I think is useful to describe the limits of nanomachine for better game (as GM i favor balance ald love player find metods for overcome limits).
- thermal: a nano is small and become hot very quick, in fluid (gas liquid) can dissipate dumping in the fluid, in vacum only radiating (principally IR but follow the rule for blackbody radiation)
- chemical reactivity: smaller is a body more surface area it have compared with volume, if nano in exposed to chemical that can react with some part of it this appen instantly (a swarm composed of carbon based nano burn as a FAE bomb in oxigen atmosphere... even diamonds burn ;) )
- energy storage: small volume means poor storage space and long trips, for a nano 1cm is some 10^7 it's dimension... even good efficency mean limited energy contenent
- sensory impairs: a nano can't directly percive wavelight longer than himself, tuely nanometric machine can percive IR or below (most molecular groups are in IR energy range) but cell sized can detect very short microwave. Shorter wavelight better the resolution and gain. Sound is detected as variation in ambient molecular density so difficult to percieve unles the nano is cell sized and detecting ultrasounds
- onboard processing power: nanocomputing is difficult for thermal issures, nano must use very simple programming and individually must be dumb but may show very hight "colony smartnes". Bigger nano are smarter than small ones
There are a lot of metod for alleviate this limitation but most are based on swarm strategy and cooperative bevior, swarms composed of specialized subraces of nanomachine working together
Welcome! These forums will be deactivated by the end of this year. The conversation continues in a new morph over on Discord! Please join us there for a more active conversation and the occasional opportunity to ask developers questions directly! Go to the PS+ Discord Server.
nanomachine limits (realistic)
Sat, 2012-08-04 19:47
#1
athanasius
nanomachine limits (realistic)
Mon, 2012-08-06 06:03
#2
Arenamontanus
I really recommend checking
I really recommend checking out Chapter 2 of Nanosystems where Drexler gives scaling laws for machinery: http://e-drexler.com/d/06/00/Nanosystems/ch2/chapter2_1.html
This kind of quick estimation already gives a lot of information, although one needs to check for corrections.
It is worth remembering that in a vacuum very small objects do not radiate freely as blackbodies, since they don't couple with wavelengths much longer than themselves. A submicrometer nanomachine will not be an efficient IR radiator, so it will actually cool down slower.
In a normal matter environment the thermal relaxation time is very fast, around 10^-13 seconds - nanomachine heat becomes environmental heat nearly instantly.
The real issue is how much energy they need to dissipate. At one extreme nanomachines are just inert objects, not using any energy at all until the right stimuli activates them. At the other extreme they are running very high reaction rates.
The upper limit is likely diffusion limited: like superefficient enzymes they will trigger a reaction whenever they meet the substrate they are supposed to deal with . For enzymes the limiting rate is about 10^9*[enzyme][substrate] reactions per second. A larger nanomachine will have a larger surface area but substrate diffusion will be the same, so they might have perhaps four orders of magnitude higher rates simply by having lots of parallel enzyme-like devices on their surface.
And if you engineer things so that diffusion is not the way of supplying stuff, like adding a conveyor belt, then the rate can go up even further. Action frequencies on the order of 10^13 movements per second are plausible.
So one way of estimating limits is heating. If there is density rho of nanomachines per cubic meter in a watery solution and we do not accept more than one degree C temperature increase per second, then the maximum allowed dissipation is 4.2e6/rho Watts. So if we want to have one nanomachine every micron (a really dense fog) the limit is 4.2*10^-12 Watt each. That is about 26 million chemical bonds (~1 eV) made or broken per second per machine. Were they to go up to a billion reactions per machine (like the enzymes), the whole energy output would jump by a thousand times and the whole volume would boil over (or, they would run out of energy nearly instantly, if powered just internally).
Yup. Pure oxygen-charged respirocytes are a fine explosive. Nanomachines intended for free use are made with inert shells, like sapphire (Al2O3).
I wonder if the high reaction rates we get from having small objects diffuse rapidly through the air or other fluids mean that fights between guardians and other swarms should be settled much faster than in a dozen turns? The "toner war" might be over nearly as fast as they swarms mix.
