Consider something like...
for (int i = 0; i < test.size(); ++i) {
test[i].foo();
test[i].bar();
}
Now consider..
for (int i = 0; i < test.size(); ++i) {
test[i].foo();
}
for (int i = 0; i < test.size(); ++i) {
test[i].bar();
}
Is there a large difference in time spent between these two? I.e. what is the cost of the actual iteration? It seems like the only real operations you are repeating are an increment and a comparison (though I suppose this would become significant for a very large n). Am I missing something?
First, as noted above, if your compiler can't optimize the size() method out so it's just called once, or is nothing more than a single read (no function call overhead), then it will hurt.
There is a second effect you may want to be concerned with, though. If your container size is large enough, then the first case will perform faster. This is because, when it gets to test[i].bar(), test[i] will be cached. The second case, with split loops, will thrash the cache, since test[i] will always need to be reloaded from main memory for each function.
Worse, if your container (std::vector, I'm guessing) has so many items that it won't all fit in memory, and some of it has to live in swap on your disk, then the difference will be huge as you have to load things in from disk twice.
However, there is one final thing that you have to consider: all this only makes a difference if there is no order dependency between the function calls (really, between different objects in the container). Because, if you work it out, the first case does:
test[0].foo();
test[0].bar();
test[1].foo();
test[1].bar();
test[2].foo();
test[2].bar();
// ...
test[test.size()-1].foo();
test[test.size()-1].bar();
while the second does:
test[0].foo();
test[1].foo();
test[2].foo();
// ...
test[test.size()-1].foo();
test[0].bar();
test[1].bar();
test[2].bar();
// ...
test[test.size()-1].bar();
So if your bar() assumes that all foo()'s have run, you will break it if you change the second case to the first. Likewise, if bar() assumes that foo() has not been run on later objects, then moving from the second case to the first will break your code.
So be careful and document what you do.
There are many aspects in such comparison.
First, complexity for both options is O(n), so difference isn't very big anyway. I mean, you must not care about it if you write quite big and complex program with a large n and "heavy" operations .foo() and bar(). So, you must care about it only in case of very small simple programs (this is kind of programs for embedded devices, for example).
Second, it will depend on programming language and compiler. I'm assured that, for instance, most of C++ compilers will optimize your second option to produce same code as for the first one.
Third, if compiler haven't optimized your code, performance difference will heavily depend on the target processor. Consider loop in a term of assembly commands - it will look something like this (pseudo assembly language):
LABEL L1:
do this ;; some commands
call that
IF condition
goto L1
;; some more instructions, ELSE part
I.e. every loop passage is just IF statement. But modern processors don't like IF. This is because processors may rearrange instructions to execute them beforehand or just to avoid idles. With the IF (in fact, conditional goto or jump) instructions, processors do not know if they may rearrange operation or not.
There's also a mechanism called branch predictor. From material of Wikipedia:
branch predictor is a digital circuit that tries to guess which way a branch (e.g. an if-then-else structure) will go before this is known for sure.
This "soften" effect of IF's, through if the predictor's guess is wrong, no optimization will be performed.
So, you can see that there's a big amount of conditions for both your options: target language and compiler, target machine, it's processor and branch predictor. This all makes very complex system, and you cannot foresee what exact result you will get. I believe, that if you don't deal with embedded systems or something like that, the best solution is just to use the form which your are more comfortable with.
For your examples you have the additional concern of how expensive .size() is, since it's compared for each time i increments in most languages.
How expensive is it? Well that depends, it's certainly all relative. If .foo() and .bar() are expensive, the cost of the actual iteration is probably minuscule in comparison. If they're pretty lightweight, then it'll be a larger percentage of your execution time. If you want to know about a particular case test it, this is the only way to be sure about your specific scenario.
Personally, I'd go with the single iteration to be on the cheap side for sure (unless you need the .foo() calls to happen before the .bar() calls).
I assume .size() will be constant. Otherwise, the first code example might not give the same as the second one.
Most compilers would probably store .size() in a variable before the loop starts, so the .size() time will be cut down.
