mirror
/
concurrentqueue
mirror of https://github.com/cameron314/concurrentqueue.git
Code Releases Activity
You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
master
1.0.2
v1.0.0-beta
v1.0.1
v1.0.2
v1.0.3
v1.0.4
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'master'
${ noResults }
concurrentqueue/benchmarks/contrib/benchmarks.aws-8.log

818 lines
50 KiB
Plaintext
Raw Permalink Blame History

# Run on a 8-core c1.xlarge AWS instance
--- New run (Sat Nov 8 19:47:00 2014) ---
Running 64-bit benchmarks on an Intel(R) Xeon(R) CPU E5506 @ 2.13GHz
(precise mode)
Note that these are synthetic benchmarks. Take them with a grain of salt.
Legend:
'Avg': Average time taken per operation, normalized to be per thread
'Range': The minimum and maximum times taken per operation (per thread)
'Ops/s': Overall operations per second
'Ops/s/t': Operations per second per thread (inverse of 'Avg')
Operations include those that fail (e.g. because the queue is empty).
Each logical enqueue/dequeue counts as an individual operation when in bulk.
balanced:
(Measures the average operation speed with multiple symmetrical threads
under reasonable load -- small random intervals between accesses)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 2.1439us Range: [2.0316us, 2.1834us] Ops/s: 932.90k Ops/s/t: 466.45k
3 threads: Avg: 3.8096us Range: [3.3842us, 4.1643us] Ops/s: 787.49k Ops/s/t: 262.50k
4 threads: Avg: 5.5406us Range: [5.0081us, 5.6616us] Ops/s: 721.94k Ops/s/t: 180.48k
8 threads: Avg: 0.0120ms Range: [0.0118ms, 0.0123ms] Ops/s: 664.05k Ops/s/t: 83.01k
12 threads: Avg: 0.0189ms Range: [0.0188ms, 0.0191ms] Ops/s: 633.72k Ops/s/t: 52.81k
16 threads: Avg: 0.0246ms Range: [0.0240ms, 0.0248ms] Ops/s: 650.46k Ops/s/t: 40.65k
Operations per second per thread (weighted average): 133.15k
With tokens
2 threads: Avg: 2.0623us Range: [2.0364us, 2.0712us] Ops/s: 969.77k Ops/s/t: 484.88k
3 threads: Avg: 3.1724us Range: [3.1542us, 3.2327us] Ops/s: 945.67k Ops/s/t: 315.22k
4 threads: Avg: 4.4476us Range: [4.2467us, 4.6332us] Ops/s: 899.36k Ops/s/t: 224.84k
8 threads: Avg: 0.0101ms Range: [9.7889us, 0.0102ms] Ops/s: 795.01k Ops/s/t: 99.38k
12 threads: Avg: 0.0155ms Range: [0.0150ms, 0.0158ms] Ops/s: 773.39k Ops/s/t: 64.45k
16 threads: Avg: 0.0213ms Range: [0.0206ms, 0.0218ms] Ops/s: 750.30k Ops/s/t: 46.89k
Operations per second per thread (weighted average): 153.72k
> boost::lockfree::queue
2 threads: Avg: 2.3908us Range: [2.2293us, 2.6284us] Ops/s: 836.54k Ops/s/t: 418.27k
3 threads: Avg: 4.2701us Range: [3.5851us, 4.3783us] Ops/s: 702.55k Ops/s/t: 234.18k
4 threads: Avg: 6.0101us Range: [5.8873us, 6.0534us] Ops/s: 665.55k Ops/s/t: 166.39k
8 threads: Avg: 0.0123ms Range: [0.0120ms, 0.0125ms] Ops/s: 647.96k Ops/s/t: 81.00k
12 threads: Avg: 0.0191ms Range: [0.0189ms, 0.0193ms] Ops/s: 626.99k Ops/s/t: 52.25k
16 threads: Avg: 0.0244ms Range: [0.0234ms, 0.0248ms] Ops/s: 656.16k Ops/s/t: 41.01k
Operations per second per thread (weighted average): 123.33k
> tbb::concurrent_queue
2 threads: Avg: 2.2272us Range: [2.1376us, 2.3070us] Ops/s: 898.01k Ops/s/t: 449.00k
3 threads: Avg: 4.0825us Range: [4.0605us, 4.1017us] Ops/s: 734.84k Ops/s/t: 244.95k
4 threads: Avg: 5.4065us Range: [5.3347us, 5.4531us] Ops/s: 739.86k Ops/s/t: 184.96k
8 threads: Avg: 0.0115ms Range: [0.0112ms, 0.0116ms] Ops/s: 692.97k Ops/s/t: 86.62k
12 threads: Avg: 0.0171ms Range: [0.0169ms, 0.0172ms] Ops/s: 702.47k Ops/s/t: 58.54k
16 threads: Avg: 0.0228ms Range: [0.0225ms, 0.0232ms] Ops/s: 701.16k Ops/s/t: 43.82k
Operations per second per thread (weighted average): 132.93k
> SimpleLockFreeQueue
2 threads: Avg: 2.2096us Range: [2.2034us, 2.2123us] Ops/s: 905.15k Ops/s/t: 452.58k
3 threads: Avg: 3.9077us Range: [3.5631us, 4.2803us] Ops/s: 767.71k Ops/s/t: 255.90k
4 threads: Avg: 5.8294us Range: [5.7747us, 5.8686us] Ops/s: 686.17k Ops/s/t: 171.54k
8 threads: Avg: 0.0125ms Range: [0.0121ms, 0.0129ms] Ops/s: 641.39k Ops/s/t: 80.17k
12 threads: Avg: 0.0187ms Range: [0.0184ms, 0.0189ms] Ops/s: 640.76k Ops/s/t: 53.40k
16 threads: Avg: 0.0255ms Range: [0.0244ms, 0.0262ms] Ops/s: 626.32k Ops/s/t: 39.15k
Operations per second per thread (weighted average): 129.20k
> LockBasedQueue
2 threads: Avg: 2.6149us Range: [2.2393us, 2.8581us] Ops/s: 764.85k Ops/s/t: 382.43k
3 threads: Avg: 5.0743us Range: [4.9853us, 5.0976us] Ops/s: 591.21k Ops/s/t: 197.07k
4 threads: Avg: 7.6434us Range: [7.3726us, 7.7969us] Ops/s: 523.33k Ops/s/t: 130.83k
8 threads: Avg: 0.0302ms Range: [0.0269ms, 0.0328ms] Ops/s: 264.68k Ops/s/t: 33.09k
12 threads: Avg: 0.0637ms Range: [0.0509ms, 0.0682ms] Ops/s: 188.37k Ops/s/t: 15.70k
16 threads: Avg: 0.1120ms Range: [0.1001ms, 0.1184ms] Ops/s: 142.90k Ops/s/t: 8.93k
Operations per second per thread (weighted average): 85.99k
> std::queue
(skipping, benchmark not supported...)