Yup. The power densities can be pretty big, though.
http://www.nanomedicine.com/NMI/6.2.htm says:
"For example, a 1 micron^3 storage device with storage density of 2 kT/nm^3 (~10^7 joules/m^3) contains sufficient energy to power a 10 picowatt (pW) nanorobot for ~1 second. Chemical storage devices (providing up to 10^11 joules/m3; Section 6.2.3) may extend this duration to 10^4 sec (~3 hours)."
Check out http://www.nanomedicine.com/NMI/4.1.htm and subsequent sections.
Generally, nanomachines are bad at "seeing" long-range. Doing image-formation using electromagnetic or acousic information doesn't work well due to diffraction. The machines are extremely myopic, and will mostly know about their immediate surroundings through "touch".
This is also why nanoswarms or colloid suspensions of nanobots are fairly stupid and inefficient. it is when they are linked together into a C3I structure they become really capable. The only benefit of swarms is that they can cover an entire volume without the need for much planning or organisation: they are quite worthless compared to integrated nanosystems for manufacturing and other productive uses. (So, in my game protean swarms typically form an ad-hoc nanofactory on surfaces rather than just slosh around, and they are orders of magnitude less efficient than a fabber)
http://www.nanomedicine.com/NMI/10.2.htm
"Thus while nanomechanical computers may be limited to switching speeds of ~50 picoseconds, electronic devices may be ~10^3-10^4 times faster. " (he then points out that nanomechanics is EMP safe)
"Drexler's benchmark mechanical nanocomputer design has 10^6 interlock gates, 10^5 logic rods, 10^4 registers, an energy-buffering flywheel and other components with total cubic volume (~400 nm)^3, mass ~10^-16 kg, and total power ~ 60 nW, giving a power density of ~10^12 watts/m^3. Power dissipation per logic operation is ~0.013 zJ per gate per cycle, giving (including register dissipation) ~2 x 10^4 operations/sec-pW. Processing speed is ~10^9 operations/sec (~1 gigaflop) or ~10^28 operations/sec-m^3; assuming one bit per register, processing speed is ~10^13 bits/sec or ~10^32 bits/sec-m^3."
This means that the 400 nm computer would have a power roughly comparable to Pentium II. But as the above discussion on heat dissipation points out, you can't pack them very densely without having plenty of cooling or running them slower.
Reversible computing can get around a lot of energy dissipation issues.
—
Mon, 2012-08-06 07:08
#3
Xagroth
Mmm... I just got this crazy
Mmm... I just got this crazy idea: a thermal conversor "tied" to a CM of sorts, to recycle the energy radiated by the nanites into electricity again. That might be a standard part of the machine, but the problem gets out of hand when a Titanian microcorp has an accident in the shipyard where they are testing this with nanoswarms to build spaceships safer, faster, and cheaper (since in space you can't just drop the nanos because of the heat dissipation issues: at the very least, you need to have some sort of pressurized hangar bay to work as some sort of huuuuge sized CM (defeating the purpose of using a nanoswarm, essentially). The idea was to derive this tech into faster mining options.
The problem, of course, is that the tech was derived from some "sanitized" TITAN tech that was not as scrubbed as it should have been. So there is now a situation quite dire in the orbital system, and let's hope we are not going to see some mass production of fractals. Or a TITAN warship!
Oh dear... imagine that the unleashed system would take all the materials in the storage area, and use the whole shipyard as part of the design... There should be some far-reaching consequences, specially so close to the Jovian Republic... So for the sake of stability, a Firewall op gets underway with as many operatives as they can muster, to try to wrap this up before it gets too noticed...
Mon, 2012-08-06 09:30
#4
Arenamontanus
Xagroth wrote:Mmm... I just
Sorry, this doesn't really work. Yes, you get waste heat from some processes and you can turn some of it into extra work. But the extra work is a fairly small fraction of the energy you put in: that is why it is called waste heat. Entropy always increases. Your fridge cannot run on the waste heat it produces.
CMs are typically aiming at using as little energy as possible. In theory it could work by disassociating all molecular bonds in the feedstock and then assembling new bonds in the product, paying just the energy difference. In practice this would not work because you need to pay big thermodynamic costs for these operations, and it is reasonable to expect that they are a decently large fraction of the total energy involved. For a largely carbon structure this means on the order of around a few hundred kJ per kilogram (the total bond energy of a kilogram diamond is around 16 MJ). A smarter way of building things is to have semi-finished parts in the feedstock and just slot them together: the fewer chemical bonds you need to make, the better.
