74
$\begingroup$

Following on from this question and assuming the race in question had developed the ability to create tools how would they approach the need for mass calculation.

Our early computers were used to calculate large amounts of financial information, how would an underwater species approach this problem (I'm assuming electrical systems are probably out of the question but I'm happy to accept answers which prove me wrong).

I'm looking for a technology which could do large arithmetic calculations and be built and function in an underwater environment.

$\endgroup$
  • 23
    $\begingroup$ I don't know, but they would likely be susceptible to damage from air ;-) $\endgroup$ – smci Nov 6 '14 at 19:12
  • 3
    $\begingroup$ I like all of the answers, but I will point out that the answer may be simpler than we think. A quick layer of insulating material would quickly reduce the complexity of building a water-borne computer to almost exactly the complexity of building an air-borne computer. Besides, most of the interesting discoveries were done in labs under highly exotic (read: expensive) conditions, just like the exotic conditions to do it under water. $\endgroup$ – Cort Ammon Nov 7 '14 at 5:40
  • $\begingroup$ If you want them to have electronics, a first step would probably be to make them develop safe uses of electricity. Given the history of electricity discovery and early uses for us land based animals, this would be the toughest part. Anything from there on is just basic evolution. $\endgroup$ – PlasmaHH Nov 7 '14 at 10:48
  • 1
    $\begingroup$ @PlasmaHH I've deliberately phrased the question to ask for a computational machine - not an electronic system. Electricity may not be the best mechanism (in fact the top answer suggests it isn't) $\endgroup$ – Liath Nov 7 '14 at 11:22
  • $\begingroup$ I'd wager that even underwater people would still use electricity. They would just face different challenges with their ambient environment. Electricity still works, it just need insulating. $\endgroup$ – dibs Nov 9 '14 at 5:45

19 Answers 19

83
$\begingroup$

Instead of electronics they could develop fluidics: http://en.wikipedia.org/wiki/Fluidics

We've built fluidics logic circuits to control ICBM's and rockets and nuclear reactors - basically environments that are harsh/destructive to electronics. But we sort of stopped developing fluidics further as hardening process and redundancy algorithms improved to the point where electronics can be used in harsh environments (a side-effect of our space program where space agencies had to develop electronics to withstand space).

One down side of fluidics is that it has a maximum clock frequency of tens of kilohertz. But as we've seen since the late 90s, electronics also have a maximum clock barrier (apparently at roughly 4GHz for practical purposes). This doesn't mean that computers are impractical. It may just mean that they would try the parallel, multi-core approach earlier compared to us.

The slow clock speed, multi-core limitation would also put pressure on software development to write programs with hundreds of small tasks where each task does very little but working together in parallel can achieve high computing troughput. We've played around with this in the 80s. A notable implementation is Thinking Machine's Connection Machine: http://en.wikipedia.org/wiki/Connection_Machine.

$\endgroup$
  • 1
    $\begingroup$ "where each task does very little but working together in parallel" sounds like Erlang. $\endgroup$ – Naftuli Kay Jan 25 '16 at 18:43
  • 1
    $\begingroup$ @NaftuliTzviKay: The Connection Machine is a much more impressive example compared to Erlang. Google it. It's a supercomputer made of thousands of 1bit CPUs (yes, 1 bit, not 8 or even 4). So doing 8bit addition for example requires the use or 8 (or is it 16?) CPUs. The assembly language is actually closer to designing digital logic (FPGAs etc) than it is a programming language. $\endgroup$ – slebetman Jan 25 '16 at 19:45
  • $\begingroup$ We're probably fairly far from the limit in the clock speed of electronics, but we've pushed silicon to it's physical limitations and started adding germanium to it. Silicon has been convenient for making micro and nano scale semiconductors, but future improvements will require more radical material changes that shake up the whole fabrication process we've been improving for so many decades. $\endgroup$ – Stephen Lujan Feb 14 '18 at 20:08
  • $\begingroup$ @StephenLujan We still haven't seen any improvement over the 10GHz CPUs we had in the 90s. It's been close to 20 years now that we're stuck at 4GHz. I bet if we come back here in 2040 and check this comment we still won't exceed 5GHz and most probably still be stuck at around 4GHz but we may be running hundreds of cores $\endgroup$ – slebetman Feb 14 '18 at 20:54
  • $\begingroup$ @Slebetman I think perhaps they found less reason to push clock speeds higher. Besides the advancements in parallelism, the number of instructions that can be processed in a single clock cycle on a single core has also increased significantly. There is also focus on power efficiency, since battery powered devices are more common, and because moore's law hasn't slowed much and we're still cooling processors with a little piece of aluminum and a tiny fan. $\endgroup$ – Stephen Lujan Feb 14 '18 at 21:33
47
$\begingroup$