only enqueue:
(Measures the average operation speed when all threads are producers)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0229us Range: [0.0229us, 0.0230us] Ops/s: 43.58M Ops/s/t: 43.58M
2 threads: Avg: 0.0418us Range: [0.0416us, 0.0419us] Ops/s: 47.84M Ops/s/t: 23.92M
4 threads: Avg: 0.0821us Range: [0.0821us, 0.0822us] Ops/s: 48.71M Ops/s/t: 12.18M
8 threads: Avg: 0.1621us Range: [0.1616us, 0.1625us] Ops/s: 49.35M Ops/s/t: 6.17M
12 threads: Avg: 0.2503us Range: [0.2498us, 0.2507us] Ops/s: 47.94M Ops/s/t: 3.99M
16 threads: Avg: 0.3417us Range: [0.3406us, 0.3428us] Ops/s: 46.82M Ops/s/t: 2.93M
Operations per second per thread (weighted average): 12.03M
With tokens
1 thread: Avg: 0.0146us Range: [0.0146us, 0.0146us] Ops/s: 68.45M Ops/s/t: 68.45M
2 threads: Avg: 0.0315us Range: [0.0313us, 0.0316us] Ops/s: 63.53M Ops/s/t: 31.77M
4 threads: Avg: 0.0614us Range: [0.0613us, 0.0615us] Ops/s: 65.10M Ops/s/t: 16.27M
8 threads: Avg: 0.1220us Range: [0.1216us, 0.1222us] Ops/s: 65.56M Ops/s/t: 8.20M
12 threads: Avg: 0.1901us Range: [0.1884us, 0.1909us] Ops/s: 63.14M Ops/s/t: 5.26M
16 threads: Avg: 0.2694us Range: [0.2658us, 0.2733us] Ops/s: 59.38M Ops/s/t: 3.71M
Operations per second per thread (weighted average): 16.96M
> boost::lockfree::queue
1 thread: Avg: 0.0745us Range: [0.0744us, 0.0745us] Ops/s: 13.43M Ops/s/t: 13.43M
2 threads: Avg: 0.7351us Range: [0.6376us, 0.8720us] Ops/s: 2.72M Ops/s/t: 1.36M
4 threads: Avg: 5.9754us Range: [5.5307us, 6.1937us] Ops/s: 669.41k Ops/s/t: 167.35k
8 threads: Avg: 0.0276ms Range: [0.0263ms, 0.0282ms] Ops/s: 289.71k Ops/s/t: 36.21k
12 threads: Avg: 0.0486ms Range: [0.0222ms, 0.0555ms] Ops/s: 246.97k Ops/s/t: 20.58k
16 threads: Avg: 0.1059ms Range: [0.1018ms, 0.1092ms] Ops/s: 151.02k Ops/s/t: 9.44k
Operations per second per thread (weighted average): 1.45M
> tbb::concurrent_queue
1 thread: Avg: 0.0443us Range: [0.0442us, 0.0443us] Ops/s: 22.60M Ops/s/t: 22.60M
2 threads: Avg: 0.3436us Range: [0.3082us, 0.3696us] Ops/s: 5.82M Ops/s/t: 2.91M
4 threads: Avg: 1.6788us Range: [1.6172us, 1.7061us] Ops/s: 2.38M Ops/s/t: 595.67k
8 threads: Avg: 8.5452us Range: [8.4615us, 8.6459us] Ops/s: 936.20k Ops/s/t: 117.03k
12 threads: Avg: 0.0475ms Range: [0.0469ms, 0.0482ms] Ops/s: 252.43k Ops/s/t: 21.04k
16 threads: Avg: 0.1398ms Range: [0.1317ms, 0.1445ms] Ops/s: 114.48k Ops/s/t: 7.16k
Operations per second per thread (weighted average): 2.61M
> SimpleLockFreeQueue
1 thread: Avg: 0.0739us Range: [0.0738us, 0.0739us] Ops/s: 13.54M Ops/s/t: 13.54M
2 threads: Avg: 0.4404us Range: [0.4023us, 0.4634us] Ops/s: 4.54M Ops/s/t: 2.27M
4 threads: Avg: 3.1988us Range: [3.1510us, 3.2464us] Ops/s: 1.25M Ops/s/t: 312.62k
8 threads: Avg: 0.0103ms Range: [0.0102ms, 0.0105ms] Ops/s: 774.38k Ops/s/t: 96.80k
12 threads: Avg: 0.0215ms Range: [0.0212ms, 0.0216ms] Ops/s: 557.77k Ops/s/t: 46.48k
16 threads: Avg: 0.0401ms Range: [0.0394ms, 0.0407ms] Ops/s: 398.68k Ops/s/t: 24.92k
Operations per second per thread (weighted average): 1.62M
> LockBasedQueue
1 thread: Avg: 0.0696us Range: [0.0696us, 0.0696us] Ops/s: 14.36M Ops/s/t: 14.36M
