views:

440

answers:

14

Hi-

I'm trying to figure out the best programming language for an analytical model I'm building. Primary consideration is speed at which it will run FOR loops.

Some detail:

  • The model needs to perform numerous (~30 per entry, over 12 cycles) operations on a set of elements from an array -- there are ~300k rows, and ~150 columns in the array. Most of these operations are logical in nature, e.g., if place(i) = 1, then j(i) = 2.
  • I've built an earlier version of this model using Octave -- to run it takes ~55 hours on an Amazon EC2 m2.xlarge instance (and it uses ~10 GB of memory, but I'm perfectly happy to throw more memory at it). Octave/Matlab won't do elementwise logical operations, so a large number of for loops are needed -- I'm relatively certain that I've vectorized as much as possible -- the loops that remain are necessary. I've gotten octave-multicore to work with this code, which makes some improvement (~30% speed reduction when I get it running on 8 EC2 cores), but ends up being unstable with file locking, etc. +I'm really looking for a step change in run-time -- I know that actually using Matlab might get me as much as a 50% improvement from looking at some benchmarks, but that is cost-prohibitive. The original plan when starting this was to actually run a Monte Carlo with this, but at 55 hours a run, that's completely impractical.
  • The next version of this is going to be a complete rebuild from the ground up (for IP reasons I won't get in to if nothing else), so I'm completely open to any programming language. I'm most familiar with Octave/Matlab, but have dabbled in R, C, C++, Java. I'm also proficient w/ SQL if the solution involves storing the data in a database. I'll learn any language for this -- these aren't complicated functionality we're looking for, no interfacing with other programs, etc., so not too concerned about learning curve.

So with all that said, what's the fastest programming language specifically for FOR loops? From a search of SO and Google, Fortran and C bubble to the top, but looking for some more advice before diving in to one or the other.

Thanks!

+1  A: 

In terms of absolute speed, probably Fortran followed by C, followed by C++. In practical application, well written code in any of the three, compiled with a descent compiler should be plenty fast.

Edit- generally you are going to see much better performance with any kind of looped or forking (e.g.- if statements) code with a compiled language, versus an interpreted language.

To give an example, on a recent project I was working on, the data sizes and operations were about 3/4 the size of what you're talking about here, but like your code, had very little room for vectorization, and required significant looping. After converting the code from matlab to C++, runtimes went from 16-18 hours, down to around 25 minutes.

MarkD
I can't think of any reason that a for loop would be different in C vs. c++. You can make it slower if you operate on some overly complicated data type in the C++, but there isn't any obvious reason to do so. Just don't go out of your way to virtual methods in the for loop that could impose a very tiny vtable overhead, and I'd expect a sanely written C++ version to compile to an identical binary as the C version for this type of work.That said, the question is wrong. for loops are an artifact of the syntax of a specific language, and the performance of them can't usefully be compared.
wrosecrans
The difference between C and C++ is not so much due to the language itself as it is due to how people use the language. As a simple example, if you are growing a vector inside a loop in C++, you may not realize how expensive of an operation that is, but if you are writing in C and allocating the memory yourself, it is a lot easier to see.
Larry Wang
A: 

How is the data stored? Your execution time is probably more effected by I/O (especially disk or worse, network) than by your language.

Assuming operations on rows are orthogonal, I would go with C# and use PLINQ to exploit all the parallelism I could.

fatcat1111
Since the OP is talking about octave, I am guessing he is using the normal matrix/array storage that octave uses, which is typically stored as an array in memory. In this case, I would guess that the language IS what is causing the slowdown, as matlab and octave are notorious for being extremely slow when it comes to looping.
MarkD
Data is all loaded from disk at the start (comprises ~10 minutes of the 55 hour runtime) and currently resides in memory for the entirety of the runtime. I believe that disk I/O time was the reason I didn't see bigger improvements with octave-multicore (which relies on writing job files to a common directory for a set of slave processes that monitor that directory, then recombining them)
Noah
@MarkD, thanks I didn't know that about Octave's disk usage.
fatcat1111
+6  A: 

This for loop looks no more complex than this when it hits the CPU:

for(int i = 0; i != 1024; i++) translates to

mov r0, 0           ;;start the counter    
top:

;;some processing

add r0, r0, 1       ;;increment the counter by 1
jne top: r0, 1024   ;;jump to the loop top if we havn't hit the top of the for loop (1024 elements)

;;continue on

As you can tell, this is sufficiently simple you can't really optimize it very well[1]... Refocus towards the algorithm level.

