[HN Gopher] Solid state active cooling on a Mini PC
       ___________________________________________________________________
        
       Solid state active cooling on a Mini PC
        
       Author : PaulHoule
       Score  : 77 points
       Date   : 2023-06-02 13:54 UTC (9 hours ago)
        
 (HTM) web link (www.theregister.com)
 (TXT) w3m dump (www.theregister.com)
        
       | nottorp wrote:
       | [quote]Frore believes small is beautiful and hopes many more
       | products will follow, selling the AirJet for tightly-packed
       | devices in which passive heat management means lesser CPUs are
       | deployed, or silicon is throttled back to avoid
       | overheating.[/quote]
       | 
       | How about leaving the throttling on and making the device even
       | cooler and quieter instead?
        
         | packetlost wrote:
         | User controllable fan-curves should be standard for any
         | actively cooled device.
        
           | nottorp wrote:
           | Even when they are the software works only in Windows...
        
             | smolder wrote:
             | For motherboards you can typically set them in the BIOS, so
             | no software needed there. For GPUs, IDK what exists outside
             | of windows, but I'd expect there to be some kind of utility
             | for it in the Linux world.
        
               | 0cf8612b2e1e wrote:
               | Fine tuning the fan curves in the BIOS is a pretty crummy
               | experience requiring multiple reboots to test the system
               | under load and play with the configuration. If there was
               | a Linux compatible tool to tweak them on demand, I would
               | love to know.
        
         | Zemtomo wrote:
         | That already exist. This product is not for the case you
         | describe.
        
           | nottorp wrote:
           | ... is not marketed for ...
        
       | Havoc wrote:
       | Struggling to see how this will not get clogged up in a week
       | flat. Tiny membranes moving air through a narrow slit sounds like
       | bad news for even moderately dusty houses. Hopefully I'm missing
       | something there though because fanless would be ideal
        
         | judge2020 wrote:
         | From the LTT video linked elsewhere in these comments
         | (timestamp 6:13), it has insanely high static pressure, which
         | means you can install dust filters on the actual laptop intake,
         | in addition to the AirJet itself having a dust filter, without
         | sacrificing airflow. In addition, supposedly there's a reverse
         | airflow mode that would allow it to blast dust out of the dust
         | filters.
        
           | Havoc wrote:
           | Oh that's actually clever. Thanks for sharing
        
       | belltaco wrote:
       | LTT hands on video with a lot more details and a teardown
       | https://www.youtube.com/watch?v=vdD0yMS40a0
        
       | yboris wrote:
       | Quick read link:
       | https://www.theverge.com/2023/5/23/23733592/frore-airjet-zot...
        
       | 2OEH8eoCRo0 wrote:
       | How well do these deal with dust?
        
       | droopyEyelids wrote:
       | Last time these were in the news I was curious how they'd handle
       | environmental dust. Did they ever release information about that?
        
         | enragedcacti wrote:
         | The units have paper filters on them and they have said that
         | the very high static pressure means that laptops designed
         | around them could start using external filters without
         | hampering the performance like they would on existing laptop
         | cooling solutions.
         | 
         | The LTT video quoted something like a 5+ year service life for
         | the filters. That could probably be extended with disassembly
         | and cleaning or potentially by replacing the filters which seem
         | to basically just be a sticker that goes on top of the unit.
        
         | alex7734 wrote:
         | In Linus' video it was said that it reverses flow for a second
         | to remove the dust from the filters
        
       | bryanlarsen wrote:
       | Related: passive cooling case that can handle a 4080:
       | https://www.anandtech.com/show/18887/streacoms-sg10-case-can...
        
         | karmakaze wrote:
         | I can't imagine wanting to run this in the summer. Unless I
         | really needed to see output in realtime, better to use it in a
         | remote data center.
        
         | nottorp wrote:
         | So many transparent panels. If those were radiators too it
         | could handle four 4080s!
        
       | mrangle wrote:
       | Been seeing press releases, like this article, for this tech over
       | awhile now. I'm now in a "start selling it or it didn't happen"
       | phase of attention to it.
        