In the past I have estimated that an Eclipse Phase CM typically runs about as hot and energy-guzzling as a typical current car engine. Desktop designs likely do things slower and with less energy, relying on smaller workpieces. A big ship manufacturing plant is going to be a major industrial system with large cooling networks and big superconductor cables distributing energy.
That said, a nano-fabrication system infected with TITAN code can do some pretty amazing and unsettling things. There are plenty of surprising energy sources when you are creative. Including the nuclear batteries of synthmorphs and weapons.
—
Mon, 2012-08-06 12:07
#5
athanasius
crazy 2.0 icecube strategy
funny idea htat can work..
Titanian strategy for hangar CM:
- make an huge habitat like sealed spacedok
- put in some big icy methane ateroid
- launch nano
- the nano convert CH4 in carbon nanotube and other useful preassembled part freeing H2
- the thermal energy is used for melting ice
- the process create an H2 atmosphere inside the spacedok that si used as thermal transfert medium and evaporative cooling
- the H3 and D found is used for energy production
final result a good spaceship produced from some big "icecube"..
for extra thermal disipation use some metane as a Droplet Radiator, ehpelling gas and recollecting icecube for reuse, for optimization is possible to mix nanoscale,micro and standard tech exploitihg the best of evrything: nano for material engeneering, micro for component positioning and robotic for macroscale placment. An even more extreme metod is use an ice asteroid as hangar revestment, if jou can obtain a microscopic gravity you can evaporate the surface as a droplet radiator and lets gravity recollect the mass, using H2O as medium is possible ignore a lot of surface for thermal dissipation.
the true power of outher rim is the cold so is logical use it at maximum extent.
Mon, 2012-08-06 12:08
#6
Prophet710
with nanomachines acting as a
with nanomachines acting as a swarm and forming a liquid-esque material could they not transfer this energy using some kind of conduction into some kind of heatsink? Or transfer the swarm itself in a body of something like liquid nitrogen? Since we're talking about in a vacuum I'm assuming we wont be too worried about combustion.
—
"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."
Mon, 2012-08-06 13:29
#7
Arenamontanus
athanasius wrote:funny idea
I think the method is energy limited. CH4 has enthalpy of formation of −74.87 kJ per mol - you need energy to separate methane into carbon and hydrogen. Obviously same thing for melting the ice (333.55 J/g). Deuterium is about 0.0156% of all hydrogen, and needs to be separated out. Then it can be burned at a few MeV per atom: 10^14 J per kg, or 15 GJ/kg of hydrogen, or 1.6 GJ/kg of ice.
So I think this hangar needs to start with a fusion plant and some fuel, crunching the ice to separate out elements and more fuel. You cannot start with just a pocket seed and some solar power (since methane asteroids are not viable in the inner system where solar power is effective).
—
Mon, 2012-08-06 13:40
#8
Arenamontanus
Prophet710 wrote:with
Yes, efficient thermal conduction is doable - from what I have heard, graphene and fullerenes can transmit heat much better than copper when handled right, and if you use superfluid helium 4 it has a thermal conductivity >100000 W/mK, about a hundred times better than graphene. And if you build in fractal nanochannels for heat transport you can cool big systems at a very high rate.
However, I don't think you want to use this for everyday devices. Having your implants explode due to a cooling failure can ruin your day.
—
Mon, 2012-08-06 16:31
#9
athanasius
Arenamontanus wrote:CH4 has
I know you need energy for the work and is not a free meal, working outside out nanomechanically you use energy for CH4 processing and use waste energy for melting: the substrate is used forcool down the process.the D-He3 mining is a paranoid method for maximize efficency...
inside a human body you can use the blood as working fluid for cooling, in space the problem is a bit wrose: no fluid no thermal carry, all the heat is radiated as photon, alternative strategy is boil off something and let's it carry away heat.
For calculate coolant issure i use thias site http://www.5596.org/cgi-bin/coolant.php.
