/sci/ - Science & Math

[Anonymous No.11778049]

what do you think of programming language theory?

[Anonymous No.11778181]

bump

i was instantly red pilled when i learned you cant add or subtract decimals without a float error.

The rabbit hole went into the realm of questioning why we abandoned analogue for digital in the first place.

I think the field will be relevant in the years to come. Modeling computations as energy minimization problems like what Dwave Quantum computers got going on looks like a promise tool to me.

[Anonymous No.11778205]

>>11778181
>Dwave
>Quantum computer
Pick 1 and only 1.

[Anonymous No.11778215]

>>11778205
how does this https://www.dwavesys.com/take-leap cloud service not fit your definition of quantum?

But based, i don't think dwave tells the whole story

[Anonymous No.11778531]

>>11778181
You're getting pretty high on the confidence/knowledge scale kiddo. We don't use analog because it's harder to make consistent and it's limited in precision by noise. So instead of adding 0.1 to 0.9 and getting 1.0004 every time like in a digital system, you get 1.00053 one time and 0.999988 the next. And the behavior will be heavily impacted by shit like temperature, age, the quality of your power supply, and noise in the environment. We left analog behind for very good reasons. The "issue" with decimal math is not a real issue either. You can do math to arbitrary precision or symbolically, it's just slow. Float operations are for when you can tolerate error and you need the calculations to be done as quickly as possible.

[Anonymous No.11779073]

>>11778531
>confidence/knowledge
if slash is 'per' than noted if else 'or' than thanks

eyy cool, sounds like its endemic to a resistance of a circuit. Would you look at analogue again, if the engineering could remove the parasitics, like making your circuits with superconducting wire?

arbitrary precision is my gripe, in fact, i can't help put feel like nature porports quite a explicit degree of it, not too hot, not too cold.

Let precision be defined as the simplest expression required to act upon purely a set of natural numbers.

I fear in our approximations we're missing out on subtle but important physics

I dunno, you ever feel though, that we get more information than we need from the mod 9 times table? I feel like primes got this special prana about them

[Anonymous No.11779286]

>>11779073
Has little to do with resistance. Noise is noise. You can get rid of resistive noise, but sometimes you NEED a resistor, and now there's noise. But there are other sources as well. Have a semiconductor? Noise is intrinsic to its operation. Noise is in the environment, in your inputs, everywhere. If you could remove all noise and have perfect analog operation then you could make a Real Computer. This is impossible because physical things in our universe are quantized when you get right down to it. To have unlimited memory you always need unlimited space. This limitation manifests itself as quantum noise in an analog system.
If you want to manipulate exact values, they must have finite precision or be symbolic.

[Anonymous No.11779953]

>>11778049
>>11778181
FINALLY SOMEBODY GETS ME

[Anonymous No.11780238]

but what's the future then?
What do I do? What language do I learn? How do I optimize computers?

[Anonymous No.11780366]

>>11778049
PLT is pretty legit, type theories and categories are fun.
>>11778181
This isn't exactly PLT.
> Modeling computations as energy minimization problems like what Dwave
Quantum computers got going on looks like a promise tool to me.
No. The *actual* work here is from theorists in both CS and physics - look up gapped local hamiltonian complexity.
>>11778205
>>11778215
dwave is a meme
>>11779073
learn more CS before spouting bullshit. There is literature about computation over arbitrary field structure, but the problem here is realizability in a way that analog computing isn't the *exact* answer.
>>11780238
>What do I do?
idk what you like
>What language do I learn?
whatever language suits your needs, and if it doesn't exist, write it
>How do I optimize computers?
this is a broad question that has no well defined answer. Optimize *what* about computers?

[Anonymous No.11780531]

>>11780366
I feel this post is generally reflective of the maligned condition of society divorcing its mathematicians from its engineers. The arbitrary field structure that i attest to here is the omni-dimensional hypersphere.

Another negative too d-wave, i've used it several times, it feels way less of a meme than ibmq and rigetti

[Anonymous No.11780594]

>>11778215
>Rodin maths
>>>/x/ is that way