         | pkd wrote:
         | It is available in a zotac mini PC.
        
       | minhazm wrote:
       | I saw this video[1] of them at CES and was wondering when people
       | would actually be able to get their hands on one. Cool to see
       | they'll actually be in a product soon.
       | 
       | [1] https://www.youtube.com/watch?v=YGxTnGEAx3E
        
       | fnord77 wrote:
       | how is this "solid state" if it is just a fancy flat fan?
        
         | amiga-workbench wrote:
         | Because its using piezoelectric elements to move the air. There
         | are no motors or bearings in here.
        
           | SV_BubbleTime wrote:
           | " We were told the vibrating membranes are made with
           | techniques similar to those used for production of
           | semiconductors and screens - hence the "solid state cooler"
           | description."
           | 
           | In other words, not at all solid-state. Some process they
           | used to make a component reminds them of solid state
           | manufacture techniques.
        
             | enragedcacti wrote:
             | Would you call a quartz clock "not at all solid-state"?
             | After all, it uses a piezoelectric material that vibrates
             | to cause an effect.
             | 
             | I think that description would be technically correct while
             | also being very misleading to the average person, and I
             | think that's why we've seen the colloquial definition of
             | "solid-state" expand to effectively mean "non-mechanical"
             | in some domains.
        
               | SV_BubbleTime wrote:
               | No. Not at all. I would call it crystal resonator.
               | Because it's not solid state, there's no semiconductor. I
               | think you might be confusing that "it looks like a chip
               | kinda".
               | 
               | If you wanted a solid state clock source, that is a MEMS.
               | 
               | And further, no one calls, Crystal resonators "solid
               | state" for marketing purposes, which is all that's going
               | on here.
        
               | enragedcacti wrote:
               | I feel like we've reached the all too common HN cliche of
               | having to address that words have multiple meanings and
               | those meanings change over time. "solid-state" only means
               | "semiconductors" because solid semiconductors replaced
               | gaseous state devices.
               | 
               | If we operate by that definition these devices are very
               | much solid so perhaps from context we can choose infer
               | that they mean "non-mechanical" and not "made of
               | semiconductors" rather than getting wrapped up in
               | pedantry.
               | 
               | > And further, no one calls, Crystal resonators "solid
               | state" for marketing purposes, which is all that's going
               | on here.
               | 
               | Yeah, a company is trying to communicate the benefits of
               | their product using words that convey them i.e. HDD->SSD,
               | fan->AirJet. We can shake our fists at the sky in
               | righteous anger or we can acknowledge that there isn't
               | really any harm to using the terms that evoke the right
               | meaning for most people in a new context.
        
               | SV_BubbleTime wrote:
               | It's not pedantry to call things what they are.
               | 
               | A crystal and this cooler - have moving parts. And don't
               | fit even a laissez-faire definition.
        
               | enragedcacti wrote:
               | so does a MEMS if that's the definition were going with.
        
               | CyberDildonics wrote:
               | I'm with you, technically all heat is vibration of atoms
               | so nothing is really solid state.
        
               | pengaru wrote:
               | > I feel like we've reached the all too common HN cliche
               | of having to address that words have multiple meanings
               | and those meanings change over time
               | 
               | It's not even an HN-specific thing.
               | 
               | Arguing semantics is a time-honored method for derailing
               | discussions. A sort of quasi-bikeshedding, except more
               | often done in bad faith for creating a winnable meta-
               | argument by someone who otherwise has nothing useful to
               | say about (or was losing a debate of) the actual
               | underlying subject.
        
       | Hamuko wrote:
       | Frore is claiming that you could slap two AirJet Minis in a Steam
       | Deck and that it'd cut down on the noise, lower the device skin
       | temperature and free up enough space for a bigger battery that'd
       | allow for 30% more game time.
       | 
       | https://www.froresystems.com/application/handheld-gaming
       | 
       | I don't think they've actually built a prototype (the PDF and the
       | website show off two separate designs) but it seems promising if
       | these stop being as expensive as they are currently.
        
