This is more of a generic, technical question. I'm just curious about what the main factors in determining how fast or slow a computer program runs are?
For example, when I time Python code, the runtime always varies by at least +/- 0.02 seconds
There are many reasons of execution time variance. Variation of ~200ms looks plausible for a python script that runs for seconds. Main contributors here would be OS/scheduler and memory/cache. OS will serve interrupts on a core your script is running, and on blocking system calls it will run the scheduler, which will run background tasks on that core. While these tasks are running, they will pollute L1,L2 and L3 caches so that some part of data and code of that python script will be evicted to RAM. So memory references will always take different time each run, because you can never reproduce memory footprint of background tasks that interrupted your script.
If you are running on Linux, you may try scheduling your script to a CPU that was offlined from scheduler using isolcpu= kernel boot option, so you have less noise from other processes. You'll have orders of magnitude less variation then, but there will be still some coming from using shared resources - memory controllers, IO buses, shared last level cache.
Related
I have a fortran code that I am using to calculate some quantities related to the work that I do. The code itself involves several nested loops, and requires very little disk I/O. Whenever the code is modified, I run it against a suite of several input files (just to make sure it's working properly).
To make a long story short, the most recent update has increased the run time of the program by about a factor of four, and running each input file serially with one CPU takes about 45 minutes (a long time to wait, just to see whether anything was broken). Consequently, I'd like to run each of the input files in parallel across the 4 cpus on the system. I've been attempting to implement the parallelism via a bash script.
The interesting thing I have noted is that, when only one instance of the program is running on the machine, it takes about three and a half minutes to crank through one of the input files. When four instances of the program are running, it takes more like eleven and a half minute to crank through one input file (bringing my total run time down from about 45 minutes to 36 minutes - an improvement, yes, but not quite what I had hoped for).
I've tried implementing the parallelism using gnu parallel, xargs, wait, and even just starting four instances of the program in the background from the command line. Regardless of how the instances are started, I see the same slow down. Consequently, I'm pretty sure this isn't an artifact of the shell scripting, but something going on with the program itself.
I have tried rebuilding the program with debugging symbols turned off, and also using static linking. Neither of these had any noticeable impact. I'm currently building the program with the following options:
$ gfortran -Wall -g -O3 -fbacktrace -ffpe-trap=invalid,zero,overflow,underflow,denormal -fbounds-check -finit-real=nan -finit-integer=nan -o [program name] {sources}
Any help or guidance would be much appreciated!
On modern CPUs you cannot expect a linear speedup. There are several reasons:
Hyperthreading GNU/Linux will see hyperthreading as a core eventhough it is not a real core. It is more like 30% of a core.
Shared caches If your cores share the same cache and a single instance of your program uses the full shared cache, then you will get more cache misses if you run more instances.
Memory bandwidth A similar case as the shared cache is the shared memory bandwidth. If a single thread uses the full memory bandwidth, then running more jobs in parallel may congest the bandwidth. This can partly be solved by running on a NUMA where each CPU has some RAM that is "closer" than other RAM.
Turbo mode Many CPUs can run a single thread at a higher clock rate than multiple threads. This is due to heat.
All of these will exhibit the same symptom: Running a single thread will be faster than each of the multiple threads, but the total throughput of the multiple threads will be bigger than the single thread.
Though I must admit your case sounds extreme: With 4 cores I would have expected a speedup of at least 2.
How to identify the reason
Hyperthreading Use taskset to select which cores to run on. If you use 2 of the 4 cores is there any difference if you use #1+2 or #1+3?
Turbo mode Use cpufreq-set to force a low frequency. Is the speed now the same if you run 1 or 2 jobs in parallel?
Shared cache Not sure how to do this, but if it is somehow possible to disable the cache, then comparing 1 job to 2 jobs run at the same low frequency should give an indication.
I'm new to Elixir, and I'm starting to read through Dave Thomas's excellent Programming Elixir. I was curious how far I could take the concurrency of the "pmap" function, so I iteratively boosted the number of items to square from 1,000 to 10,000,000. Out of curiosity, I watched the output of htop as I did so, usually peaking out with CPU usage similar to that shown below:
After showing the example in the book, Dave says:
And, yes, I just kicked off 1,000 background processes, and I used all the cores and processors on my machine.