2 threads: Avg: 0.8199us Range: [0.6931us, 0.8818us] Ops/s: 2.44M Ops/s/t: 1.22M
4 threads: Avg: 5.0074us Range: [4.9400us, 5.0589us] Ops/s: 798.82k Ops/s/t: 199.70k
8 threads: Avg: 0.0233ms Range: [0.0220ms, 0.0240ms] Ops/s: 342.66k Ops/s/t: 42.83k
12 threads: Avg: 0.0427ms Range: [0.0218ms, 0.0492ms] Ops/s: 281.23k Ops/s/t: 23.44k
16 threads: Avg: 0.0845ms Range: [0.0782ms, 0.0872ms] Ops/s: 189.45k Ops/s/t: 11.84k
Operations per second per thread (weighted average): 1.53M
> std::queue
1 thread: Avg: 0.0109us Range: [0.0108us, 0.0109us] Ops/s: 92.13M Ops/s/t: 92.13M
Operations per second per thread (weighted average): 92.13M
only enqueue (pre-allocated):
(Measures the average operation speed when all threads are producers,
and the queue has been stretched out first)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0175us Range: [0.0175us, 0.0176us] Ops/s: 56.98M Ops/s/t: 56.98M
2 threads: Avg: 0.0343us Range: [0.0337us, 0.0368us] Ops/s: 58.31M Ops/s/t: 29.15M
4 threads: Avg: 0.1314us Range: [0.1059us, 0.1515us] Ops/s: 30.43M Ops/s/t: 7.61M
8 threads: Avg: 0.5502us Range: [0.4973us, 0.5676us] Ops/s: 14.54M Ops/s/t: 1.82M
Operations per second per thread (weighted average): 16.37M
With tokens
1 thread: Avg: 8.0792ns Range: [8.0657ns, 8.0857ns] Ops/s: 123.77M Ops/s/t: 123.77M
2 threads: Avg: 0.0199us Range: [0.0198us, 0.0200us] Ops/s: 100.56M Ops/s/t: 50.28M
4 threads: Avg: 0.0362us Range: [0.0361us, 0.0363us] Ops/s: 110.40M Ops/s/t: 27.60M
8 threads: Avg: 0.0730us Range: [0.0724us, 0.0734us] Ops/s: 109.61M Ops/s/t: 13.70M
Operations per second per thread (weighted average): 39.88M
> boost::lockfree::queue
1 thread: Avg: 0.0477us Range: [0.0476us, 0.0478us] Ops/s: 20.96M Ops/s/t: 20.96M
2 threads: Avg: 0.7644us Range: [0.6611us, 0.8702us] Ops/s: 2.62M Ops/s/t: 1.31M
4 threads: Avg: 6.5308us Range: [6.0638us, 6.9217us] Ops/s: 612.48k Ops/s/t: 153.12k
8 threads: Avg: 0.0276ms Range: [0.0192ms, 0.0292ms] Ops/s: 290.08k Ops/s/t: 36.26k
Operations per second per thread (weighted average): 3.21M
> tbb::concurrent_queue
1 thread: Avg: 0.0440us Range: [0.0439us, 0.0440us] Ops/s: 22.75M Ops/s/t: 22.75M
2 threads: Avg: 0.2923us Range: [0.2799us, 0.3120us] Ops/s: 6.84M Ops/s/t: 3.42M
4 threads: Avg: 1.5782us Range: [1.4991us, 1.6202us] Ops/s: 2.53M Ops/s/t: 633.63k
8 threads: Avg: 8.5111us Range: [8.4543us, 8.6016us] Ops/s: 939.95k Ops/s/t: 117.49k
Operations per second per thread (weighted average): 4.03M
> SimpleLockFreeQueue
1 thread: Avg: 0.0563us Range: [0.0563us, 0.0564us] Ops/s: 17.76M Ops/s/t: 17.76M
2 threads: Avg: 0.6085us Range: [0.6031us, 0.6223us] Ops/s: 3.29M Ops/s/t: 1.64M
4 threads: Avg: 4.8504us Range: [3.0550us, 5.2267us] Ops/s: 824.68k Ops/s/t: 206.17k
8 threads: Avg: 0.0211ms Range: [0.0205ms, 0.0215ms] Ops/s: 378.88k Ops/s/t: 47.36k
Operations per second per thread (weighted average): 2.85M
> LockBasedQueue
1 thread: Avg: 0.0698us Range: [0.0698us, 0.0698us] Ops/s: 14.33M Ops/s/t: 14.33M
2 threads: Avg: 0.7667us Range: [0.7002us, 0.8237us] Ops/s: 2.61M Ops/s/t: 1.30M
4 threads: Avg: 5.0945us Range: [4.9227us, 5.1724us] Ops/s: 785.17k Ops/s/t: 196.29k
8 threads: Avg: 0.0233ms Range: [0.0224ms, 0.0241ms] Ops/s: 343.59k Ops/s/t: 42.95k