Awesome! I can talk about what I think is one of the coolest inventions of all time: the analog computer.

An analog (more properly, a mechanical) computer was actually the world's first "computer" - the Antikythera Mechanism. It was an ancient Greek device that predicted the motions of the planets and other astronomical objects. Only parts of it have been found, but we can figure out a few of its basic properties. It made its "computations" using an elaborate system of gears.

Fast forward a couple millennia. Let's go to England, and meet a man by the name of Charles Babbage. I'm guessing you've heard of him. He shows up in a lot of books on alternative history, because he's hailed as one of the first computer pioneers. Babbage's first "computer" was the Difference Engine, a glorified calculator that could work with polynomials. It used a lot of gears to work with polynomials and do advanced calculations. Babbage got some funding from the government, but not a lot.

Later, he worked on his more advanced Analytical Engine, a machine that never got off the drawing board. It would have used punchcards and a more advanced gear system to perform "general purpose" calculations. Unfortunately, funding dried up, and Babbage never built it. Parts of it have been built, but it has never appeared in full form.

An underwater civilization could certainly create one of these machines. The one issue would be ocean currents, which could disturb the mechanisms. Perhaps they could shield it with something, or else put it in an area of still water. They could build it out of metal - assuming they could make tools, as per your other question.

$\endgroup$
45
$\begingroup$

Don't take it personally, but if you think computers are only made of silicon, you must have sand for brains!

The first computational apparatus, primitive neural systems evolved in an aquatic environment. For 1.5 million years (or so) they have developed environments that protect them from electrical and mechanical failure due to dehydration. These systems use some elaborate osmotic pressure control mechanisms that not only maintain a narrow band of ionic concentrations (aq) but also run on their transient fluctuations.

To your specific question about computers in an aquatic environment think of the real, original neural nets. Think axons, dendrites, and synapses. Throw in a little myelin to insulate yourself and reduce repolarization times and you're on your way. Your silicon devices are chips of the old aquatic blocks and in a few decades may catch up.

$\endgroup$
  • $\begingroup$ As a former Pyschology student I have much love for the mention of axons, dendrites (yay trees) and synapses. $\endgroup$ – Pharap Nov 9 '14 at 0:58
  • 1
    $\begingroup$ Don't you mean 1.5 billion years, not million? $\endgroup$ – feetwet Sep 13 '15 at 17:53
23
$\begingroup$

(I'm assuming electrical systems are probably out of the question but I'm happy to accept answers which prove me wrong).

Let me take a crack at this one.

There are two problems with running your standard PC in water:

  1. Conductivity
  2. Corrosion

Conductivity is bad, as it allows the electricity in the wires to not actually follow the wires and end up where you need it to go. The nice thing about conductivity, is that it's all about degrees. The wires in your machine are always going to be more conductive than the surrounding material (or else they wouldn't work and nobody would make wires out of them). Salt water is very conductive. Fresh water is not terribly conductive (but air is still 10^12 better). Distilled water is even better (but still likely not good enough).

