[HN Gopher] AirJet makes a MacBook Air perform like a MacBook Pro
___________________________________________________________________
AirJet makes a MacBook Air perform like a MacBook Pro
Author : MBCook
Score : 154 points
Date : 2023-11-27 15:35 UTC (7 hours ago)
(HTM) web link (www.macworld.com)
(TXT) w3m dump (www.macworld.com)
| rgrmrts wrote:
| I've been following Frore Systems (https://www.froresystems.com/)
| for a while and have wondered if Apple might consider
| acquisitions of next-gen components like theirs. Presumably Apple
| wants to wait for them to be proven in production first, but can
| you imagine an even more power efficient and silent cooling
| system on MacBook Pros? That would get you current M3 laptop
| performance in a form factor closer to the last Intel generation
| MacBook Pro (slimmer).
| Workaccount2 wrote:
| I would feel physical pain if apple locked in tech like this.
| So dumb and annoying that we have to cripple so many potential
| applications just so apple can exclusively put it in a couple
| high-end laptops.
| sgjohnson wrote:
| > So dumb and annoying that we have to cripple so many
| potential applications just so
|
| My brother in christ, have you heard of the patent system in
| general?
|
| Try starting a new phone or laptop company. Your step 1 will
| be to acquire a patent war chest, so if someone sues you for
| patent infringement (and they will), you can sue them right
| back for the same.
|
| In this world, if that company is for sale, someone much
| bigger is going to buy it, and make their technology their
| own exclusive. Be it Apple or Samsung or whatever. Getting
| hands on this tech _so your competitors don't_ would also be
| a valid reason for acqusition.
| smoldesu wrote:
| Shh, any louder and you'll rouse the Butterfly keyboard or
| the Touch Bar from their graves!
| wnoise wrote:
| What about the FingerWorks TouchStream? _sigh_
| orik wrote:
| I think if the tech does not get acquired by someone big you
| probably will never see it come to market in any meaningful
| way. AirJet should just let people buy these for DIY to start
| and build hype and start partnering with OEM's for
| preconfigured options.
| dangus wrote:
| DIY is probably the worst option from the business
| perspective or AirJet.
|
| The market for that is minuscule, and they likely can't
| make a solution that works universally for every laptop.
| londons_explore wrote:
| If the product was any good, Apple would have already signed an
| exclusivity deal and NDA. If they didn't, another manufacturer
| would have.
|
| The fact that hasn't happened, either means the product has a
| big downside (eg. stops working after 1 hour use as it blocks
| with dust), or because the manufacturer is offering silly
| terms.
| MBCook wrote:
| It's brand new, it was shown at CES. I think they've worked
| with one PC OEM but their devices are designed for much lower
| heat output at the moment than what a MBP generates.
|
| I suspect it's more of a "does it work, how does it hold up,
| and can they scale to our volumes/heat levels" question.
|
| The fact they could hack this is still interesting. Fun to
| watch.
| pier25 wrote:
| I still haven't heard the fans of my 14'' M2 MBP. And it's the
| thinnest laptop I've ever owned.
| rgrmrts wrote:
| I currently have the 2019 16" MacBook Pro and it's thinner
| than the new chassis. I like the form factor better, though
| I'll likely upgrade soon happily for better performance and
| power efficiency.
| Someone wrote:
| They're moving air through a channel that's about a mm high,
| yet claim "dustproof" on that page. Is that realistic?
|
| Also, could a hacker build something like this (but likely much
| less efficient) at home by taking apart a DLP video projector
| and playing a carefully constructed video loop on its
| micromirror
| (https://en.wikipedia.org/wiki/Digital_micromirror_device)?
| nomel wrote:
| According to LTT video, the 10x static pressure allows for
| dust filters.
| acomar wrote:
| such claims without tests to validate are kind of useless.
| they're just taking the company at their word.
| MBCook wrote:
| They're already in a commercial product:
|
| https://www.zotac.com/us/page/zbox-pico-PI430AJ-airjet
| orik wrote:
| where can I buy this?
| xenonite wrote:
| Being spared a fan with its noise is the exact reason that I use
| an MBA. I surely do not want to retrofit a fan. But nice idea
| anyways to look after your cooling approach.