My question is, how come on my machine only cores 1, 3, 5, and 7 are lighting up? My guess would be that it has to do with my iex process being only a single OS-level process and OSX is managing the reach of that process. Is that what's going on here? Is there some way to ensure all cores get utilized for performance-intensive tasks?
Great comment by #Thiago Silveira about first line of iex's output. The part [smp:8:8] says how many operating system level processes is Erlang using. You can control this with flag --smp if you want to disable it:
iex --erl '-smp disable'
This will ensure that you have only one system process. You can achieve similar result by leaving symmetric multiprocessing enabled, but setting directly NumberOfShcedulers:NumberOfSchedulersOnline.
iex --erl '+S 1:1'
Each operating system process needs to have its own scheduler for Erlang processes, so you can easily see how many of them do you have currently:
:erlang.system_info(:schedulers_online)
To answer your question about performance. If your processors are not working at full capacity (100%) and non of them is doing nothing (0%) then it is probable that making the load more evenly distributed will not speed things up. Why?
The CPU usage is measured by probing the processor state at many points in time. This states are either "working" or "idle". 82% CPU usage means that you can perform couple of more tasks on this CPU without slowing other tasks.
Erlang schedulers try to be smart and not migrate Erlang processes between cores unless they have to because it requires copying. The migration occurs for example when one of schedulers is idle. It can then borrow a process from others scheduler run queue.
Next thing that may cause such a big discrepancy between odd and even cores is Hyper Threading. On my dual core processor htop shows 4 logical cores. In your case you probably have 4 physical cores and 8 logical because of HT. It might be the case that you are utilizing your physical cores with 100%.
Another thing: pmap needs to calculate result in separate process, but at the end it sends it to the caller which may be a bottleneck. The more you send messages the less CPU utilization you can achieve. You can try for fun giving the processes a task that is really CPU intensive like calculating Ackerman function. You can even calculate how much of your job is the sequential part and how much is parallel using Amdahl's law and measuring execution times for different number of cores.
To sum up: the CPU utilization from screenshot looks really great! You don't have to change anything for more performance-intensive tasks.
Concurrency is not Parallelism
In order to get good parallel performance out of Elixir/BEAM coding you need to have some understanding of how the BEAM scheduler works.
This is a very simplistic model, but the BEAM scheduler gives each process 2000 reductions before it swaps out the process for the next process. Reductions can be thought of as function calls. By default a process runs on the core/scheduler that spawned it. Processes only get moved between schedulers if the queue of outstanding processes builds up on a given scheduler. By default the BEAM runs a scheduling thread on each available core.
What this implies is that in order to get the most use of the processors you need to break up your tasks into large enough pieces of work that will exceed the standard "reduction" slice of work. In general, pmap style parallelism only gives significant speedup when you chunk many items into a single task.
The other thing to be aware of is that some parts of the BEAM use a spin/wait loop when awaiting work and that can skew usage when you use
a tool like htop to examine CPU usage. You'll get a much better understanding of your program's performance by using :observer.
I am working on a calculation intensive C# project that implements several algorithms. The problem is that when I want to profile my application, the time it takes for a particular algorithm varies. For example sometimes running the algorithm 100 times takes about 1100 ms and another time running 100 times takes much more time like 2000 or even 3000 ms. It may vary even in the same run. So it is impossible to measure improvement when I optimize a piece of code. It's just unreliable.
Here is another run:
So basically I want to make sure one CPU is dedicated to my app. The PC has an old dual core Intel E5300 CPU running on Windows 7 32 bit. So I can't just set process affinity and forget about one core forever. It would make the computer very slow for daily tasks. I need other apps to use a specific core when I desire and the when I'm done profiling, the CPU affinities come back to normal. Having a bat file to do the task would be a fantastic solution.
My question is: Is it possible to have a bat file to set process affinity for every process on windows 7?