Operations per second per thread (weighted average): 2.30M
> std::queue
1 thread: Avg: 0.0107us Range: [0.0107us, 0.0107us] Ops/s: 93.28M Ops/s/t: 93.28M
Operations per second per thread (weighted average): 93.28M
only enqueue bulk:
(Measures the average speed of enqueueing an item in bulk when all threads are producers)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 8.7326ns Range: [8.7251ns, 8.7431ns] Ops/s: 114.51M Ops/s/t: 114.51M
2 threads: Avg: 0.0174us Range: [0.0171us, 0.0176us] Ops/s: 114.97M Ops/s/t: 57.49M
4 threads: Avg: 0.0346us Range: [0.0342us, 0.0347us] Ops/s: 115.60M Ops/s/t: 28.90M
8 threads: Avg: 0.0669us Range: [0.0605us, 0.0692us] Ops/s: 119.51M Ops/s/t: 14.94M
12 threads: Avg: 0.1146us Range: [0.1120us, 0.1169us] Ops/s: 104.72M Ops/s/t: 8.73M
16 threads: Avg: 0.1741us Range: [0.1700us, 0.1769us] Ops/s: 91.92M Ops/s/t: 5.75M
Operations per second per thread (weighted average): 28.44M
With tokens
1 thread: Avg: 8.7317ns Range: [8.7133ns, 8.7451ns] Ops/s: 114.53M Ops/s/t: 114.53M
2 threads: Avg: 0.0186us Range: [0.0184us, 0.0187us] Ops/s: 107.81M Ops/s/t: 53.91M
4 threads: Avg: 0.0382us Range: [0.0378us, 0.0383us] Ops/s: 104.74M Ops/s/t: 26.19M
8 threads: Avg: 0.0793us Range: [0.0790us, 0.0796us] Ops/s: 100.83M Ops/s/t: 12.60M
12 threads: Avg: 0.1229us Range: [0.1217us, 0.1232us] Ops/s: 97.64M Ops/s/t: 8.14M
16 threads: Avg: 0.1745us Range: [0.1677us, 0.1802us] Ops/s: 91.69M Ops/s/t: 5.73M
Operations per second per thread (weighted average): 27.74M
> boost::lockfree::queue
(skipping, benchmark not supported...)
> tbb::concurrent_queue
(skipping, benchmark not supported...)
> SimpleLockFreeQueue
(skipping, benchmark not supported...)
> LockBasedQueue
(skipping, benchmark not supported...)
> std::queue
(skipping, benchmark not supported...)
only enqueue bulk (pre-allocated):
(Measures the average speed of enqueueing an item in bulk when all threads are producers,
and the queue has been stretched out first)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 9.0071ns Range: [8.9924ns, 9.0178ns] Ops/s: 111.02M Ops/s/t: 111.02M
2 threads: Avg: 0.0187us Range: [0.0184us, 0.0188us] Ops/s: 106.79M Ops/s/t: 53.40M
4 threads: Avg: 0.0395us Range: [0.0393us, 0.0397us] Ops/s: 101.25M Ops/s/t: 25.31M
8 threads: Avg: 0.0831us Range: [0.0823us, 0.0835us] Ops/s: 96.28M Ops/s/t: 12.04M
Operations per second per thread (weighted average): 37.45M
With tokens
1 thread: Avg: 8.7146ns Range: [8.7099ns, 8.7187ns] Ops/s: 114.75M Ops/s/t: 114.75M
2 threads: Avg: 0.0185us Range: [0.0181us, 0.0187us] Ops/s: 108.27M Ops/s/t: 54.14M
4 threads: Avg: 0.0383us Range: [0.0376us, 0.0384us] Ops/s: 104.54M Ops/s/t: 26.13M
8 threads: Avg: 0.0798us Range: [0.0794us, 0.0801us] Ops/s: 100.25M Ops/s/t: 12.53M
Operations per second per thread (weighted average): 38.52M
> boost::lockfree::queue
(skipping, benchmark not supported...)
> tbb::concurrent_queue
(skipping, benchmark not supported...)
> SimpleLockFreeQueue
(skipping, benchmark not supported...)
> LockBasedQueue
(skipping, benchmark not supported...)
> std::queue
(skipping, benchmark not supported...)
