Rust is slower than LuaJIT

MATLAB still hanging in there

Lua with just-in-time compilation
basically it's compiled on the spot to bytecode and then it runs.

source of these numbers?

Muh python

Attached: 7FC15082-B94B-4AE6-B685-A7D91D4E9A99.jpg (694x521, 52K)

is r really slower than python?

Attached: 1523251000740.jpg (600x480, 50K)

Yup. Most serious R libraries are C modules because of that.

>Python is slower than Matlab
>AHAHAHAHAHAHAAH

FTFY

>Python

Attached: 1367103809458.gif (355x201, 2.99M)

I have my doubts about the method here, because Fortran isn't just as fast as C in this list.

julialang.org/benchmarks/

>basically it's compiled on the spot to bytecode and then it runs.
what you said is a regular interpreter you probably meant compiled to machine code.

Yes, it is. By a statistically insignificant amount, even if the values were logarithmic, and very close to C's geomean.

Fortran is usually noticeably slower than C except in specific cases (involving matrix operations in particular). I am wondering how spot 2 is higher than spot 3, and thus also question the methodology, but that part at least is consistent.

...

Why is COBOL not there?

You sure though? From what I understand, it's as compiled as .net / mono gets.

CRINGE
BASED
BASED
BASED
BASED
BASED
BASED
BASED
CRINGE
CRINGE
BASED
CRINGE
CRINGE
BASED
BASED
CRINGE
BASED
CRINGE
BASED
CRINGE
BASED
BASED
BASED
CRINGE
CRINGE
CRINGE

>Octave
>338
>MATLAB
>9.5

FREETARDS BTFO

>Javascript almost 5x slower than C
Time to drag the "Electron is the future of desktop apps" fags in here and make them answer to this.

octave is such a dead project
I think they basically caved in to python and, to a much lesser extent, julia

MATLAB itself isn't even a good language to copy, it's just that it has tons of great toolboxes but you don't get those with Octave anyways

uhmm why is python almost 4 times slower than javascript

It isn't. Did you guys even read the numbers?

Jupyter or Sage is probably what you want if you don't want R.
Why would you use matlab derived syntax that is shittier than even R?

lmfao i like how no one even asks what these numbers mean

That is actually normal? Even pypy isn't too great.

>using anything other than c
Peasants

Using Rust makes me feel like a real woman.

I had no idea Javascript was this fast ?
Is that because of the new work google did on the intepreter ?

I assumed it wasn't retarded and was put in order.

>It combines a high-speed interpreter, written in assembler, with a state-of-the-art JIT compiler.
luajit.org/luajit.html

I guess it's because it's less work to develop with ?
I am saying this as someone who has only a beginners understanding of javascript and no other programming languages besides absolute basic stuff.

To be fair, LuaJIT is really fast.

It's number 4 after C, LuaJIT, and Rust, but way ahead of Go or Java. It's definitely doing very well in this benchmark, and it's easier to write predictably fast code in big codebases than in LuaJIT which relies on a tracing JIT.

As for why it hasn't taken over the world yet, it just hit 1.0 this summer.

It’s also no longer actively developed and is partially incompatible with the current version of Lua, a problem that is likely to get worse with every update that useless language gets.

Also, its library ecosystem is fucked.

It is roughly as fast as C for most of the benchmarks that this is averaging over. It just does really poorly in the two benchmarks which include more string handling and I/O.

Julia libraries are hit-or-miss, either best-in-class/about to become so, or shit/nonexistant with not a lot in between.

The good libraries are enough to draw me to it though. For example, if you do anything that involves ODE's, then Julia's libraries are amazing.

I personally use it for Yao.jl since part of my research is in quantum algorithms, and Julia probably became the single best option for simulating them over the last year.

I tried to get into Julia, I really did, but for a language that’s often toted as a python killer, its ability to work with common office files or do webscraping is pretty terrible.

I guess they never miss, huh?

>68989958

Attached: 1544442358662.webm (640x360, 1.42M)

THEY use samplest non optimising C compiler
and only algoritmas without
>next generation proprietary memory allocation techniques.

You misunderstand. JIT means your program gets converted to machine code as it runs. LuaJIT more or less works like this:
>convert Lua source code to bytecode when it's loaded
>start executing bytecode using interpreter
>compile anything that gets executed multiple times
>replace bytecode with a call to the compiled machine code once it's available

That's a man

Where is Haskell?

It's a girl that looks like a man cosplaying a man that looks like a girl

you're welcome user

Attached: lars-bak-computer-programmer-9429892b-a94e-4968-8116-06481f18b3d-resize-750.jpg (253x275, 15K)

and was all written by one wacky pajeet? how does it do that...

It's just a new language alright? It hasn't matured, it's still in beta, it only came out 3 years ago, give it time.

Attached: rust_brainlet.jpg (709x821, 146K)

everyone and their grandma knows Python is as slow as a Python.
it's used for pseudocode and prototyping.

which are buzzwords for "RAII and bounds checking" and "just put everything in sequentially-consistent mode hahaha pro concurrency :DDD"

LuaJIT is blazing fucking fast and Rust is only 10% slower in total than C (but is much more powerful and has safety guarantees)

It need to check CoC compliance in runtime

>higher performance on women's computers to account for inequality

Lua has safety guarantees too :^)
Why would I ever use an unused language like Rust over Lua?

Python is meant to deliver code fast, not to run code fast.
But it would be interesting to measure how fast you can write a script in python vs other languages.

Have you ever seen Lua? It's a severely limited language. Not only that, but you'd have to ship around the runtime to anybody that wants to run your scripts.

what the fuck is octave?

Open source version of MATLAB.

>implying you don't need to ship runtimes in other languages
>implying Lua has any limits

>Go is significantly faster than Fortran
Somehow I'm not entirely convinced that this benchmark is reliable.

but why it's so slow?
shouldn't math oriented language be faster than normal?

What is this cockery

10 dimensional differential equations in the presence of non free-form flux, etc...

>Lua has safety guarantees too :^)
You'd use a monster truck in place of a backhoe because they both have big tires
The ownership model and borrow checker go a lot further than Lua's runtime checks
I would kill to have their thread safety guarantees in a compile-as-you-type language

Guys should I learn python

I prefer rattlesnake

Not at all, R is slow as fuck too. That’s why Julia devs harp so much about the speed of their language, everything else that’s used for scientific computing is incredibly slow.

Pretty much this.
Currently cpu time is cheaper than dev's work hours, for large amount of business solutions Python is ok (maybe + numpy/scipy). You rarely really need something blazingly fast. Sure, it shits itself if you try to write e.g. physics simulator, but how many companies are really making those? Or can even affort to make one (good C++ devs ain't fucking cheap and cheap ones will only give you a buggy barely working piece of shit deserving to br thrown into bin)?

E R E C T

There are many cases where a high performance code helps, like server stuff.
But there are many, many, many cases where you need a "run once in a red moon" code that would be an absolute waste of time using C.

>there are languages slower than python
W H E W