PS: The algorithm is correct and every time runs the same code path. I created some object pool so after first run, zero memory is allocated. I also profiled memory allocation with dottrace and it showed no allocation after first run. So I don't believe GC is triggered when the algorithm is working. Physical memory is available and system is not running low on RAM.
Result: The answer by Chris Becke does the job and sets process affinities exactly as intended. It resulted in more uniform results specially when background apps like visual studio and dottrace are running. Further investigation into the divergent execution time revealed that the root for the unpredictability is CPU overheat. The CPU overheat alarm was off while the temperature was over 100C! So after fixing the malfunctioning fan, the results became completely uniform.
You mean SetProcessAffinityMask?
I see this question, while tagged windows, is c#, so... I see the System.Diagnostics.Process object has a ThreadAffinity member that should perform the same function.
I am just not sure that this will stabilize the CPU times quite in the way you expect. A single busy task that is not doing IO should remain scheduled on the same core anyway unless another thread interrupts it, so I think your variable times are more due to other threads / processes interrupting your algorithm than the OS randomly shunting your thread to a different core - so unless you set the affinity for all other threads in the system to exclude your preferred core I can't see this helping.
Let us say we have a fictitious single core CPU with Program Counter and basic instruction set such as Load, Store, Compare, Branch, Add, Mul and some ROM and RAM. Upon switching on it executes a program from ROM.
Would it be fair to say the work the CPU does is based on the type of instruction it's executing. For example, a MUL operating would likely involve more transistors firing up than say Branch.
However from an outside perspective if the clock speed remains constant then surely the CPU could be said to be running at 100% constantly.
How exactly do we establish a paradigm for measuring the work of the CPU? Is there some kind of standard metric perhaps based on the type of instructions executing, the power consumption of the CPU, number of clock cycles to complete or even whether it's accessing RAM or ROM.
A related second question is what does it mean for the program to "stop". Usually does it just branch in an infinite loop or does the PC halt and the CPU waits for an interupt?
First of all, that a CPU is always executing some code is just an approximation these days. Computer systems have so-called sleep states which allow for energy saving when there is not too much work to do. Modern CPUs can also throttle their speed in order to improve battery life.
Apart from that, there is a difference between the CPU executing "some work" and "useful work". The CPU by itself can't tell, but the operating system usually can. Except for some embedded software, a CPU will never be running a single job, but rather an operating system with different processes within it. If there is no useful process to run, the Operating System will schedule the "idle task" which mostly means putting the CPU to sleep for some time (see above) or jsut burning CPU cycles in a loop which does nothing useful. Calculating the ratio of time spent in idle task to time spent in regular tasks gives the CPU's business factor.
So while in the old days of DOS when the computer was running (almost) only a single task, it was true that it was always doing something. Many applications used so-called busy-waiting if they jus thad to delay their execution for some time, doing nothing useful. But today there will almost always be a smart OS in place which can run the idle process than can put the CPU to sleep, throttle down its speed etc.
Oh boy, this is a toughie. It’s a very practical question as it is a measure of performance and efficiency, and also a very subjective question as it judges what instructions are more or less “useful” toward accomplishing the purpose of an application. The purpose of an application could be just about anything, such as finding the solution to a complex matrix equation or rendering an image on a display.
In addition, modern processors do things like clock gating in power idle states. The oscillator is still producing cycles, but no instructions execute due to certain circuitry being idled due to cycles not reaching them. These are cycles that are not doing anything useful and need to be ignored.
Similarly, modern processors can execute multiple instructions simultaneously, execute them out of order, and predict and execute which instructions will be executed next before your program (i.e. the IP or Instruction Pointer) actually reaches them. You don’t want to include instructions whose execution never actually complete, such as because the processor guesses wrong and has to flush those instructions, e.g. as due to a branch mispredict. So a better metric is counting those instructions that actually complete. Instructions that complete are termed “retired”.
So we should only count those instructions that complete (i.e. retire), and cycles that are actually used to execute instructions (i.e. unhalted).)
Perhaps the most practical general metric for “work” is CPI or cycles-per-instruction: CPI = CPU_CLK_UNHALTED.CORE / INST_RETIRED.ANY. CPU_CLK_UNHALTED.CORE are cycles used to execute actual instructions (vs those “wasted” in an idle state). INST_RETIRED are those instructions that complete (vs those that don’t due to something like a branch mispredict).