only dequeue:
(Measures the average operation speed when all threads are consumers)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0479us Range: [0.0478us, 0.0479us] Ops/s: 20.90M Ops/s/t: 20.90M
2 threads: Avg: 0.4099us Range: [0.3972us, 0.4170us] Ops/s: 4.88M Ops/s/t: 2.44M
4 threads: Avg: 2.7666us Range: [2.0147us, 2.8988us] Ops/s: 1.45M Ops/s/t: 361.46k
8 threads: Avg: 9.0300us Range: [8.8027us, 9.1078us] Ops/s: 885.94k Ops/s/t: 110.74k
12 threads: Avg: 0.0179ms Range: [0.0178ms, 0.0179ms] Ops/s: 672.15k Ops/s/t: 56.01k
16 threads: Avg: 0.0338ms Range: [0.0332ms, 0.0340ms] Ops/s: 473.79k Ops/s/t: 29.61k
Operations per second per thread (weighted average): 1.56M
With tokens
1 thread: Avg: 0.0404us Range: [0.0404us, 0.0404us] Ops/s: 24.75M Ops/s/t: 24.75M
2 threads: Avg: 0.0827us Range: [0.0826us, 0.0827us] Ops/s: 24.18M Ops/s/t: 12.09M
4 threads: Avg: 0.1745us Range: [0.1738us, 0.1748us] Ops/s: 22.92M Ops/s/t: 5.73M
8 threads: Avg: 0.3536us Range: [0.3514us, 0.3559us] Ops/s: 22.62M Ops/s/t: 2.83M
12 threads: Avg: 0.5554us Range: [0.5457us, 0.5628us] Ops/s: 21.60M Ops/s/t: 1.80M
16 threads: Avg: 0.7653us Range: [0.7439us, 0.7776us] Ops/s: 20.91M Ops/s/t: 1.31M
Operations per second per thread (weighted average): 5.07M
> boost::lockfree::queue
1 thread: Avg: 0.0356us Range: [0.0355us, 0.0356us] Ops/s: 28.11M Ops/s/t: 28.11M
2 threads: Avg: 0.4965us Range: [0.4550us, 0.5404us] Ops/s: 4.03M Ops/s/t: 2.01M
4 threads: Avg: 4.5768us Range: [4.4660us, 4.6482us] Ops/s: 873.96k Ops/s/t: 218.49k
8 threads: Avg: 0.0189ms Range: [0.0129ms, 0.0203ms] Ops/s: 423.34k Ops/s/t: 52.92k
12 threads: Avg: 0.0424ms Range: [0.0411ms, 0.0432ms] Ops/s: 282.73k Ops/s/t: 23.56k
16 threads: Avg: 0.0767ms Range: [0.0523ms, 0.0825ms] Ops/s: 208.48k Ops/s/t: 13.03k
Operations per second per thread (weighted average): 1.85M
> tbb::concurrent_queue
1 thread: Avg: 0.0290us Range: [0.0286us, 0.0292us] Ops/s: 34.47M Ops/s/t: 34.47M
2 threads: Avg: 0.2343us Range: [0.2270us, 0.2470us] Ops/s: 8.54M Ops/s/t: 4.27M
4 threads: Avg: 2.0731us Range: [1.9837us, 2.1597us] Ops/s: 1.93M Ops/s/t: 482.36k
8 threads: Avg: 0.0103ms Range: [9.5712us, 0.0106ms] Ops/s: 779.17k Ops/s/t: 97.40k
12 threads: Avg: 0.0451ms Range: [0.0433ms, 0.0480ms] Ops/s: 266.36k Ops/s/t: 22.20k
16 threads: Avg: 0.1436ms Range: [0.1360ms, 0.1520ms] Ops/s: 111.41k Ops/s/t: 6.96k
Operations per second per thread (weighted average): 2.51M
> SimpleLockFreeQueue
1 thread: Avg: 0.0275us Range: [0.0275us, 0.0276us] Ops/s: 36.31M Ops/s/t: 36.31M
2 threads: Avg: 0.8440us Range: [0.7192us, 0.8839us] Ops/s: 2.37M Ops/s/t: 1.18M
4 threads: Avg: 6.3769us Range: [5.6198us, 6.7313us] Ops/s: 627.27k Ops/s/t: 156.82k
8 threads: Avg: 0.0292ms Range: [0.0263ms, 0.0300ms] Ops/s: 274.34k Ops/s/t: 34.29k
12 threads: Avg: 0.0642ms Range: [0.0548ms, 0.0662ms] Ops/s: 186.79k Ops/s/t: 15.57k
16 threads: Avg: 0.1196ms Range: [0.1172ms, 0.1206ms] Ops/s: 133.76k Ops/s/t: 8.36k
Operations per second per thread (weighted average): 2.22M
> LockBasedQueue
1 thread: Avg: 0.0617us Range: [0.0617us, 0.0618us] Ops/s: 16.20M Ops/s/t: 16.20M
2 threads: Avg: 1.4892us Range: [1.3851us, 1.5315us] Ops/s: 1.34M Ops/s/t: 671.48k
4 threads: Avg: 8.0076us Range: [4.8463us, 8.5494us] Ops/s: 499.53k Ops/s/t: 124.88k
8 threads: Avg: 0.0295ms Range: [0.0202ms, 0.0314ms] Ops/s: 271.52k Ops/s/t: 33.94k
12 threads: Avg: 0.0594ms Range: [0.0364ms, 0.0639ms] Ops/s: 201.87k Ops/s/t: 16.82k
16 threads: Avg: 0.1049ms Range: [0.0564ms, 0.1191ms] Ops/s: 152.58k Ops/s/t: 9.54k
Operations per second per thread (weighted average): 1.02M
> std::queue
1 thread: Avg: 3.8947ns Range: [3.8927ns, 3.8964ns] Ops/s: 256.76M Ops/s/t: 256.76M
Operations per second per thread (weighted average): 256.76M
only dequeue bulk:
(Measures the average speed of dequeueing an item in bulk when all threads are consumers)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 1.8063ns Range: [1.8008ns, 1.8111ns] Ops/s: 553.61M Ops/s/t: 553.61M
2 threads: Avg: 5.3528ns Range: [5.3256ns, 5.3836ns] Ops/s: 373.63M Ops/s/t: 186.82M
4 threads: Avg: 0.0278us Range: [0.0257us, 0.0288us] Ops/s: 143.65M Ops/s/t: 35.91M
8 threads: Avg: 0.0860us Range: [0.0839us, 0.0868us] Ops/s: 93.02M Ops/s/t: 11.63M
12 threads: Avg: 0.1500us Range: [0.1429us, 0.1535us] Ops/s: 79.98M Ops/s/t: 6.66M
16 threads: Avg: 0.2670us Range: [0.2490us, 0.2774us] Ops/s: 59.92M Ops/s/t: 3.74M
Operations per second per thread (weighted average): 63.04M
With tokens
1 thread: Avg: 1.6650ns Range: [1.6632ns, 1.6664ns] Ops/s: 600.59M Ops/s/t: 600.59M
2 threads: Avg: 4.6593ns Range: [4.5874ns, 4.7052ns] Ops/s: 429.25M Ops/s/t: 214.62M
4 threads: Avg: 0.0159us Range: [0.0158us, 0.0160us] Ops/s: 251.62M Ops/s/t: 62.91M
8 threads: Avg: 0.0515us Range: [0.0498us, 0.0528us] Ops/s: 155.40M Ops/s/t: 19.43M
12 threads: Avg: 0.0787us Range: [0.0764us, 0.0798us] Ops/s: 152.54M Ops/s/t: 12.71M
16 threads: Avg: 0.1374us Range: [0.1252us, 0.1435us] Ops/s: 116.47M Ops/s/t: 7.28M
Operations per second per thread (weighted average): 79.24M
> boost::lockfree::queue
(skipping, benchmark not supported...)