The first cut at the problem is to look at cache locality. Look up the classic example of matrix multiplication and swapping the i and j indexes.

edit: As a second cut, I would suggest evaluating the algorithm for data-dependencies between iterations and data dependency between localities in your 'matrix' of data. It may be a good candidate for parallelization.

[1] There are some micro-optimizations possible, but those will not produce the speedsups you're looking for.

Paul Nathan
Sure, if the OP is using octave or matlab, and can change his algorithm to vectorize everything, he can see a huge increase in speed. If, however, he truly can't vectorize any more (as stated in the original post), and must rely on loops, the algorithm probably is not well suited to matlab, and moving to a compiled language can yield order-of-magnitude increases in speed.
MarkD
+3  A: 

For what you're discussing, Fortran is probably your first choice. The closest second place is probably C++. Some C++ libraries use "expression templates" to gain some speed over C for this kind of task. It's not entirely certain those will do you any good, but C++ can be at least as fast as C, and possibly somewhat faster.

At least in theory, there's no reason Ada couldn't be competitive as well, but it's been so long since I used it for anything like this that I hesitate to recommend it -- not because it isn't good, but because I just haven't kept track of current Ada compilers well enough to comment on them intelligently.

Jerry Coffin
+5  A: 

~300k * ~150 * ~30 * ~12 = ~16G iterations, right? This number of primitive operations should complete in a matter of minutes (if not seconds) in any compiled language on any decent CPU. Fortran, C/C++ should do it almost equally well. Even managed languages, such as Java and C#, would only fall behind by a small margin (if at all).

If you have a problem of ~16G iterations running 55 hours, this means that they are very far from being primitive (80k per second? this is ridiculous), so maybe we should know more. (as was already suggested, is disk access limiting performance? is it network access?)

Rotsor
Well, Octave isn't a compiled language, so not sure if that changes the time perspective. No disk / network access, as Octave stores all of the data in memory. There are sort and find operations in the time consuming loops -- e.g., sort a list of 10 elements, then find a given elements position in the sorted array (this could be done with a rank function, but Octave/Matlab doesn't have a straightforward rank function that I've been able to find). Not sure if that qualifies as non-primitive...
Noah
Of course sorting of 10 elements is not exactly primitive operation (more like ~40 operations), but it still does not explain the performance bottleneck for me. I never used Octave and don't know anything about its performance, but knowing what you told us, I believe that any compiled language can do orders or magnitude better. You should just select the one which is more convenient (which you know better). By the way, it's not `for` who will be limiting your performance in a compiled language, it's sorting and memory access.
Rotsor
@Noah FYI, in Matlab you can get the rank order from the second return argument of the sort() function.
A: 

Might you not be best with a hand-coded assembler insert? Assuming, of course, that you don't need portability.

That and an optimized algorithm should help (and perhaps restructuring the data?).

You might also want to try multiple algorithms and profile them.

LeonixSolutions
A: 

what about a lazy loading language like clojure. it is a lisp so like most lisp dialects lacks a for loop but has many other forms that operate more idiomatically for list processing. It might help your scaling issues as well because the operations are thread safe and because the language is functional it has fewer side effects. If you wanted to find all the items in the list that were 'i' values to operate on them you might do something like this.

(def mylist ["i" "j" "i" "i" "j" "i"])
(map #(= "i" %) mylist)

result

(true false true true false true)

Jed Schneider
Isn't it the worst language for the OP purposes? Doesn't have loops at all, and their alternatives have quite an overhead.
Rotsor
+3  A: 

Any compiled language should perform the loop itself on roughly equal terms.

If you can formulate your problem in its terms, you might want to look at CUDA or OpenCL and run your matrix code on the GPU - though this is less good for code with lots of conditionals.

If you want to stay on conventional CPUs, you may be able to formulate your problem in terms of SSE scatter/gather and bitmask operations.

Pete Kirkham
A: 

APL.

Even though it's interpreted, its primitive operators all operate on arrays natively, therefore you rarely need any explicit loops. You write the same code, whether the data is scalar or array, and the interpreter takes care of any looping needed internally, and thus with the minimum overhead - the loops themselves are in a compiled language, and will have been heavily optimised for the specific architecture of the CPU it's running on.

Here's an example of the simplicity of array handling in APL:

    A <- 2 3 4 5 6 8 10
    ((2|A)/A) <- 0
    A
2 0 4 0 6 8 10

The first line sets A to a vector of numbers. The second line replaces all the odd numbers in the vector with zeroes. The third line queries the new values of A, and the fourth line is the resulting output.