No problem if you use a lot of nano with small energy output, hi energy method unapplicable with standard conception.. and you must put the energy for nanowork.
EP standard setting is a logical macroscale for mining and large scale fast building and nano for subsystem and tech, i favor this taste too. Self replicating swarms can use active cooling strategy and structural building of needed infrastructure, ehergy thermal and logistical. By setting this kind of swarm is TITAN designed and logically neel singularity level intelligence for programming, is more like think about this as a megascale architectural progect executed by nanoscale workers, thinked for be applicable in a lot of different enviroment.
Tue, 2012-08-07 18:33
#10
athanasius
what we are talking abouth
A nano for fluid enviroments (air-whater-solid surface in fluid atmosphere).
[img]http://fc00.deviantart.net/fs71/i/2012/220/b/7/a_nanomachine_by_demigogo...
this can be considered a "general purpose" nano, functionig as a programmable matter compiler, .1 µm max linear dimension.
The "mouth" is composed of ative compilers and is backed by a lasing cristal for ionizing atom, sensor package is a phased array of optical nanodots for UV and IR using active or passive mode.
The arms use geko derived van der Waals feeler-adesor, used for station keeping and large scale positioning.
The bulbous body is the main energy and chemical storage,the power reciver microwave antenna feed power to reserve for conversion and storage for safety purpose (transitionary pulse are filtered by capacitor like elements used for power storage). The mesh is composed of conductive nanotube, used also as radiator for thermal dumping.
The computer is composed by doped graphene sheets layered for create a 3d structure acting as a cpu with integrated OS and memory, programmign and internano comunication are recived by frequency optimized wireles recivers (also used for cooling).
The propeller is a flagellum designed for air-whater propulsion with a mak 17000 rpm, max velocity is 600 lengths/second.
This kind of nano is usable in evry liquid enviroment (air-whater) and surface of object in liquid enviroment.
More limited nano can be much smaller but loosing smartnes and speed, operative time can also decrease for very limited storage. Single purpose nano can be only enzime scale but control is limited to on-off, are free floathing and must be external powered: this is not as bad as sound, respirocite are of this kind, a buckyball with glucose powered pumps for oxygen and CO2 and a simple automated reflex calibrated for mantain optimal distribution of gas mixure.
For space capable nano delete the propeller and make power reciver bigger, MUCH bigger, the propulsion is too expensive in term of mass/energy to be used so they must walk on solid or floath inertially in space.
For CM using nanoassemblers is unuseful, multiple fractal digit arms with nanochannals for feedstock and cooling, power is externally feed. Evry nanoarms is provided with sensor and manipulator and directed by the comp in the upper branch, evry branch is controlled by their respective upper lavel comp, this for distributed micromanagment. The CM computer act as programmer for lower levels, evry level recive discrete objectivized steps and report for complete or errors at the uppermost level for correction and coordination.
If someone have better design i need new prospective... ;)
Thu, 2012-08-09 16:37
#11
athanasius
update
sorry, some time to figure active site mechanic, i have 2 proposal:
1 Scanning Tunneling Microscope: the tip can tead single atoms nad manipulate them, with atomic precision. The energy for broke molecular bond can be feed as electric potential or as EM radiation of UV or X-Ray laser. This process need controlled enviroment, this can be obtained in a CM machine or in a limited space suc as beetwen mandibles of my pruposed nano.
STM [url]http://www.nanoscience.com/education/STM.html[/url]
2 enzimatic catalizer: the active site is actually an artificial enzime, this is much more limited but some order of magnitude more fast, you don't need to catch the atoms tey are actracted to enzime receptor and the bonding is a single multiproces. This process is more useful for biological works.
An ipotetical process for TITANs active sites is quantum uncertain manipulation for controlled waveform collapse: instantaneus positioning and bonding...
The limitation are intrinsecally bound by the phisical nature of the process:
1: the atom must be taken by storage and placed one by one, for disassembly you must find the atom and supply energy for ionize it and then store... thermal waste and time
2: the enzime is designed for a limited set of ambient phisical pharameter (wharer presence, PH, temperature, pressure...)
The 2 metod can be both in use for fractal digits of CM machines synthesys.
A gift an article speaking of the first experiment of nanotech (by IBM)
[url]http://www.scilogs.com/from_the_lab_bench/super-hero-experiment-3-what-l...