> tbb::concurrent_queue
(skipping, benchmark not supported...)
> SimpleLockFreeQueue
(skipping, benchmark not supported...)
> LockBasedQueue
(skipping, benchmark not supported...)
> std::queue
(skipping, benchmark not supported...)
mostly enqueue:
(Measures the average operation speed when most threads are enqueueing)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.0937us Range: [0.0927us, 0.0963us] Ops/s: 21.35M Ops/s/t: 10.68M
4 threads: Avg: 0.1862us Range: [0.1816us, 0.1900us] Ops/s: 21.48M Ops/s/t: 5.37M
8 threads: Avg: 0.8762us Range: [0.7998us, 0.9349us] Ops/s: 9.13M Ops/s/t: 1.14M
Operations per second per thread (weighted average): 4.66M
With tokens
2 threads: Avg: 0.0843us Range: [0.0841us, 0.0847us] Ops/s: 23.72M Ops/s/t: 11.86M
4 threads: Avg: 0.1088us Range: [0.1029us, 0.1111us] Ops/s: 36.75M Ops/s/t: 9.19M
8 threads: Avg: 0.3251us Range: [0.2290us, 0.4168us] Ops/s: 24.61M Ops/s/t: 3.08M
Operations per second per thread (weighted average): 7.02M
> boost::lockfree::queue
2 threads: Avg: 0.1152us Range: [0.1118us, 0.1171us] Ops/s: 17.37M Ops/s/t: 8.68M
4 threads: Avg: 2.5790us Range: [1.8601us, 3.1117us] Ops/s: 1.55M Ops/s/t: 387.75k
8 threads: Avg: 0.0170ms Range: [0.0166ms, 0.0176ms] Ops/s: 469.45k Ops/s/t: 58.68k
Operations per second per thread (weighted average): 2.12M
> tbb::concurrent_queue
2 threads: Avg: 0.3593us Range: [0.3540us, 0.3657us] Ops/s: 5.57M Ops/s/t: 2.78M
4 threads: Avg: 1.1523us Range: [1.1393us, 1.1810us] Ops/s: 3.47M Ops/s/t: 867.86k
8 threads: Avg: 5.9296us Range: [5.4979us, 6.1961us] Ops/s: 1.35M Ops/s/t: 168.65k
Operations per second per thread (weighted average): 984.95k
> SimpleLockFreeQueue
2 threads: Avg: 0.2142us Range: [0.1944us, 0.2233us] Ops/s: 9.34M Ops/s/t: 4.67M
4 threads: Avg: 2.7030us Range: [2.4932us, 2.7637us] Ops/s: 1.48M Ops/s/t: 369.96k
8 threads: Avg: 0.0127ms Range: [9.8947us, 0.0138ms] Ops/s: 628.60k Ops/s/t: 78.58k
Operations per second per thread (weighted average): 1.21M
> LockBasedQueue
2 threads: Avg: 0.2057us Range: [0.1952us, 0.2165us] Ops/s: 9.72M Ops/s/t: 4.86M
4 threads: Avg: 6.1602us Range: [3.7945us, 7.2017us] Ops/s: 649.33k Ops/s/t: 162.33k
8 threads: Avg: 0.0375ms Range: [0.0334ms, 0.0419ms] Ops/s: 213.17k Ops/s/t: 26.65k
Operations per second per thread (weighted average): 1.17M
> std::queue
(skipping, benchmark not supported...)
mostly enqueue bulk:
(Measures the average speed of enqueueing an item in bulk under light contention)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.0172us Range: [0.0172us, 0.0173us] Ops/s: 116.00M Ops/s/t: 58.00M
4 threads: Avg: 0.0366us Range: [0.0363us, 0.0369us] Ops/s: 109.38M Ops/s/t: 27.34M
8 threads: Avg: 0.0797us Range: [0.0794us, 0.0801us] Ops/s: 100.37M Ops/s/t: 12.55M
Operations per second per thread (weighted average): 27.58M
With tokens
2 threads: Avg: 0.0171us Range: [0.0171us, 0.0171us] Ops/s: 117.11M Ops/s/t: 58.55M
4 threads: Avg: 0.0304us Range: [0.0280us, 0.0320us] Ops/s: 131.48M Ops/s/t: 32.87M
8 threads: Avg: 0.0714us Range: [0.0612us, 0.0770us] Ops/s: 112.07M Ops/s/t: 14.01M
Operations per second per thread (weighted average): 30.14M
> boost::lockfree::queue
(skipping, benchmark not supported...)