Note that no explicit looping was required, as scalar operators such as '|' (remainder) automatically extend to arrays as required. APL also has built-in primitives for searching and sorting, which will probably be faster than writing your own loops for these operations.

Wikipedia has a good article on APL, which also provides links to suppliers such as IBM and Dyalog.

Dave Gordon
+5  A: 

As @Rotsor said, 16G operations / 55 hours is about 80,000 operations per second, or one operation every 12.5 microseconds. That's a lot of time per operation.

That means your loops are not the cause of poor performance, it's what's in the innermost loop that's taking the time. And Octave is an interpreted language. That alone means an order of magnitude slowdown.

If you want speed, you first need to be in a compiled language. Then you need to do performance tuning (aka profiling) or, just single step it in a debugger at the instruction level. That will tell you what it is actually doing in its heart of hearts. Once you've got it to where it's not wasting cycles, fancier hardware, cores, CUDA, etc. will give you a parallelism speedup. But it's silly to do that if your code is taking unnecessarily many cycles. (Here's an example of performance tuning - a 43x speedup just by trimming the fat.)

I can't believe the number of responders talking about matlab, APL, and other vectorized languages. Those are interpreters. They give you concise source code, which is not at all the same thing as fast execution. When it comes down to the bare metal, they are stuck with the same hardware as every other language.


Added: to show you what I mean, I just ran this C++ code, which does 16G operations, on this moldy old laptop, and it took 94 seconds, or about 6ns per iteration. (I can't believe you baby-sat that thing for 2 whole days.)

void doit(){
  double sum = 0;
  for (int i = 0; i < 1000; i++){
    for (int j = 0; j < 16000000; j++){
      sum += j * 3.1415926;
    }
  }
}
Mike Dunlavey
Agreed - 12.5 microseconds is a LONG time on modern processors. I've benchmarked my own laptop (Lenovo W500) as taking 5 *nano*seconds per loop iteration (specifically a C# foreach loop initializing values in a large 1 dimensional array of int).
Bevan
@Bevan: Yeah, that's comparable to what I got.
Mike Dunlavey
+2  A: 

Probably the assembly language for whatever your platform is. But compilers (especially special-purpose ones that specifically target a single platform (e.g., Analog Devices or TI DSPs)) are often as good as or better than humans. Also, compilers often know about tricks that you don't. For example, the aforementioned DSPs support zero-overhead loops and the compiler will know how to optimize code to use those loops.

Chris
A: 

Any modern compiled or JITted language is going to render down to pretty much the same machine language code, giving a loop overhead of 10 nano seconds or less, per iteration, on modern processors.

Quoting @Rotsor:

If you have a problem of ~16G iterations running 55 hours, this means that they are very far from being primitive (80k per second? this is ridiculous), so maybe we should know more.

80k operations per second is around 12.5 microseconds each - a factor of 1000 greater than the loop overhead you'd expect.

Assuming your 55 hour runtime is single threaded, and if your per item operations are as simple as suggested, you should be able to (conservatively) achieve a speedup of 100x and cut it down to under an hour very easily.

If you want to run faster still, you'll want to look at writing multi-threaded solution, in which case a language that provides good support would be essential. I'd lean towards PLINQ and C# 4.0, but that's because I already know C# - YMMV.

Bevan
+1  A: 

Matlab will do element-wise logical operations and they are generally quite fast.

Here is a quick example on my computer (AMD Athalon 2.3GHz w/3GB) :

d=rand(300000,150);
d=floor(d*10);

>> numel(d(d==1))
ans =
     4501524

>> tic;d(d==1)=10;toc;
Elapsed time is 0.754711 seconds.

>> numel(d(d==1))
ans =
     0
>> numel(d(d==10))
ans =
     4501524

In general I've found matlab's operators are very speedy, you just have to find ways to express your algorithms directly in terms of matrix operators.

A: 

C++ is not fast when doing matrixy things with for loops. C is, in fact, specifically bad at it. See good math bad math.

I hear that C99 has __restrict pointers that help, but don't know much about it.

Fortran is still the goto language for numerical computing.

"Goto statement considered harmful" -- Edsger Dijkstra, 1968
Jason S
My typo. I wrote "goto" instead of "go to". This is an issue with pointers and how they eviscerate optimization opportunities, not goto statements.