| rswskg wrote:
| it doesn't use a fan
| unsupp0rted wrote:
| Video says it expels air, generating sound at about 25 db, so
| not much "air noise".
| rowanG077 wrote:
| Frore systems claims 21 dBA. That's about the level in a quiet
| area in nature without wind. Much less then even a whisper.
| Unless you are sitting in an extremely quiet room you won't
| hear it.
| torginus wrote:
| Generally it won't spare you of the noise, as the exit velocity
| of the stream will be quite a bit higher, which will result in
| a higher pitch, which is quite a bit more noticeable due to the
| quirks of human hearing.
| simbolit wrote:
| FYI: This is a proof of concept, not a product, not a product
| announcement.
| LunarAurora wrote:
| If I understand correctly, CPU clock speed is really relevant
| only for bursts in CPU use. Any meaningful steadying of
| processing speed at that peak,or near it, seems to be a question
| of power efficiency and cooling in tandem.
|
| So now if Apple is well ahead of the competition in power
| efficiency, and if, on top of that, the Aijet is really a
| revolution in (silent?) cooling, then we have here a maximization
| of the value of all those Ghz.
|
| As an amateur computer musician, this is exactly what I need: a
| continuous silent performance at a "budget" price.
| toast0 wrote:
| There's a couple wrinkles here. For one, power efficiency
| varies with temperature, keeping chips cool, increases their
| efficiency, although you'd need a lot of data to determine if
| watts spent on cooling were comparable to watts saved by being
| cool at a given performance level.
|
| The other is Apple chips target a lower clock rate than Intel
| and AMD do for laptop chips, this is a tradeoff. You get better
| perf/watt at lower clock rates, and a lower targer design can
| be smaller[1] and lower power. OTOH, you miss out on the top
| end of performance for those willing to power and cool the
| beast. It's a good tradeoff for Apple, who never provided ample
| cooling for their Intel based laptops, but it's hard to market
| for Intel and AMD when PC laptop makers do want to sell some
| devices that can clock to the moon all day if users want it.
| AMD is expirementing with their compact cores as power
| efficient cores in some laptop skus, so we'll see something
| there; Intel is going with the BIG.little approach of less
| capable little cores, rather than just clock limited little
| cores.
|
| [1] see AMD's Zen4c which is about half the size as a regular
| Zen4 for the core+L1+L2, L3 cache footprint remains the same
| for a given size
| jeroenhd wrote:
| On many laptops you can disable boosting speeds and lock down
| the frequency (thermals permitting). I don't know if Apple's
| devices allow for such measures, but close-to-constant
| performance without fan noise can be a matter of configuration.
|
| You can try running these chips at their boost clock all the
| time by attaching coolers, but the hardware wasn't designed for
| that and neither was the software. It's easier (and more power
| efficient) to clock down if you care about maintaining a set
| speed. With a battery capacity designed for short bursts,
| you'll also lose out on battery life if you decide to crank the
| hardware.
|
| The Macbook Air and its competition (MS Surface etc.) are
| completely silent because they are fanless, and I don't expect
| their performance to be inadequate for anything but the most
| CPU intensive tasks. The Airjet will add some airflow, so while
| the difference may be minjmal, it'll always end up louder than
| before.
| huijzer wrote:
| For people interested in the Frore system, Linus Tech Tips did a
| video on it: https://youtu.be/vdD0yMS40a0?si=UwAeqbTxxnDRf1_V.
| miffe wrote:
| They just released a new video about it.
|
| https://www.youtube.com/watch?v=ji5kZ3VdFDY
| Kirby64 wrote:
| As cool as this tech is, I question the efficiency. They claim
| '10.5W of heat removal using 1.75W of power' in their press
| release. I'm not entirely sure how this compares to regular fans,
| but data points I can find suggests that regular fans use much
| less power (even the tiny laptop ones) for similar amounts of
| cooling performance. Could possibly be very cool for a tethered
| application (mac mini, mac studio?), but increasing power draw
| seems like it'd be a bad thing for battery life.
