrate commercial applicability; rather, to prove the underlying physical principle. So certain corners were cut. For example, the production model will have to produce its own air bearing hydrodynamically (i.e., by "spinning up" -- a well-trodden mechanical engineering path) but the prototype used a valved source of pressurized dry nitrogen gas to establish the air bearing. This is not a "cheat", but a simple experimental convenience to enable the experimenter to control the gap distance of the air bearing to find optimal settings. Basic science. With these parameters defined on the prototype and the technology proven, the design has to be optimized for effectiveness and cheap mass manufacture, which are the licensee's job, not Sandia's. And just to nail this down, those tasks are not trivial, because you are dealing with two hairy problem sets: thermodynamics and fluid dynamics. Change one little thing, and some baroque-ass equations have to be completely re-run. So I am looking forward to this product reaching the market, but I am not surprised at the pace things are moving.
Click to expand...

It has been almost three years since that PDF was made and we still haven't seen a single useful demonstration of the technology. This is an over designed idea that is doomed to vaporware status.
 
The delay was not due to to any flaw in the technology. The cooler was hung up in DOE technology transfer red tape for much of the three years you're referring to. Licenses were finally granted to commercial partners in mid-2012. For manufactured products (i.e., physical things, not software) it's not unreasonable to expect a year or more of R&D and design for manufacture before a product sees the market.
 
Edit: I need to learn how to click links sooner.

It is still very odd that there has been no real new information in 3 years.
 
Last edited:
darkstarcow said:
So in three years they have gathered no hard data, its all theoretical... If the cooler was capable of even a tenth of what is being claimed, they would have real world data to back it up.


There is something very wrong if in three years they aren't willing to test the prototype. It would be much more likely that the prototype was tested, and failed.
Click to expand...

Even in that paper it speaks about the rubbish thermal transfer of the air gap (but that they "cant measure" it in motion).

It's also pretty tiny. Even if it had wonderful thermal transfer, it's a tiny assed heatsink, and would be beaten by a "low tech" big assed heatsink every time without the "danger" and potential difficulty. It might just about beat a stock heatsick, but that's not saying much. Then again, it's taking something simple, and making it needlessly complicated (and with a giant singular point of failure, the motor. Instead of replacing a fan, you'll be replacing the whole thing). If you wanted a more complicated solution you'd just go watercooling, which would cool much much better. If you wanted a low power silent solution, passive heatsinks have been available for a while. It also uses lots of power and makes lots of noise compared to other solutions.

Which is another thing, it wont reduce power consumption. I don't get how a company apparently so qualified can make such a basic error. Energy has to go somewhere. A computer with a crappy heatsink and the same computer with a factory sized water cooler will still produce the same amount of heat. No cooler can make energy disappear. The component might (although doubtful) reduce the heat more than others at the local level, but that will just make the surrounding air hotter, so you cannot claim "places will need less air conditioning" because that isn't true. So any "energy reduction" would have to come from the fan vs motor...and this motor looks like it uses lots more energy than lots of smaller fan motors. So again, this thing will probably increase energy and heat. The only type of cooler than doesn't is passive, as it doesn't use any energy on cooling. So the claims of "7% reduction of energy usage in the US" sounds like "marketing" used to get simple mined investors, that is utterly false.

That they can get something so simple so wrong makes the whole thing pretty doubtful... :p
 
The prototype has been thoroughly tested and the underlying principles proven. That's what the 2009 report is about.

"It's also pretty tiny. Even if it had wonderful thermal transfer, it's a tiny assed heatsink, and would be beaten by a "low tech" big assed heatsink every time" This is the whole point of the product. No matter how big the heatsink, the ultimate bottleneck preventing efficient heat transfer is the boundary layer of air that clings to the surface of the fins. Increasing this surface area or increasing the fan wattage increases noise, expense, power consumption, dust fouling, etc., an exercise in diminishing returns. The Sandia cooler thins the boundary air layer, promoting efficient heat transfer at the air-fin interface. This is the 30x cooling factor that gets trotted out in marketing literature.

The air conditioning claim is not based on using this fan to cool processors, then using a conventional air conditioner to remove that heat. If you review the claims made, they are very careful to precede claims about A/C with such conditionals as: if the technology scales to HVAC applications... and the like, because, as of the last I heard, they had not done the thermodynamics and fluid dynamics necessary to validate the concept. The claims about potential savings in the A/C space are based on the idea of using this technology to eliminate dust fouling from A/C compressors. In other words, using Sandia Coolers instead of a traditional heat-sink/radiator + fan combination to increase the efficiency of air conditioners. This is discussed in detail in the 2009 report.
 
Still, the important thin (to me) is that something new is being tried and tested thoroughly. The whole point of science is to try and then measure. It's not possible to count how many times something ends up working against the common logic or gut feeling, and I may not have a use for this right now but there will be plenty of uses for a very small and efficient cooler in business / enterprise units. I's say about a third of the damaged computers out on the flight line (SDNI) are caused at least partly by simple fouling of the heatsink and fan. Strive on!
 
You must log in or register to reply here.
Back
Top