Therefore the time of the stuff inside the two for loops will be the same, but the other part will be twice as much.
Performance tag, right.
As long as you are concentrating on the "cost" of this or that minor code segment, you are oblivious to the bigger picture (isolation); and your intention is to justify something that, at a higher level (outside your isolated context), is simply bad practice, and breaks guidelines. The question is too low level and therefore too isolated. A system or program which is set of integrated components will perform much better that a collection of isolated components.
The fact that this or that isolated component (work inside the loop) is fast or faster is irrelevant when the loop itself is repeated unnecessarily, and which would therefore take twice the time.
Given that you have one family car (CPU), why on Earth would you:
sit at home and send your wife out to do her shopping
wait until she returns
take the car, go out and do your shopping
leaving her to wait until you return
If it needs to be stated, you would spend (a) almost half of your hard-earned resources executing one trip and shopping at the same time and (b) have those resources available to have fun together when you get home.
It has nothing to do with the price of petrol at 9:00 on a Saturday, or the time it takes to grind coffee at the café, or cost of each iteration.
Yes, there is a large diff in the time and the resources used. But the cost is not merely in the overhead per iteration; it is in the overall cost of the one organised trip vs the two serial trips.
Performance is about architecture; never doing anything twice (that you can do once), which are the higher levels of organisation; integrated of the parts that make up the whole. It is not about counting pennies at the bowser or cycles per iteration; those are lower orders of organisation; which ajust a collection of fragmented parts (not a systemic whole).
Masseratis cannot get through traffic jams any faster than station wagons.
Related
The following question is more related to design, rather than actual coding. I don't know if there's a technical term for such problem, so I'll proceed with an example.
I have some openCL code not optimized at all, and in the Kernel there's essentially a switch statement similar to the following
switch(const) {
case const_a : do_something_a(...); break;
case const_b : do_something_b(....); break;
... //etc
}
I cannot write the actual statement since is quite long. As a simple example consider the following switch statement:
int a;
switch(input):
case 13 : {a = 3; break;}
case 1 : {a = 7; break;}
case 23 : {a = 1; break;}
default : {...}
The question is... would it be better to change such switch with an expression like
a = (input == 13)*3 + (input == 1)*7 + (input == 23)
?
If it's not, is it possible to make it more efficient anyway?
You can assume input only takes values in the set of cases of the switch statement.
You've discovered an interesting question that GPU compilers wrestle with. The general advice is try not to branch. Tricks to make that possible are splitting kernels up (as suggested above) and preprocessor (program-time definitions). Research in GPU algorithm development basically works from this axiom.
Branching all over the place won't get great efficiency because of the inherent divergence (channel = work item within the SIMD thread/warp). Remember that all these channels must execute together. So in a switch where all are taking different paths everyone else goes along for the ride silently waiting for their "case" to execute. Now, if input is always always the same value, it can still be a win.
Another popular option is a table indirection.
kernel void foo(const int *t, ...)
...
a = tbl[input];
This case has a few problems too depending on hardware, inputs, and problem size.
Without more specific context, I can conjure up a case where any of these can run well or poorly.
Switching (or big if-then-else chains).
PROS: If all work items generally take the same path (input is mostly the same value), it's going to be efficient. You could also write an if-then-else chain putting the most common cases first. (On GPUs a switch is not necessarily as easy as an indirect jump since there are multiple work items and they may take different paths.)
CONS: Might generate lots of program code and could blow out the instruction cache. Branching all over the place can get a little costly depending on how many cases need to be evaluated. It might just be better to grind through the compute with the predicated code.
Predicated Code (Your (input == 13)*3 ... code).
PROS: This will probably generate smaller programs and stress the I$ less. (Lookup the OpenCL select function to see a more general approach for your case.)
CONS: We've basically punted and decided to evaluate every "case in the switch". If input is usually the same value, we're wasting time here.
Lookup-table based approaches (my example).