| brundolf wrote:
| It seems like the kind of thing a company like Apple could
| acquire, tightly integrate, and squeeze out a small margin
| above existing tech on a couple axes (eg. maybe it draws more
| power but makes room for a bigger battery, or maybe software
| can hyper-optimize when and how it's turned on)
|
| Seems like a harder sell as an off-the-shelf component. I
| wonder if they're angling for an acquisition (by apple or
| otherwise)
| oddevan wrote:
| This smells _exactly_ like a move to drum up acquisition
| talks.
| RajT88 wrote:
| Exactly my read as well:
|
| > AirJet isn't a consumer product-consumers can't buy it
| and modify their own MacBook Air, nor does Frore plan to
| create a kit for consumer sale.
|
| This is aimed at Apple, and he doesn't want to offer it for
| sale? Acquisition.
| alt227 wrote:
| Apple have already shown from their crippled Intel Macbook
| Pros that they care nothing for good cooling systems, and are
| more than happy to accept sub par performance as long as the
| design of the product is right. I cant see them jumping at
| the chance to aquire a cooling device company.
| criddell wrote:
| I doubt Apple is all that interested in something like this for
| the Air because they have a faster, actively-cooled laptop
| available - the MacBook Pro. The Air is their light-duty,
| fanless machine and I doubt they are all that interested in
| changing either of those qualities.
|
| It would be nice for consoles and other living room
| electronics. My PS5 fans are annoyingly loud.
| freedomben wrote:
| Why is this downvoted? It's speculation and may be wrong (I
| don't think it is), but it's a perfectly legitimate thought
| that _any_ other company on the planet would receive from HN
| without the downvotes.
|
| The downvoting anything critical of Apple is getting
| ludicrous. Maybe you believe they are a benevolent user-
| loving company who always does what's best for their users
| regardless of how it affects their business, but viewpoints
| who _don 't_ share that position aren't automatically wrong
| because they have a different opinion than you. Skeptical
| people are what built the hacker community. If you don't want
| to embrace that spirit, at least don't try to silence it.
|
| Edit: And yes, whether Apple would want this or not is very
| relevant, because if you read the article, they do _not_ sell
| to consumers. The only customer for this would be Apple, and
| having Apple buy their product and put it in macbooks is the
| strategy they are pursuing.
| jameshart wrote:
| 1) it's not downvoted
|
| 2) it's not critical of Apple
| freedomben wrote:
| > _1) it 's not downvoted_
|
| To quote Doc Brown, "you're not thinking fouth-
| dimensionally!" It was when I posted the original
| comment.
| Kirby64 wrote:
| You can make large fans quiet, people do it all the time...
| it's just a design choice to not do it (since usually it
| involves size tradeoff). I assume this would be quite
| expensive to cool a PS5 though, since PS5s need hundreds of
| watts of cooling (so, you'd need like 20 of these modules?).
| ComputerGuru wrote:
| > You can make large fans quiet, people do it all the time
|
| This is unintentionally funny, because the law of fans is
| that the _smaller_ they are, the louder they have to be to
| achieve the same cooling.
|
| Air moved is a function of blade size and speed (and shape
| and count and more, all applying to big and small fans
| ~alike) -- and those two are indirectly proportionate. The
| smaller the blade, the faster you have to spin it. The
| faster you spin, the more turbulence you make, the more
| sound you get.
|
| Biggest hack for silent _active_ PC cooling is to use
| upsized fans spinning slower (see Notcua).
| Kirby64 wrote:
| Understood, however the PS5 uses a single giant 120mm fan
| (120x45mm) which should in theory be quiet. It's much
| larger than the fans used in many consoles. And yet...
| I'm not sure what other constraints caused this trade
| off.
| ComputerGuru wrote:
| One fan might not be enough for the amount of heat
| generated. The airflow engineering could be very poor,
| requiring high velocity and not just high cfm to
| overcome. It could be positioned poorly or attached to a
| skimpy heatsink to save on copper (or even worse, to save
| on another metal). The hottest components may have very
| small surface area so they don't diffuse heat
| efficiently. Lots of reasons, usually having to do with
| thermal engineering not being a priority until late in
| the game.
| jdietrich wrote:
| The PS5 uses a centrifugal blower rather than an axial
| fan. Blowers can push air effectively through restrictive
| heatsinks and vents, but at the cost of greater
| turbulence and therefore noise. The Xbox Series X/S
| consoles use axial fans that move a lot of air at low
| RPMs; this requires large intake and exhaust vents and an
| optimised internal layout, because the flow rate of an
| axial fan falls rapidly as static pressure increases.