The main idea though, is that as long as the wires are insulated by something to prevent the electricity from jumping wires, it will work just fine. We use air for that because it's cheap and ubiquitous. Aquatic peoples would need something else. (though maybe not flourinert)

Corrosion I know less about, but expect salt water to be a larger issue here as well. The wires would need to be chosen with corrosive properties in mind as well as their conductivity.

Most likely, the aquatic race would simply need to seal their computers with some non-conductive, non-corrosive contents, and have the electric plugs well insulated - but they'd probably still work the same way. If anything they might be beefier since the water could serve as an effective heat sink for the entire machine rather than depending on air to carry the heat away.

$\endgroup$
  • $\begingroup$ The name escapes me, but there exist water resistant PCBs. You coat the finished board in something, that covers all exposed surfaces, and can later go swimming with the thing. $\endgroup$ – Vorac Nov 7 '14 at 11:22
  • $\begingroup$ @Vorac You might be thinking of Parylene. $\endgroup$ – s0rce Nov 7 '14 at 23:08
18
$\begingroup$

Oil immersed PCs and servers exist. They conduct heat but not electricity and prevent water corrosion.

http://www.pugetsystems.com/submerged.php

We imagine upside-down bins to contain the oil (oil tends to float up). The oil is water-displacing( WD-40 can remove rust from a jammed hinge). Electrical (batteries) and electronic (computers) apparatuses can be assembled in water and then placed in the bins, then turned on.

We can adapt most of our electronics to this method. And hey, upside-down bins!

$\endgroup$
11
$\begingroup$

I'd consider the moniac and water integrators and early analog computers.

You'd probably have a very different environment for computer design - consider that in many ways the jacquard loom was the first programmable computer, and that it used punchcards. And while there were alternatives to binary designs, they are consigned to the dusty basements of history. Without this basis, you're more likely to find analog machines.

You'd probably start with fluid-analog computers or simple hydraulic switches, possibly used to control simple machines.Eventually you may progress to logic gates, or fluidics.

Computers would necessarily be room sized since even with eventual development of microfluidics, you can't hit the same process sizes as a electrical system.

As for alternatives, I'd consider electrical circuits as being impractical in a marine environment but the techniques that would lead to microfludics (etching channels in your substrate) might eventually lead to photonic computing.

As for storage, delay line memory of some sort could work, assuming we're using analog computers, eventually. One might also consider the use of empty and full fluid channels as a equivilent of 0 and 1 in storage.

$\endgroup$
  • $\begingroup$ Excellent point on micro-fluidics possibly leading to photonic based computing... or @ least photonic information transmission (some fiber optic equivalent). As I understand it the inspiration for fiber optics was an early system of water filled pipes designed to transport light around a house (the water makes a reflective surface inside the pipes allowing the transmission of light across a distance of pipe) $\endgroup$ – MER Nov 9 '14 at 18:43
8
$\begingroup$

I think this would depend upon the depth of the aquatic civilization.

However, I would imagine that an underwater species might rely on a symbiotic relationship with other species in order to achieve something like this. Perhaps a species that uses its own bioelectricity, and based on certain stimuli could return some type of output. (Bioelectricity and bioluminescence are apparently common in very deep waters, as no sunlight penetrates there, at least on Earth.)

Depending on how your aquatic race communicates, I'm guessing that arithmetic could be achieved through light waves or sound waves. (In fact, I think it'd be interesting to see some type of mathematical system implemented with sine waves and Fourier transforms.)

$\endgroup$
6
$\begingroup$

I think they would tend towards biological computers. Many underwater creatures have finely tuned senses that can detect a myriad of impulses. Groups of creatures could link together through this network. Most species already do this to some extent. I am suggesting that the Aquarians take it to the next level. Maybe when a bunch of them get together in a small space they can zen themselves into a state whereby they really can crunch some data. A Think Tank, if you will.