PROS: If the switch you are evaluating has a massive number of cases (branches), but can be indexed by integer you might be ahead to just use a lookup table. On some hardware this means a read from global memory (far far away). Other architectures have a dedicated constant cache, but I understand that a vector lookup will serialize (K cycles for each channel). So it might be only marginally better than the global memory table. However, the code table-lookup generated will be short (I$ friendly) and as the number of branches (case statements) grow this will win in the limit. This approach also deals well with uniform/scattered distributions of input's value.
CONS: The read from global memory (or serialized access from the constant cache) has a big latency even compared to branching. In some cases, to eliminate the extra memory traffic I've seen compilers convert lookup tables into if-then-else/switch chains. It's rare that we have 100 element case statements.
I am now inspired to go study this cutoff. :-)
This is a question about performance of code written in Scala.
Consider the following two code snippets, assume that x is some collection containing ~50 million elements:
def process(x: Traversable[T]) = {
processFirst x.head
x reduce processPair
processLast x.last
}
Versus something like this (assume for now we have some way to determine if we're operating on the first element versus the last element):
def isFirstElement[T](x: T) = ???
def isLastElement[T](x: T) = ???
def process(x: Traversable[T]) = {
x reduce {
(left, right) =>
if (isFirstElement(left)
processFirst(left)
else if (isLastElement(right))
processLast(right)
processPair(left, right)
}
}
Which approach is faster? and for ~50 million elements, how much faster?
It seems to me that the first example would be faster because there are fewer conditional checks occurring for all but the first and last elements. However for the latter example there is some argument to suggest that the JIT might be clever enough to optimize away those additional head/last conditional checks that would otherwise occur for all but the first/last elements.
Is the JIT clever enough to perform such operations? The obvious advantage of the latter approach is that all business can be placed in the same function body while in the latter case business must be partitioned into three separate function bodies invoked separately.
** EDIT **
Thanks for all the great responses. While I am leaving the second code snippet above to illustrate its incorrectness, I want to revise the first approach slightly to reflect better the problem I am attempting to solve:
// x is some iterator
def process(x: Iterator[T]) = {
if (x.hasNext)
{
var previous = x.next
var current = null
processFirst previous
while(x.hasNext)
{
current = x.next
processPair(previous, current)
previous = current
}
processLast previous
}
}
While there are no additional checks occurring in the body, there is an additional reference assignment that appears to be unavoidable (previous = current). This is also a much more imperative approach that relies on nullable mutable variables. Implementing this in a functional yet high performance manner would be another exercise for another question.
How does this code snippet stack-up against the last of the two examples above? (the single-iteration block approach containing all the branches). The other thing I realize is that the latter of the two examples is also broken on collections containing fewer than two elements.
If your underlying collection has an inexpensive head and last method (not true for a generic Traversable), and the reduction operations are relatively inexpensive, then the second way takes about 10% longer (maybe a little less) than the first on my machine. (You can use a var to get first, and you can keep updating a second far with the right argument to obtain last, and then do the final operation outside of the loop.)
If you have an expensive last (i.e. you have to traverse the whole collection), then the first operation takes about 10% longer (maybe a little more).
Mostly you shouldn't worry too much about it and instead worry more about correctness. For instance, in a 2-element list your second code has a bug (because there is an else instead of a separate test). In a 1-element list, the second code never calls reduce's lambda at all, so again fails to work.
This argues that you should do it the first way unless you're sure last is really expensive in your case.
Edit: if you switch to a manual reduce-like-operation using an iterator, you might be able to shave off up to about 40% of your time compared to the expensive-last case (e.g. list). For inexpensive last, probably not so much (up to ~20%). (I get these values when operating on lengths of strings, for example.)
First of all, note that, depending on the concrete implementation of Traversable, doing something like x.last may be really expensive. Like, more expensive than all the rest of what's going on here.
Second, I doubt the cost of conditionals themselves is going to be noticeable, even on a 50 million collection, but actually figuring out whether a given element is the first or the last, might again, depending on implementation, get pricey.
Third, JIT will not be able to optimize the conditionals away: if there was a way to do that, you would have been able to write your implementation without conditionals to begin with.
Finally, if you are at a point where it starts looking like an extra if statement might affect performance, you might consider switching to java or even "C". Don't get me wrong, I love scala, it is a great language, with lots of power and useful features, but being super-fast just isn't one of them.