|
| The PS5 is fairly loud, but the noise is quite
| directional - it's much louder from the rear than the
| front. By keeping all of the vents on the front and rear
| of the console, it can work effectively inside a TV
| cabinet or media center that would suffocate an Xbox.
| It's not the set of tradeoffs I would have preferred, but
| it isn't a thoughtless design.
| alt227 wrote:
| Why speed up the cheap model when we have a faster model for
| more money?!
| losteric wrote:
| Silence is part of the Air's appeal, as well as the shape
| which wouldn't allow for an internal fan
| LordDragonfang wrote:
| IIRC, LTT did a video touching on these a bit ago and said
| there was still a noticeable hum (at an annoying frequency),
| so the noise savings might be less than you'd expect.
| tiltowait wrote:
| Is your PS5 dust-clogged? Mine is effectively silent.
| criddell wrote:
| I don't think so. I just got it back from Sony and I would
| guess that when they replace the power supply, they would
| also blow out any dust.
| geoelectric wrote:
| Unless it's calmed down since launch, PS5 noise is all over
| the place. I don't usually hear fan so much as coil whine
| on mine though.
| throwawaymobule wrote:
| at the very least, getting at the fan for cleaning is
| toolless. they really learned their lesson from how bad
| some ps4s got.
| WillPostForFood wrote:
| You are trading some efficiency for smaller size.
| londons_explore wrote:
| That ratio is pretty okay for a laptop... Sure, it's far worse
| than fans, but you are only paying for all of that 1.75 watts
| when at full CPU load for extended periods, which is rare.
| inasio wrote:
| Note that they also had to drill vents into the laptop and use
| the speakers as outlets.
| duped wrote:
| Is it audible to animals?
| megraf wrote:
| Likely edible as well
| dghughes wrote:
| Humans are animals.
| MBCook wrote:
| In the video interview on the page they said that it's
| something like 25 dB. So technically it's audible to humans
| too, it's just low enough that no one really cares.
| duped wrote:
| In the product page they mention it's 25dBa, but if that's
| measured using a reference microphone for the human spectrum
| it's not going to reflect if the ultrasonic frequencies its
| running at are audible to say, a dog.
|
| The reason I asked is because this is sick, unless it makes
| my dog cry.
| baz00 wrote:
| My "original out of box" OCD couldn't stare at a machine modded
| like that.
| unsupp0rted wrote:
| My 2015 13" MacBook Pro spins up fans hard just to check email.
| Maybe it's the aftermarket OWC Aura Pro SD I've got in it? As
| soon as I run VS Code with a Firebase emulator it starts
| throttling down.
|
| I'd love to get this Airjet thing installed in it to see if it
| makes a difference.
| pornel wrote:
| Get an Apple Silicon Mac. The latest Sonoma beta has a bug that
| makes every WebView use 99% CPU all the time, and the fans
| don't even spin :)
| jeroenhd wrote:
| The Intel Macbooks had woefully underdesigned cooling
| solutions. They're also less efficient than the model
| demonstrated here.
|
| This "airjet" is a kind of almost solid state fan using
| membranes to generate airflow. They may be able to perform
| similarly to normal fans with less noise and a smaller physical
| footprint, but they still need to get that air moving through a
| heatsink somehow. I don't think it'll work for laptops that are
| designed to thermal throttle after a few seconds.
| nicoburns wrote:
| A 2015 MacBook is also approximately 10x less powerful at max
| capacity than a 2023 laptop, so it has to work much harder to
| achieve the same result.
| jwells89 wrote:
| With that laptop now being 8 coming up on 9 years old, I
| wouldn't be surprised if its thermal paste has dried out, which
| means that heat isn't being transferred away from the CPU into
| the heatsink as effectively, resulting in the fan running
| harder. If you want to keep it going a while longer it might be
| worth picking up a tube of Arctic MX-4 and replacing the stock
| paste, cleaning out any accumulated dust in the process.
| pier25 wrote:
| The 2015 models used 4th gen Intel chips which ran quite hot.
| And the cooling system on those models is super noisy.