> tbb::concurrent_queue
(skipping, benchmark not supported...)
> SimpleLockFreeQueue
(skipping, benchmark not supported...)
> LockBasedQueue
(skipping, benchmark not supported...)
> std::queue
(skipping, benchmark not supported...)
mostly dequeue:
(Measures the average operation speed when most threads are dequeueing)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.1035us Range: [0.1034us, 0.1036us] Ops/s: 19.32M Ops/s/t: 9.66M
4 threads: Avg: 1.4781us Range: [1.4440us, 1.4955us] Ops/s: 2.71M Ops/s/t: 676.56k
8 threads: Avg: 3.8335us Range: [2.7691us, 4.1910us] Ops/s: 2.09M Ops/s/t: 260.86k
Operations per second per thread (weighted average): 2.52M
With tokens
2 threads: Avg: 0.0843us Range: [0.0842us, 0.0844us] Ops/s: 23.72M Ops/s/t: 11.86M
4 threads: Avg: 0.7221us Range: [0.6558us, 0.7557us] Ops/s: 5.54M Ops/s/t: 1.38M
8 threads: Avg: 1.0831us Range: [0.8750us, 1.2979us] Ops/s: 7.39M Ops/s/t: 923.26k
Operations per second per thread (weighted average): 3.55M
> boost::lockfree::queue
2 threads: Avg: 0.4447us Range: [0.3836us, 0.4890us] Ops/s: 4.50M Ops/s/t: 2.25M
4 threads: Avg: 2.3882us Range: [1.9014us, 2.7969us] Ops/s: 1.67M Ops/s/t: 418.73k
8 threads: Avg: 0.0132ms Range: [7.8331us, 0.0142ms] Ops/s: 604.38k Ops/s/t: 75.55k
Operations per second per thread (weighted average): 677.76k
> tbb::concurrent_queue
2 threads: Avg: 0.1613us Range: [0.1609us, 0.1618us] Ops/s: 12.40M Ops/s/t: 6.20M
4 threads: Avg: 1.6693us Range: [1.5484us, 1.7851us] Ops/s: 2.40M Ops/s/t: 599.07k
8 threads: Avg: 7.8647us Range: [7.4879us, 8.0231us] Ops/s: 1.02M Ops/s/t: 127.15k
Operations per second per thread (weighted average): 1.65M
> SimpleLockFreeQueue
2 threads: Avg: 0.4576us Range: [0.4205us, 0.4906us] Ops/s: 4.37M Ops/s/t: 2.19M
4 threads: Avg: 3.7064us Range: [3.5114us, 3.8460us] Ops/s: 1.08M Ops/s/t: 269.81k
8 threads: Avg: 0.0188ms Range: [0.0186ms, 0.0189ms] Ops/s: 426.30k Ops/s/t: 53.29k
Operations per second per thread (weighted average): 605.60k
> LockBasedQueue
2 threads: Avg: 0.7979us Range: [0.7404us, 0.8972us] Ops/s: 2.51M Ops/s/t: 1.25M
4 threads: Avg: 7.7466us Range: [7.5002us, 7.9109us] Ops/s: 516.35k Ops/s/t: 129.09k
8 threads: Avg: 0.0258ms Range: [0.0248ms, 0.0267ms] Ops/s: 310.04k Ops/s/t: 38.76k
Operations per second per thread (weighted average): 342.85k
> std::queue
(skipping, benchmark not supported...)
mostly dequeue bulk:
(Measures the average speed of dequeueing an item in bulk under light contention)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 8.3404ns Range: [7.7191ns, 8.9024ns] Ops/s: 239.80M Ops/s/t: 119.90M
4 threads: Avg: 0.0264us Range: [0.0223us, 0.0274us] Ops/s: 151.37M Ops/s/t: 37.84M
8 threads: Avg: 0.1042us Range: [0.0994us, 0.1064us] Ops/s: 76.75M Ops/s/t: 9.59M
Operations per second per thread (weighted average): 43.63M
With tokens
2 threads: Avg: 4.9164ns Range: [4.8544ns, 5.0084ns] Ops/s: 406.80M Ops/s/t: 203.40M
4 threads: Avg: 0.0156us Range: [0.0155us, 0.0158us] Ops/s: 256.40M Ops/s/t: 64.10M
8 threads: Avg: 0.0518us Range: [0.0448us, 0.0607us] Ops/s: 154.57M Ops/s/t: 19.32M
Operations per second per thread (weighted average): 75.37M
> boost::lockfree::queue
(skipping, benchmark not supported...)
> tbb::concurrent_queue
(skipping, benchmark not supported...)
> SimpleLockFreeQueue
(skipping, benchmark not supported...)
> LockBasedQueue
(skipping, benchmark not supported...)