$\endgroup$
  • 2
    $\begingroup$ Quite a literal version of "Think Tank" $\endgroup$ – user2813274 Mar 17 '15 at 19:15
5
$\begingroup$

As HDE226868 pointed out they could certainly start with physical calculators, (think of the abacus!, and then much later the Slide Rule). However, we need to think of a few things first.

As an underwater species how or why would they even develop writing? Chiseling into stone? possible, but paper? Writing came about to extend our memories and computers were extensions of that with the ability to perform calculations for us. So by bring this up it might mean that an advanced water race might be extremely intelligent with an incredible memory. Having to get that far without cheap 'books'.

Now as far as electricity, the electric eel does just fine in water so something might be designed along those lines, but for the big problem of refined metals to use, and being submerged in water, would leave things like gold and platinum to not rust away. So I think something with sound and tuning forks or crystals would be much more likely to be used. Of course these would have to be fairly high range sounds to have the least interference from ambient noise.

$\endgroup$
  • $\begingroup$ Electric eels don't have to limit their electricity to specific places, such as binary switches, nor do they have to keep their surroundings safe for other creatures, so I'm not sure what you're trying to prove from them. If I dropped a laptop into the bathtub the electricity would also do something, but it certainly wouldn't work the same after ;-) $\endgroup$ – Shokhet Nov 7 '14 at 3:05
  • $\begingroup$ Why would the concept of "writing" require paper? Seen in an abstract way, "writing" just means the representation of ideas or statements in an agreed-upon encoding. This may be done by coloring parts of a surface (and that might be the most straightforward way in our environment), but it is not restricted to just that way. Think of the Incan quipus, for example; that system could be used under water just as well. $\endgroup$ – O. R. Mapper Nov 7 '14 at 10:52
  • 1
    $\begingroup$ @O.R.Mapper That's true! and I forgot about the Irish used wooden rods as well... $\endgroup$ – bowlturner Nov 7 '14 at 13:29
5
$\begingroup$

We use water for infinity of things despite of we do not live inside it.

So the first thing it comes to my mind is they could go to the surface, take some air and use it inside their computers.

If the whole planet would made only of water, they could take the oxygen from the water and create air.

$\endgroup$
5
$\begingroup$

They may be able to make optical computers with lasers, diamonds, fiber, etc.

They could also remove the water from the computers - just like how we create a vacuum in order to make light bulbs, they could create a vacuum and remove water (and insert air, or something non-conductive like mineral oil).

I should say that the "tools" needed for such devices/production would be more advanced than what is linked to, however the human race didn't start making computers as soon as they could forge something either, and the previous answers include the "geared" and "mechanical" computers that could develop in a very similar way already.

One more possibility would be fluidics - basically transistor and other logic gate equivalents via hoses/pressure

$\endgroup$
5
$\begingroup$

Insulators can allow for, effectively the same method of electrical transmission we use now... in fact internet and phone cables are laid on the bottom of the ocean to connect the continents of earth.

Fun Side Effect (well to me anyway :D ):
Having stated the above I would add that it would even be possible to pass a controlled current through a tube built from an insulator that is filled with nothing more than sea water. Wires attached to, or as part of, a device would be most effectively water tight. Though, also pretty interesting, I don't think a 'power leak' would actually prevent power transmission. Instead, I'd conjecture, this would cause a great deal of attenuation in the amount of power carried.

Using water as the means of power transmission may be valuable but I suspect it would have limits in a computational device. Forming tubes of smaller and smaller sizes would be one likely limit. The boiling point of water might be another... I don't believe steam has the same conductivity profile. So in the end, for building small, fast and efficient computers, I suspect the race would discover some other, semiconducting, material (silicone does seem like a very likely choice) much as we did.

Short answer, just insulate.

Oh, also, while I had thought it before I saw the comment, I wanted to point out that this was said in a comment by Cort Ammon as well.

$\endgroup$
3
$\begingroup$

There is no need to assume that water creatures cannot use processes or technologies that require air environment.