       | 6DM wrote:
       | I think the use case for this would be ultra thin laptops, this
       | could also be a quiet alternative to the current fans loudly
       | spinning up.
        
         | seltzered_ wrote:
         | I was actually thinking high-end tablet PCs (like the Asus ROG
         | Flow Z13, Microsoft surface) where they have thermal
         | constraints due to having both a medium-TDP CPU (and GPU) AND a
         | large display to backlight.
         | 
         | Tablet setups also can facilitate better keyboards and an eye-
         | level screen, as shown by members in
         | https://reddit.com/r/ergomobilecomputers
        
         | wil421 wrote:
         | It's the reason I love my M1 air so much. No fan and barely any
         | heat.
         | 
         | My company issued HP zbook for Software Engineers is very loud
         | and pushes lots of heat out. I would love a quieter alternative
         | for large Windows laptops.
        
           | CharlesW wrote:
           | FWIW, Windows on Parallels is great on my M1 Air.
        
           | rimunroe wrote:
           | I'm looking forward to using one of those. My mom's old Dell
           | laptop died and she finally made the switch to the new
           | MacBook Air. It seems fantastic.
           | 
           | I had the 2015 MacBook. It was an oddball model which afaik
           | never got a refresh: it didn't have any active cooling, had a
           | 12" screen, a single USB-C port and a headphone jack, and was
           | the and the first version of the terrible butterfly keyboard
           | (which had to get replaced three times over its life!). I got
           | it through my company for half price after an executive had
           | gotten it for travel but couldn't get adapters to reliably
           | work for presentations. The main things I remember about it
           | (other than the terrible keyboard) were that it was
           | delightfully portable and quiet, and that I had to consider
           | the weather when I was deciding what to do. I haven't lived
           | in a place with air conditioning in years, and in the summers
           | I'd sometimes have to stop my chronic multitasking and decide
           | if I wanted to browse/run other programs, or stream videos.
           | On the rare occasions I'd do some programming and needed to
           | compile something, I'd open the laptop halfway and put it on
           | its side with a fan pointed at the back. I also put it on an
           | ice pack on more than one occasion. I've always been
           | surprised that condensation didn't kill it.
           | 
           | The new Airs seems like they solidly solve all of my
           | complaints with the 2015 MacBook, which means I might finally
           | have found a computer which can make me as happy as my
           | Fujitsu LifeBook P1510D did. My only quibble might be that
           | it's still a smidge larger than what I'd like, but it's a
           | minor quibble.
        
           | kitsunesoba wrote:
           | Yeah, reducing heat output and fan noise should be more of a
           | goal for mainstream PC vendors. My ThinkPad X1 Nano is a
           | lovely machine... super lightweight, tiny, solidly built,
           | decent keyboard and screen, but it starts getting hot and
           | spins up its fan with the slightest hint of work. I have to
           | keep its CPU throttled by putting it in power saver mode to
           | keep its fan off (most of the time), which brings major
           | performance tradeoffs.
        
         | JamesAdir wrote:
         | Why do we need thinner laptops? You can hardly work on any of
         | the current models today, which their almost zero height keys.
         | We need better keyboards and not thinner laptops.
        
           | znpy wrote:
           | i have mixed feelings about thin&light laptops... i thought i
           | wouldn't mind a bit more weight, but after getting a 14"
           | macbook pro for work i appreciated my personal x270 (and my
           | old T440) so much more.
           | 
           | the macbook pro is crazy heavy.
        
           | bryanlarsen wrote:
           | Weight is a much more useful metric than height for a laptop.
        
           | packetlost wrote:
           | Hard disagree. Shorter key travel is a matter of preference.
           | I quite like the ultra-short-travel of my laptops even though
           | I have half a dozen mechanical keyboards that I use on my
           | desktops. It would be nice to have more options, but I'd take
           | a thin and light over a chunky laptop any day.
        
           | topikk wrote:
           | > You can hardly work on any of the current models today
           | 
           | Tell that to the millions of people who work on them daily
           | without issue.
        
           | dietr1ch wrote:
           | I'd like if we got to the point where the computer was on the
           | back of the monitor so it'd be easier to cool, and then the
           | bottom would be dedicated to peripherals, including way
           | better keyboards.
        