I really tried to find something about this kind of operations but I don't find specific information about my question... It's simple: Are boolean operations slower than typical math operations in loops?
For example, this can be seen when working with some kind of sorting. The method will make an iteration and compare X with Y... But is this slower than a summatory or substraction loop?
Example:
Boolean comparisons
for(int i=1; i<Vector.Length; i++) if(Vector[i-1] < Vector[i])
Versus summation:
Double sum = 0;
for(int i=0; i<Vector.Length; i++) sum += Vector[i];
(Talking about big length loops)
Which is faster for the processor to complete?
Do booleans require more operations in order to return "true" or "false" ?
Short version
There is no correct answer because your question is not specific enough (the two examples of code you give don't achieve the same purpose).
If your question is:
Is bool isGreater = (a > b); slower or faster than int sum = a + b;?
Then the answer would be: It's about the same unless you're very very very very very concerned about how many cycles you spend, in which case it depends on your processor and you need to read its documentation.
If your question is:
Is the first example I gave going to iterate slower or faster than the second example?
Then the answer is: It's going to depend primarily on the values the array contains, but also on the compiler, the processor, and plenty of other factors.
Longer version
On most processors a boolean operation has no reason to significantly be slower or faster than an addition: both are basic instructions, even though comparison may take two of them (subtracting, then comparing to zero). The number of cycles it takes to decode the instruction depends on the processor and might be different, but a few cycles won't make a lot of difference unless you're in a critical loop.
In the example you give though, the if condition could potentially be harmful, because of instruction pipelining. Modern processors try very hard to guess what the next bunch of instructions are going to be so they can pre-fetch them and treat them in parallel. If there is branching, the processor doesn't know if it will have to execute the then or the else part, so it guesses based on the previous times.
If the result of your condition is the same most of the time, the processor will likely guess it right and this will go well. But if the result of the condition keeps changing, then the processor won't guess correctly. When such a branch misprediction happens, it means it can just throw away the content of the pipeline and do it all over again because it just realized it was moot. That. does. hurt.
You can try it yourself: measure the time it takes to run your loop over a million elements when they are of same, increasing, decreasing, alternating, or random value.
Which leads me to the conclusion: processors have become some seriously complex beasts and there is no golden answers, just rules of thumb, so you need to measure and profile. You can read what other people did measure though to get an idea of what you should or should not do.
Have fun experimenting. :)
Now that std::experimental::optional has been accepted (or is about to be accepted), I wonder what is the overhead and the consequences on the assembly generated when the inner value is get by the following operators :
->
*
value
value_or
compared to the case without std::optional. It could be particularly important for computationaly intensive programs.
For example, what would be order of magnitude of the overhead on operations on a std::vector<std::experimental::optional<double>> compared to a std::vector<double> ?
-> and * ought to have zero overhead.
value and value_or ought to have the overhead of one branch: if(active)
Also, copy/move constructor, copy/move assignment, swap, emplace, operator==, operator<, and the destructor ought to also have the overhead of one branch.
However, one banch of overhead is so small it probably can't even be measured. Seriously, write pretty code, and don't worry about the performance here. Odds are making the code pretty will result in it running faster than if you tried to make it fast. Counter-intuitive, but do it anyway.
There are definitely cases where the overhead becomes noticible, for instance sorting a large number of optionals. In these cases, there's four situations,
(A) all the optionals known to be empty ahead of time, in which case, why sort?
(B) Some optionals may or may not be active, in which case the overhead is required and there is no better way.
(C) All optionals are known to have values ahead of time and you don't need the sorted-data in place, in which case, use the zero overhead operators to make a copy of the data where the copy is using the raw type instead of optional, and sort that.
(D) All optionals are known to have values ahead of time, but you need the sorted data in-place. In this case, optional is adding unnecessary overhead, and the easiest way to work around it is to do step C, and then use the no-overhead operators to move the data back.
Besides the other answer, you should also consider that std::optional requires additional memory.