|
| Just get an M1 Air on sale, second hand, or refurbished. It
| will be night and day.
| unsupp0rted wrote:
| I have a 16" M1 Pro as well and it's night and day. But I
| still love my little 13" 2015 Pro- there's something
| comforting about it.
| dghughes wrote:
| My 2012 17" MacBook Pro video card died about 2017 but even
| then and even when new the fan was on a lot. HD was SSD, lots
| of RAM, discrete GPU but overall the thing was toasty.
| ralfd wrote:
| When was the last time you opened it? In my experience it is
| just dust/hair in the fans.
| someNameIG wrote:
| I have a 2019 16" i7 pro and a M1 Air. The 16" is always warm
| even just idling, and the fans go when you do anything more
| intense that basic terminal tasks.
|
| Air gets warm only with some long compiling.
| gcr wrote:
| When they write about putting this chip in the MacBook, do they
| mean physically cracking the device open and putting the chip
| inside? If so, where does the space come from? doesn't that void
| the warranty?
| runjake wrote:
| While the article itself answers all your questions, the
| article links to more details at
| https://www.froresystems.com/blog/how-do-you-cool-ultra-thin...
| andrei_says_ wrote:
| Would placing the air on top of an ice pack temporarily improve
| performance? For example when needing to render a video?
|
| Don't judge if this is a very stupid question - about to buy a
| MacBook Air to substitute an M1 Max for travel.
| mrweiner wrote:
| Iirc, an issue with this and e.g putting it in the fridge is
| potential for condensation. Probably worth looking into before
| putting your machine on ice.
| Turj48tj wrote:
| Not really, thermal conductivity between CPU and outer layer is
| not best. M1 is OKish, M2 are really bad.
| MBCook wrote:
| I think I saw that they added little standoffs on the M3
| between the case and the cooling mechanism to help prevent
| the case from coming into contact and getting too hot.
| droopyEyelids wrote:
| Sort of. The problem is that the chip is what needs to be
| cooled, and that is separated from the ice pack by a thermal
| pad and the laptop case.
|
| both of those are way less efficient than something that cools
| the chip directly
| smodo wrote:
| Yes. The aluminum chassis is basically a heat sink. The
| underside of the M1 Air heats up when the soc is under load.
| Cooling that spot will help cool the soc.
| rhinoceraptor wrote:
| If the chassis were a heat sink, it would be a serious burn
| risk.
| richardwhiuk wrote:
| That's basically why it throttles.
| MBCook wrote:
| It's not directly connected to the CPU, but it clearly gets
| quite hot.
|
| Removing that heat should be beneficial either way.
| crazygringo wrote:
| Can anyone explain how "tiny membranes that vibrate at an
| ultrasonic frequency" is able to produce "AirJet's 1750 pascals
| of backpressure" which is "more than 10 times the suction force
| of a high-end notebook fan"?
|
| I understand how you could have a bunch of membranes that vibrate
| ultrasonically -- it's basically an array of miniaturized audio
| speakers. What I don't understand is how any localized
| oscillation of air gets transformed into massive, consistent air
| suction in a single direction.
| evntdrvn wrote:
| The answer, as you might expect, is "valves" as part of the
| integrated MEMS :)
| https://patents.google.com/?assignee=frore+systems
| pavedwalden wrote:
| "valves" is a good short answer. If you want a long
| explanation, this video gets into a lot of details.
| https://www.youtube.com/watch?v=vdD0yMS40a0
| crazygringo wrote:
| Thanks, but the video doesn't actually explain it. It
| explains a lot _around_ it, but it just says that the chip
| uses a "piezoelectric fan" and shows the intake slits and
| the outtake jets, but I'm still not finding an explanation
| of the mechanism itself.
|
| I can vaguely imagine how valves _might_ work, but I 'm
| still curious how they _do_. I looked at the linked patent
| applications but the descriptions seem way too complex and
| abstract to get any kind of intuitive sense of the
| mechanism without already being an expert.
|
| There's a related video they link to: https://youtu.be/NY-
| gA_zA_os?si=9ULOds_TqUO56MGs which demonstrates one type of
| piezoelectric fan, but it seems to be operating on totally
| different principles -- one end is freely swinging, rather
| than a membrane with fixed edges oscillating. And there are
| definitely no valves involved in that one.