             | MiddleEndian wrote:
             | The Surface Book 2 sort of did this, most of the computer
             | was behind the screen, the base only had a keyboard, more
             | batteries, and a GPU (which could be disabled so you could
             | detach the screen).
             | 
             | It got really hot really fast if you did any graphics/video
             | work on it, even more than regular laptops do.
        
             | Zanni wrote:
             | We're already there with tablets (Surface, iPad) with
             | attached keyboards. Personally, I prefer the bulk and
             | weight of the device to be on the table, not propped up,
             | but some folks love this configuration.
        
           | stronglikedan wrote:
           | If you can make the frame thinner under the keyboard area by
           | moving some computing components around to areas that are
           | normally housing cooling components, then you can have a
           | longer travel distance for the keys. So, not necessarily a
           | thinner laptop, but thinner where it would count.
        
             | _a_a_a_ wrote:
             | "Talk about thin laptops long enough and people will start
             | to buy into it" (A variation on Goebbels' big lie)
             | 
             | I reckon you bought into it.
        
               | JadeNB wrote:
               | > "Talk about thin laptops long enough and people will
               | start to buy into it" (A variation on Goebbels' big lie)
               | 
               | I think that analogizing persistent marketing propaganda
               | to genocidal fascist propaganda is going a bit far.
        
         | gizmo wrote:
         | No, the primary use is in dusty industrial environments where
         | regular fans get clogged up.
         | 
         | Noise currently is worse than regular fans because it's higher
         | pitched.
        
         | Enginerrrd wrote:
         | As someone that can hear a lot of "ultrasonic" frequencies, I
         | would not assume it would be quiet. Those ultrasonic occupancy
         | sensors are TERRIBLE. Your "quiet" laptop might be worse since
         | the sound would be continuous instead of loud bursts.
        
           | rimunroe wrote:
           | I really wonder about what the impact of having a lot of
           | these devices in a house with animals which can hear higher
           | frequency ranges would be. Most (all?) rodents, cats, and--to
           | a lesser extent--dogs all hear well into the ultrasonic
           | range. I don't know if this is the sort of ambient noise that
           | they'd habituate to easily or if it would cause constant
           | stress.
        
           | noizz wrote:
           | You are correct. LTT's coverage mentions that while it's not
           | loud, it's a bit unpleasant due to high frequencies. Although
           | things might get tweaked with case design probably.
           | Regardless, I am super excited about this tech.
        
             | moffkalast wrote:
             | After decades of jokes about laptops "lifting off" when
             | under load, they will finally sound like a proper jet
             | airplane.
        
         | bitwize wrote:
         | I still prefer the Apple approach: design a CPU so powerful and
         | efficient, it can handle desktop workloads and larger and not
         | need active cooling at all.
        
           | jayd16 wrote:
           | The higher performing skus are still actively cooled.
        
           | gamblor956 wrote:
           | But Apple's Mx chips can't handle desktop workloads without
           | active cooling, and are demonstrably worse at workloads that
           | are demanding enough to require active cooling.
        
           | Tade0 wrote:
           | I think it's less Apple and more TSMC.
           | 
           | Through overzealous cleaning I broke the fans on my Ryzen
           | 4800HS laptop but with the right fan-control software even
           | that TSMC 7nm CPU operated largely fanless(they only spooled
           | up above 80degC, which was rare) until replacements came in
           | the mail.
           | 
           | The M1 in the Air is throttled to roughly half the TDP of
           | other devices with the same chip, so it's not like you're
           | getting all the performance it can offer.
           | 
           | That being said Apple's design is somewhat faster than
           | competition using the same lithography.
        
             | kitsunesoba wrote:
             | > The M1 in the Air is throttled to roughly half the TDP of
             | other devices with the same chip, so it's not like you're
             | getting all the performance it can offer.
             | 
             | Throttling only kicks in on the M1 Air when the CPU has
             | been running full tilt for ~10m. Outside of that it
             | performs identically the the regular M1 in a Mac Mini,
             | iMac, or 13" MBP.
             | 
             | Even a lot of heavy users rarely keep the CPU pegged for
             | that long. Most workloads are instead short and bursty.
             | There's stuff like compiling Chromium that'll do it, but
             | people who do that regularly should probably be looking at
             | something like a 14" M1/M2 Pro MBP instead if portability
             | is a priority.
        