[img]http://www.scilogs.com/import-data/images/11/IBM.jpg[/img]
Sun, 2012-08-19 01:01
#12
NewtonPulsifer
Considering a buckyball is 1
Considering a buckyball is 1 nanometer.....isn't 100nm too small for a general purpose nanite?
To compare:
Smallest virus - 42nm
Smallest bacterium - 200-300nm
Edit:
Also, microwaves are in the millimeter wave range. A nanoscale version of a microwave rectenna would probably have to convert frequencies in the visible light range:
http://en.wikipedia.org/wiki/Nantenna
—
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto
Sun, 2012-08-19 04:05
#13
Arenamontanus
General purpose nanites doesn
General purpose nanites doesn't make much sense in most situations, either. General purpose machines are usually far bigger and more complex than specialised ones, so if you want to do something specific you construct a specialised nanite for the job. General purpose nanites will be useful in a few situations where you have no idea from the start what challenges you will want to meet, but they are not going to be very good at meeting it. It makes more sense having a general purpose assembler system to make required specialised nanites.
I think the nanite sketch above is pretty small, but it is not out of line with existing sketches for nanites.
—
Wed, 2012-08-22 18:08
#14
Prophet710
OK so, left field question.
OK so, left field question. Smart materials as it concerns wearable clothes and other general purpose textiles. Would these basically just be typical polyester/silkthetic materials that are laced with an array of nanites to change over to what ever you like?
As far as manipulating things on the nano level, we've covered that the typical way to do this would be chemically (I think, if I've absorbed this right), what about using energy? If you can use a swarm in a medium to conduct energy and bleed away thermal energy you should have an efficient process to use something more fundamentally powerful than a chemical reaction to construct and change future materials no?
—
"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."
Thu, 2012-08-23 08:48
#15
Arenamontanus
Prophet710 wrote:OK so, left
You can do this in a lot of ways. You could have a lacing of nanofibers that can signal and contract, changing the properties and shape of the clothes. Or the clothes could just be nanofibers. Or even indpendent nano- or micromachines that link together in different configurations. It is mostly a design choice.
I don't think this makes physical sense (or I misunderstood you). Chemical reactions are just changes of where energy and different atoms are. There is no firm boundary between different ways of manipulation. You can move atoms using optical tweezers, shooting them with a little accelerator, moving them to or from a STM tool tip, random diffusion, biomolecules binding at particular surfaces and a myriad other ways. The key problem with controlling chemical reactions without having some nano-thing right at them is that they are essentially random: thermal energy is random motion, atoms and molecules absorb and emit energy randomly. And you cannot direct energy well enough over long distances to control nanosystems unless you have a lot of feedback.
The reason nanotech is so exciting is that it may allow reactions to happen without needing to throw things hard at each other and hope some stick, and instead just use precision control to get things to the right spot. Controlling thermal motion on the nanoscale is part of the challenges in designing nanomachinery - things must not shake apart, and the processes that are done by the machinery must not cause too much random shaking.
—
Thu, 2012-08-23 13:15
#16
Prophet710
I was thinking instead of
I was thinking instead of using chemical reactions you could use magnetic fields and other forms of energy like mini lasers. They seem to be a bit more precise and efficient than chemical cocktails.
—
"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."
Thu, 2012-08-23 14:59
#17
Arenamontanus
Prophet710 wrote:I was
More precise than chemical cocktails, but less precise than nanomachines.
Magnetic fields produce forces roughly proportional to the number of magnetic atoms in the thing being subjected to them. So most molecules are not affected at all. The big problem is that things making magnetic fields have to be pretty large (either to have atoms with unpaired spins making magnetic fields, or to have currents circulating to make the field). There are some cool microdevices that make micron-sized magnetic fields, but they are too big to manipulate molecules.
Lasers can also move atoms - you can keep ions suspended in vacuum by setting up an optical trap where they get locked into the troughs of intersecting wavefields, and optical tweezers work fine. Assuming the atoms or molecules are charged so they react to the field. And again, the resolution is optical - hundreds of nanometers.
A STM tip can place atoms with 0.1 nm lateral resolution and 0.01 nm depth resolution. And of course enzymes move functional groups from one place to another in a molecule.