|
| Is there anywhere else that explains how oscillating
| membranes generate smooth airflow?
| torginus wrote:
| my guess would be that they move an absolutely tiny amount of
| air, and pressure is force * area. They explained that since
| air is a very poor conductor of heat, most of the heat tends to
| pile up in a thin boundary layer next to the fins, and just
| exhausting that will get you most of the cooling performance.
| mannyv wrote:
| There was a hack (using the most liberal idea of a hack) a few
| years ago that was even simpler: just put a thermal pad into your
| MBA. Apparently it still works.
|
| https://forums.macrumors.com/threads/m1-air-with-thermal-pad...
| Reason077 wrote:
| Yep, can confirm this works great in the M1 Air. I've been
| running mine this way since I got it way back in 2020.
| _Greatly_ lengthens the time you can play graphics-intensive
| games before thermal throttling starts to occur and your frame
| rate goes choppy.
| evanjrowley wrote:
| I wonder what the performance improvement would be for a
| newer PTM7950 thermal pad. These have a different chemistry
| versus traditional thermal pads. They actually get _better_
| over time.
| rowanG077 wrote:
| That would be way too thin. PTM7950 is not intended to be
| used as a thermal pad.
| germinalphrase wrote:
| How do you find surface temps under non-gaming loads, i.e.
| does this limit your use of the MacBook to desk only?
| TazeTSchnitzel wrote:
| To pre-empt the inevitable question about why Apple didn't do
| something so seemingly obvious: it's a trade-off to avoid the
| device getting uncomfortably, maybe skin-burningly hot under
| load.
| summarity wrote:
| Indeed, there are also surface temp regulations for consumer
| devices (eg why certain action cams shut down much earlier in
| the EU)
| rhinoceraptor wrote:
| Even a very reasonable CPU temperature under load like 60C,
| you can only touch a surface that hot for about 5 seconds
| before being burned.
| someNameIG wrote:
| Reading that thread it also seems to heat the battery more,
| prematurely degrading it.
| rhinoceraptor wrote:
| This hack will always work, they must thermally insulate the
| laptop case, otherwise the laptop is a huge burn hazard. About
| 43 C is the threshold for pain, anything above that will cause
| burns.
| Spunkie wrote:
| I've been pretty excited for the airjet to start making it into
| products but also I hope they open up D2C orders for tinkerers
| soon. As a normal consumer I can buy every other part of my
| laptop or desktop all the way down to most individual electrical
| components.
|
| I get that it's a new product that requires good engineered
| integration to really show off how amazing it is, so they want to
| start with some flagship products... But if Frore Systems starts
| treating the airjet like it's a secret sauce to be slowly doled
| out to manufactures it'll just make me suspicious of their claims
| and the airjets real world usability.
| killjoywashere wrote:
| piezoelectric cooling has left a history of ruins because it
| doesn't account for condensation, particularly in small,
| consumer-grade applications where drainage paths are hard to
| design for. You need one drop of water in the wrong spot. Not a
| game I'd play with a metal-cased laptop, personally.
| chasing wrote:
| I used to set my MacBook Pro on ice packs when working on certain
| processor-intensive tasks... Seemed to help!
| aeyes wrote:
| Doesn't this cause condensation?
| acheong08 wrote:
| Yes. I broke my Mac a few years back this way
| bastawhiz wrote:
| Their business model to try to get Apple to license their tech
| seems kind of misguided. Why wouldn't Apple simply design their
| own system? It's one thing to get licensed or acquired by a
| company worth $100M. It's another to try to do the same with a
| company worth 10,000x times that much.
| alt227 wrote:
| If they dont have a patent on this thing then I would be very
| surprised that they are shouting this loudly about it.
| bastawhiz wrote:
| Even if they do have a patent, it seems like Apple has enough
| lawyers that they could simply build around it. It's a
| trillion dollar company. They can do what they want, really.
| mustaflex wrote:
| I don't know how it will pan out in commercial products, but it
| has been a long time that I've been this excited about hardware,
| it does feel like real innovation for cooling and especially for
| mobile devices.
___________________________________________________________________
(page generated 2023-11-27 23:02 UTC)