Often it's not just an extra byte, but (at least for "small" types) a 2x space overhead due to padding .
Maybe RAM isn't a problem but that also means fewer values available in the cache.
A sentinel value, if specific knowledge allows to use it, could be a better choice (probably in the form of markable to keep type safety).
An interesting reading is: Boost optional - Performance considerations
I have a function that I use to look up a value based on an index. The value takes some time to calculate, so I want to do it with ParallelMap, and references another similar such function that returns a list of expressions, also based on an index.
However, when I set it all up in a seemingly reasonable fashion, I see some very bizarre behaviour. First, I see that the function appears to work, albeit very slowly. For large indexes, however, the processor activity in Taskmangler stays entirely at zero for an extended period of time (i.e. 2-4 minutes) where all instances of Mathematica are seemingly inert. Then, without the slightest blip of CPU use, a result appears. Is this another case of Mathematica spukhafte Fernwirkung?
That is, I want to create a variable/function that stores an expression, here a list of integers (ListOfInts), and then on the parallel workers I want to perform some function on that expression (here I apply a set of replacement rules and take the Min). I want the result of that function to also be indexed by the same index under another variable/function (IndexedFunk), whose result is then available back on the main instance of Mathematica:
(*some arbitrary rules that will convert some of the integers to negative values:*)
rulez=Dispatch[Thread[Rule[Range[222],-Range[222]]]];
maxIndex = 333;
Clear[ListOfInts]
Scan[(ListOfInts[#]=RandomInteger[{1,999},55])&,Range[maxIndex ]]
(*just for safety's sake:*)
DistributeDefinitions[rulez, ListOfInts]
Clear[IndexedFunk]
(*I believe I have to have at least one value of IndexedFunk defined before I Share the definition to the workers:*)
IndexedFunk[1]=Min[ListOfInts[1]]/.rulez
(*... and this should let me retrieve the values back on the primary instance of MMA:*)
SetSharedFunction[IndexedFunk]
(*Now, here is the mysterious part: this just sits there on my multiprocessor machine for many minutes until suddenly a result appears. If I up maxIndex to say 99999 (and of course re-execute the above code again) then the effect can more clearly be seen.*)
AbsoluteTiming[Short[ParallelMap[(IndexedFunk[#]=Min[ListOfInts[#]/.rulez])&, Range[maxIndex]]]]
I believe this is some bug, but then I am still trying to figure out Mathematica Parallel, so I can't be too confident in this conclusion. Despite its being depressingly slow, it is nonetheless impressive in its ability to perform calculations without actually requiring a CPU to do so.
I thought perhaps it was due to whatever communications protocol is being used between the master and slave processes, perhaps it is so slow that it just appears that the processors are doing nothing when if fact they are just waiting to send the next bit of some definition or other. In which case I thought ParallelMap[..., Method->"CoarsestGrained"] would be of some use. But no, that doesn't work neither.
A question: "Am I doing something obviously wrong, or is this a bug?"
I am afraid you are. The problem is with the shared definition of a variable. Mathematica maintains a single coherent value in all copies of the variable across kernels, and therefore that variable becomes a single point of huge contention. CPU is idle because kernels line up to the queue waiting for the variable IndexedFunk, and most time is spent in interprocess or inter-machine communication. Go figure.
By the way, there is no function SetSharedDefinition in any Mathematica version I know of. You probably intended to write SetSharedVariable. But remove that evil call anyway! To avoid contention, return results from the parallelized computation as a list of pairs, and then assemble them into downvalues of your variable at the main kernel:
Clear[IndexedFunk]
Scan[(IndexedFunk[#[[1]]] = #[[2]]) &,
ParallelMap[{#, Min[ListOfInts[#] /. rulez]} &, Range[maxIndex]]
]
ParallelMap takes care of distributing definition automagically, so the call to DistributeDefinitions is superfluous. (As a minor note, it is not correct as written, omitting the maxIndex variable, but the omission is automatically taken care of by ParallelMap in this particular case.)
EDIT, NB!: The automatic distribution applies only to the version 8 of Mathematica. Thanks #MikeHoneychurch for the correction.