—
Thu, 2012-08-23 18:35
#18
Prophet710
OK, so another left field
OK, so another left field question. How do you get seamless construction then? Simply bonding the atoms together at said seam using any of the methods detailed above?
—
"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."
Fri, 2012-08-24 04:39
#19
Arenamontanus
Prophet710 wrote:OK, so
Works well. A seam is after all just a plane where the structure is slightly different from the bulk: if you can put atoms in exactly the right places you can make the two bulks continuous with each other.
In practice this is likely overkill. A few dislocations won't kill you if you manage them right. So it might be more useful to make parts in the form of "lego pieces" that snap together: protrusions make sure they are aligned right, shapes are made such that there is an irreversible step as they snap together. Remember that super-clean surfaces often bond together spontaneously (contact welding), so if you make your nano-pieces in a controlled environment you can put them together fairly straightforwardly.
The real mess is when pieces get large, have to be made in a messy, shaking environment or involve soft materials.
—
Sat, 2012-08-25 11:37
#20
NewAgeOfPower
Ah. How strong are contact
Ah. How strong are contact welds? I seem to recall they require a fairly high pressure to conduct anyways...
—
As mind to body, so soul to spirit.
As death to the mortal man, so failure to the immortal.
Such is the price of all ambition.
Sat, 2012-08-25 12:26
#21
Arenamontanus
NewAgeOfPower wrote:Ah. How
Depends on what surfaces you have. Consider that two molecules reacting are a case of single-atom contact welding. So you could for example set up the two sides so they react and form covalent bonds when touching.
—
Fri, 2012-09-07 11:17
#22
Prophet710
OK so, another left field
OK so, another left field question. Since we are looking at reconstructing things at the molecular level. Would there be a way to reconstruct molecules at the sub-atomic level without the use of pico/femto technology? If so, then, could you literally reconstruct atoms into different elements, i.e. making silicon into carbon, making carbon into iron, etc. From the way I see it, its a simple process of adding or subtracting sub-atomic particles and seems feasible.
If the above is true, couldn't you then construct anything you want using basically local sand and basalt?
—
"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."
Fri, 2012-09-07 13:52
#23
NewtonPulsifer
No.
No.
What you're proposing is called nuclear fusion (fusing carbon to iron) or fission (fissioning iron to carbon).
The difference in energy between atomic nucleus bonds and bonds dealing with electrons (ionic and covalent bonds) is on a vastly different scale (millions of times the difference). If you want to read up on it:
http://en.wikipedia.org/wiki/Nuclear_binding_energy
http://en.wikipedia.org/wiki/Electron_binding_energy
Realistically, even de/re/constructing certain molecules on a molecular level with nanomachines is questionable, if the ionic and/or covalent bond is one of the strongest available in nature.
To give an example, how would one propose to remove one atom of carbon from a 2.5cmx2.5cmx2.5cm synthetic diamond by 400nm long nanomachines? Or write "NANOBOT WAS HERE" as small as possible (1 atom deep) on that diamond? I cannot think of how.
—
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto
Fri, 2012-09-07 14:03
#24
Arenamontanus
Prophet710 wrote:Would there
Well, changing the nucleus *is* pico/femto technology. Currently we do it like we do "nanotechnology" - we throw a lot of things hard at each other, and carefully pick up the ones that stick right. The promise of real nanotechnology is to actually put molecules together in a deliberate way. Similarly for atomic transmutation technology.
However, it is a tough goal. We do not know any tools that work well on the nuclear level. We do know tools that work on molecules - enzymes, STM tips, various proposed molecular machines - but nothing like it for nuclei.
Any technology also needs energy and getting rid of entropy. Changing the nuclei of sand into aluminium means that some protons and neutrons need to go somewhere - and that you need to pay a nuclear binding energy cost for doing the fission. Maybe you can balance things by fusing those excess particles with hydrogen "fuel", but it is unlikely that you will be able to confine all the released gamma photons and neutrinos. I suspect working femtotech is actually pretty dangerous to be close, the equivalent to mechanical vibrations will be gamma-rays. And the energies involved will be like the ones involved in nuclear reactions: you better have good cooling.
—