[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math


View post   

File: 3 KB, 250x83, IMG-20180730-WA0001.jpg [View same] [iqdb] [saucenao] [google]
9931745 No.9931745[DELETED]  [Reply] [Original]

Computer science, computer systems engineering or electronic engineering? UK btw

>> No.9931783

>>9931745
CE here. Depends on what you want to do, obviously.

>> No.9931793
File: 5 KB, 262x292, 0000.png [View same] [iqdb] [saucenao] [google]
9931793

>>9931745

>> No.9931963

>>9931793
This is a shitpost. Undergrad CS is wildly inconsistent, but

>CS theory/grad CS
>bachelor's degree into industry work with a heavy focus on compilers, OS, HPC, scientific computing, cryptography, algorithm design and analysis, and working on engineering projects

These are all fine fields. It just depends on what you want to do.

Do Computer Science (and probably a double major in math for good measure. My uni was non brainlet, so we had the hard math courses as prereqs anyway) if you like solving hard math problems. Writing code is something borne of habit. Learning how to solve hard problems, how to determine what parts are "solvable" and which ones can be estimated, etc. is actually pretty difficult in practice. If you double major in math and CS, it gives you the world back in usefulness. If in the future you want to do engineering, you can just do a masters in an engineering and apply for R&D departments (most other positions will be locked behind the PE exam, so you're probably gonna be targeting research in the industry). CS theory leads into lots of pure math, lots of natural science (complexity theory is being used to study everything from the physical complexity of black holes to why protein folding has metastable states), algorithms, etc.

Do Computer Engineering if you like the implementation of computer systems on both the physical and software side. It's the EE theory behind the construction of a computer and learning enough CS to understand the construction of software interfacing between machine language and userspace. There's this meme on /sci/ that it's CS+, but not really; it's just a different focus. Problems are mostly about finding new ways to make existing computers more efficient, graphics card design research, etc. etc. To be honest, there's actually a good amount of intersection between CS and CE jobs, especially on parallel computing, GPU interfacing and programming, and distributed computation.
(cont.)

>> No.9931973

>>9931963
(cont.)
Mircocontroller and embedded programming sees a lot of CS and CE people, depending on their interests. Grad school will have you working on developing increasingly complex computer systems. it's also a great field that's in the middle of EE and CS, though definitely more on the EE side

Do EE if you more electronics than just computers. There's power (though that's a hard field to break into), computers, robotics, etc. EE gives you a pretty wide skill set to do a lot in engineering. You learn some pretty interesting stuff, especially on signals and communication. As far as what you'd see with the intersection with CS, there are a lot of cross listed jobs in information theory, encoding, and communication, though RF is really just EE specific. You can do grad school in whatever type of electronic system (small devices, controllers, computers, etc) you please and it'll be fine.

I hope this clears things up. All three fields are pretty great, and if you really care about them, you'll go out of your way to take the hard and interesting stuff they all have to offer. CS (especially in grad school) is the most theoretical of the 3 and has its foot in the heart of math and science, while the other two are proper engineering fields that teach a lot of interesting and important skills for the field. Follow your dreams.

PS: take a lot of math and physics with CS. This will give you a big edge. Math/CS and Physics/CS are powerful combos, especially if you get into theoretical CS

>> No.9932000
File: 297 KB, 836x1136, 1513067679808.png [View same] [iqdb] [saucenao] [google]
9932000

>>9931963

>> No.9932004

>>9932000
your image does not apply the the anon you're quoting

>> No.9932025

>>9932000
Okay, lemme redpill you on this:
Nobody past their freshman year (and subsequently their intro to proofs class) actually talks like this. Nobody in academia is talking about mergesort; that was developed in the 1940s. Hard, non-intuitive algorithms and their proofs exist, even in undergrad.

Basic CS proofs use induction and structural induction, yes, but as this is CS, most proofs are either by contradiction (prove X isn't in Y complexity class) or a constructive proof (devise an algorithm to solve a small to prove that a property in a class of problems exists). It goes from algorithmic proofs to things like the proof of a computable representation of the nth digit of pi in hex

Once you get to analytical combinatorics, numerical analysis (FFT comes to mind), ML theory and design, etc., you encounter a huge amount of analysis on continuous mathematics.

Nobody claims that CS uses more math than anything else, but they do claim that it is *closer* than many other disciplines, which is obvious once you get to grad school. (Non-brainlet) CS departments are offshoots of mathematics departments, sometimes in collaboration with the EE departments. Given the way CS theory is being used to study lots of natural science these days, it's not even a ballsy claim.