Some manufacturing processes do require vacuum, or have required in the past (vacuum tubes, for instance). Humans live in the air but were capable of making vacuum when needed. And, for sure, they create water environment when needed (most of the chemistry).

Aquatic creatures could make air environment, or even vacuum environment in they laboratories as required. For instance, diving bell may be possible to make from some shell or water plant, allowing to experiment with air technologies to discover lots of amazing things.

In the Earth, spiders create air environment under water without any intelligence.

$\endgroup$
2
$\begingroup$

So...don't think the question is "How do you build a computer without electricity?" as much as it is "How would an aquatic species build a computer?" So the Difference engine is a good start, but: 1) There is no reason to believe an aquatic species could not go on land and develop electricity, much as we have run electrical cables across the Atlantic Ocean or gone in electrical powered vehicles under water or used water to generate electricity ourselves. No reason to believe the opposite couldn't be done (although obviously it's easier if your species has a stronger skeletal structure, etc., point for another time). 2) As I said before analog computers starts down this road but to take that point a bit further you do not need electricity to build a computer, only some way to create a register that holds a value -- which could quite easily be done with a hydraulic system. This concept is explained in some depth in Daniel Hillis' excellent book Pattern On The Stone. (amazon.com/The-Pattern-On-Stone-Computers/dp/046502596X), and a Soviet engineer built a hydraulic computer in the 1930's (http://makezine.com/2012/01/24/early-russian-hydraulic-computer/).

$\endgroup$
2
$\begingroup$

Lot of interesting answers here, but it really is very simple: even if they are "aquatic" and live underwater, they could still have their computers - and anything else - on land, just like we have diving equipment and can do sophisticated things underwater.

Having said that, they would probably start off underwater so their technological trajectory would certainly look very different from ours.

$\endgroup$
2
$\begingroup$

Well simple and it already exists on the sea floor. there live bacteria that are linked together by bio-fibers. They are able to exchange electrons over those wires. As a way of feeding (fermenting) garbage on the sea floor. So where there is an overdosis of + they exchange that with bacteria where there is an overdosis of - charge, they can live on give-away and taking electrons (its remarkable). And they are all connected. Its far more dense then the internet.. Structure has something in common with neurons, but we dont know if it thinks yet...

$\endgroup$
  • $\begingroup$ +1, Good answer! Welcome to the site! I think you could even improve it, if you would include some ressources (maybe some scientific paper?) on these bacteria $\endgroup$ – T3 H40 Feb 12 '16 at 15:39
1
$\begingroup$

Cetaceans in general have the capability of using sound underwater in highly sophisticated ways. Maybe it would be possible to simply excavate resonant cavities in the sea floor and have schools of specially talented cetacean-like "choruses" sing a program so that the sound waves interacting with the cavities would produce output that could be directly apprehended.

$\endgroup$
0
$\begingroup$

OK, so this is assuming that an electric source has been developed.

Currently, we do have water-proof technology. This is made waterproof by developing the technology in question, and then coating the entire device in a very thin layer of atoms. This layer of atoms has the power to repel water and/or other liquids.

However, applying this process underwater may be much harder. In the current world, we can develop it and then apply the atoms. This means there is two options for the under water race. They can find a way to use the above ground temporarily - in a way that humans use scuba diving suits to go underwater, they could develop suits to go out of water - and develop the technology there, or use waterproofed primitives (NAND gates), which means that everything they build is waterproofed from the very start.

Another problem is these atomic layers work for brief exposures. The underwater race may coat it in many layers of the material, meaning it can be exposed for however long necessary (although this will stop the end device from being touch-screen, because after so many layers, touches will not reach the touch sensor.)

$\endgroup$
0
$\begingroup$

First, you should think about the evolution of the species along with technology. They probably don't make fire as a primitive technology. Do they have hands? Tentacles? Which materials and tools they have?

Then, they could be manufactured using living beings or their parts, as I propose in this answer.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.