           | dimmke wrote:
           | I was never into "PC Building", but had a friend who was and
           | a few years ago they helped me build out a "Small form
           | factor" PC, using this case: https://imgur.com/a/7xLBiC3
           | 
           | The case was designed to have two columns, and one side was
           | dedicated to fitting a full sized graphics card. The other
           | column was the rest of the computer.
           | 
           | The graphics card had its own fan system, but I also needed a
           | fan glued to the CPU and then an extra one at the bottom to
           | help spit out exhaust. The thing ran so hot, and the cords
           | ended up just jammed into the case where they fit.
           | 
           | I had no experience with building PCs and was amazed I could
           | fit a 2TB Hard drive in a tiny little stick on the
           | motherboard, whereas X amount of years before, I would have
           | had to install a spinning disk, much bigger. Such a huge leap
           | forward.
           | 
           | The main reason I wanted this form factor was because the
           | perceived elegance of it appealed to me. But it became clear
           | to me through the process that "small form factor" for
           | enthusiasts is usually more about the challenge of building
           | in a tight space. The thing is super loud with the fans and
           | runs super hot when using it for gaming, even with a 1080ti
           | card.
           | 
           | That's one of the reasons Apple's progress with CPUs is so
           | great to me. I don't want the entire industry to move to SoC
           | style where everything is soldered to the motherboard, but
           | the more efficient you can make the rest of the computer, the
           | better cooling system you can design. Just like the M.2 form
           | factor for SSDs was a huge leap. But even if you can't
           | eliminate the need for active cooling, something like this
           | would make things a lot better. I think you'd still have to
           | design around where the exhaust blows out though.
           | 
           | I think it'd be better to have the ARM increase in efficiency
           | and then maybe combine it with something like this for extra
           | powerful models, the way graphics cards are sold with the
           | fans attached.
           | 
           | Someday I'd like to be able to upgrade the build and optimize
           | it. The dream would be to not have any fans, but some other
           | form of active cooling, and for the machine to be basically
           | silent, while packing a 3090+ tier graphics card.
        
         | idontwantthis wrote:
         | Once this is affordable and proven reliable, why would laptops
         | that are not thin and light still exist?
        
           | bick_nyers wrote:
           | Structural integrity?
        
             | tempodox wrote:
             | But only until we can crumple up our electronics like a
             | piece of cloth.
        
           | gambiting wrote:
           | I have a work laptop with a 12850HX + 3080Ti and 128GB of ram
           | for work - even if you remove all the cooling from it, all
           | the components still take a huge amount of space inside. The
           | 128GB CAMM module for ram alone is larger than the entire
           | motherboard in macbooks. Also the cooling would need to be
           | able to deal with ~200W of heat, even solid state cooling
           | would be quite large for such an amount of heat.
           | 
           | https://imgur.com/a/L24ATjp
           | 
           | that green module on the right is just ram. And Dell is
           | trying to make this the future standard for ram modules in
           | laptops. (ignore the dust on fans, I was in the middle of
           | cleaning it).
        
             | kylixz wrote:
             | What model Dell? Satisfied? We've tried some other Linux
             | friendly brands and the quality control has been less than
             | stellar.
        
               | gambiting wrote:
               | Yeah it's a Dell Precision 7670 - not had any problems
               | with it, and it's an absolute beast performance wise.
        
         | corysama wrote:
         | Interesting point from the Linus Tech Tips video: A lot of the
         | thermal throttling in laptops is based on user comfort (don't
         | burn the user) rather than CPU limits. He had a demo of a
         | laptop CPU running faster and _hotter_ with this system because
         | the body of the laptop was the focus of the cooling.
        
       ___________________________________________________________________
       (page generated 2023-06-02 23:01 UTC)