Trying to get a more specific metric, such as the instructions that contribute to the solution of a matrix multiple, and excluding instructions that don’t directly contribute to computing the solution, such as control instructions, is very subjective and difficult to gather statistics on. (There are some that you can, such as VECTOR_INTENSITY = VPU_ELEMENTS_ACTIVE / VPU_INSTRUCTIONS_EXECUTED which is the number of SIMD vector operations, such as SSE or AVX, that are executed per second. These instructions are more likely to directly contribute to the solution of a mathematical solution as that is their primary purpose.)
Now that I’ve talked your ear off, check out some of the optimization resources at your local friendly Intel developer resource, software.intel.com. Particularly, check out how to effectively use VTune. I’m not suggesting you need to get VTune though you can get a free or very discounted student license (I think). But the material will tell you a lot about increasing your programs performance (i.e. optimizing), which is, if you think about it, increasing the useful work your program accomplishes.
Expanding on Michał's answer a bit:
Program written for modern multi-tasking OSes are more like a collection of event handlers: they effectively setup listeners for I/O and then yield control back to the OS. The OS wake them up each time there is something to process (e.g. user action, data from device) and they "go to sleep" by calling into the OS once they've finished processing. Most OSes will also preempt in case one process hog the CPU for too long and starve the others.
The OS can then keep tabs on how long each process are actually running (by remembering the start and end time of each run) and generate the statistics like CPU time and load (ready process queue length).
And to answer your second question:
To stop mostly means a process is no longer scheduled and all associated resource (scheduling data structures, file handles, memory space, ...) destroyed. This usually require the process to call a special OS call (syscall/interrupt) so the OS can release the resources gracefully.
If however a process run into an infinite loop and stops responding to OS events, then it can only be forcibly stopped (by simply not running it anymore).
Has anyone else noticed terrible performance when scaling up to use all the cores on a cloud instance with somewhat memory intense jobs (2.5GB in my case)?
When I run jobs locally on my quad xeon chip, the difference between using 1 core and all 4 cores is about a 25% slowdown with all cores. This is to be expected from what I understand; a drop in clock rate as the cores get used up is part of the multi-core chip design.
But when I run the jobs on a multicore virtual instance, I am seeing a slowdown of like 2x - 4x in processing time between using 1 core and all cores. I've seen this on GCE, EC2, and Rackspace instances. And I have tested many difference instance types, mostly the fastest offered.
So has this behavior been seen by others with jobs about the same size in memory usage?
The jobs I am running are written in fortran. I did not write them, and I'm not really a fortran guy so my knowledge of them is limited. I know they have low I/O needs. They appear to be CPU-bound when I watch top as they run. They run without the need to communicate with each other, ie., embarrasingly parallel. They each take about 2.5GB in memory.
So my best guess so far is that jobs that use up this much memory take a big hit by the virtualization layer's memory management. It could also be that my jobs are competing for an I/O resource, but this seems highly unlikely according to an expert.
My workaround for now is to use GCE because they have single-core instance that actually runs the jobs as fast as my laptop's chip, and are priced almost proportionally by core.
You might be running into memory bandwidth constraints, depending on your data access pattern.
The linux perf tool might give some insight into this, though I'll admit that I don't entirely understand your description of the problem. If I understand correctly:
Running one copy of the single-threaded program on your laptop takes X minutes to complete.
Running 4 copies of the single-threaded program on your laptop, each copy takes X * 1.25 minutes to complete.
Running one copy of the single-threaded program on various cloud instances takes X minutes to complete.
Running N copies of the single-threaded program on an N-core virtual cloud instances, each copy takes X * 2-4 minutes to complete.
If so, it sounds like you're either running into a kernel contention or contention for e.g. memory I/O. It would be interesting to see whether various fortran compiler options might help optimize memory access patterns; for example, enabling SSE2 load/store intrinsics or other optimizations. You might also compare results with gcc and intel's fortran compilers.