> std::queue
(skipping, benchmark not supported...)
single-producer, multi-consumer (measuring all but 1 thread):
(Measures the average speed of dequeueing with only one producer, but multiple consumers)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.1041us Range: [0.0978us, 0.1122us] Ops/s: 9.60M Ops/s/t: 9.60M
4 threads: Avg: 1.9386us Range: [1.7129us, 2.3281us] Ops/s: 1.55M Ops/s/t: 515.83k
8 threads: Avg: 0.7837us Range: [0.5718us, 1.4367us] Ops/s: 8.93M Ops/s/t: 1.28M
16 threads: Avg: 0.8587us Range: [0.8096us, 1.0077us] Ops/s: 17.47M Ops/s/t: 1.16M
Operations per second per thread (weighted average): 1.99M
With tokens
2 threads: Avg: 0.1034us Range: [0.0984us, 0.1079us] Ops/s: 9.67M Ops/s/t: 9.67M
4 threads: Avg: 1.4904us Range: [1.3998us, 1.6162us] Ops/s: 2.01M Ops/s/t: 670.97k
8 threads: Avg: 0.9243us Range: [0.3801us, 6.2399us] Ops/s: 7.57M Ops/s/t: 1.08M
16 threads: Avg: 0.7863us Range: [0.6860us, 0.8744us] Ops/s: 19.08M Ops/s/t: 1.27M
Operations per second per thread (weighted average): 2.01M
> boost::lockfree::queue
2 threads: Avg: 0.1390us Range: [0.1325us, 0.1545us] Ops/s: 7.19M Ops/s/t: 7.19M
4 threads: Avg: 0.4625us Range: [0.2754us, 0.5561us] Ops/s: 6.49M Ops/s/t: 2.16M
8 threads: Avg: 0.7324us Range: [0.5028us, 1.2586us] Ops/s: 9.56M Ops/s/t: 1.37M
16 threads: Avg: 0.4086us Range: [0.3950us, 0.4355us] Ops/s: 36.71M Ops/s/t: 2.45M
Operations per second per thread (weighted average): 2.60M
> tbb::concurrent_queue
2 threads: Avg: 0.0396us Range: [0.0395us, 0.0396us] Ops/s: 25.27M Ops/s/t: 25.27M
4 threads: Avg: 0.4785us Range: [0.3555us, 2.2952us] Ops/s: 6.27M Ops/s/t: 2.09M
8 threads: Avg: 6.6917us Range: [2.8804us, 8.9630us] Ops/s: 1.05M Ops/s/t: 149.44k
16 threads: Avg: 6.2738us Range: [4.4906us, 0.0135ms] Ops/s: 2.39M Ops/s/t: 159.39k
Operations per second per thread (weighted average): 3.23M
> SimpleLockFreeQueue
2 threads: Avg: 0.1923us Range: [0.1744us, 0.2023us] Ops/s: 5.20M Ops/s/t: 5.20M
4 threads: Avg: 1.5853us Range: [0.4197us, 2.7908us] Ops/s: 1.89M Ops/s/t: 630.79k
8 threads: Avg: 1.1113us Range: [0.8406us, 1.7108us] Ops/s: 6.30M Ops/s/t: 899.83k
16 threads: Avg: 0.5356us Range: [0.5155us, 0.5548us] Ops/s: 28.01M Ops/s/t: 1.87M
Operations per second per thread (weighted average): 1.72M
> LockBasedQueue
2 threads: Avg: 0.1914us Range: [0.1719us, 0.2080us] Ops/s: 5.22M Ops/s/t: 5.22M
4 threads: Avg: 2.8230us Range: [2.6594us, 2.9641us] Ops/s: 1.06M Ops/s/t: 354.23k
8 threads: Avg: 8.1605us Range: [7.5421us, 9.6067us] Ops/s: 857.79k Ops/s/t: 122.54k
16 threads: Avg: 0.0330ms Range: [0.0316ms, 0.0339ms] Ops/s: 455.18k Ops/s/t: 30.35k
Operations per second per thread (weighted average): 678.72k
> std::queue
(skipping, benchmark not supported...)
single-producer, multi-consumer (pre-produced):
(Measures the average speed of dequeueing from a queue pre-filled by one thread)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0479us Range: [0.0478us, 0.0479us] Ops/s: 20.89M Ops/s/t: 20.89M
3 threads: Avg: 1.0918us Range: [0.9990us, 1.1343us] Ops/s: 2.75M Ops/s/t: 915.91k
7 threads: Avg: 6.1592us Range: [6.0475us, 6.1973us] Ops/s: 1.14M Ops/s/t: 162.36k
15 threads: Avg: 0.0243ms Range: [0.0240ms, 0.0245ms] Ops/s: 617.10k Ops/s/t: 41.14k
Operations per second per thread (weighted average): 2.49M
With tokens
1 thread: Avg: 0.0404us Range: [0.0404us, 0.0404us] Ops/s: 24.74M Ops/s/t: 24.74M
3 threads: Avg: 0.9008us Range: [0.8752us, 0.9113us] Ops/s: 3.33M Ops/s/t: 1.11M
7 threads: Avg: 4.7648us Range: [4.7195us, 4.8146us] Ops/s: 1.47M Ops/s/t: 209.87k
15 threads: Avg: 0.0194ms Range: [0.0191ms, 0.0196ms] Ops/s: 772.72k Ops/s/t: 51.51k
Operations per second per thread (weighted average): 2.96M
> boost::lockfree::queue
1 thread: Avg: 0.0356us Range: [0.0355us, 0.0357us] Ops/s: 28.11M Ops/s/t: 28.11M
3 threads: Avg: 2.1016us Range: [1.8879us, 2.2375us] Ops/s: 1.43M Ops/s/t: 475.82k
7 threads: Avg: 0.0161ms Range: [0.0149ms, 0.0163ms] Ops/s: 435.41k Ops/s/t: 62.20k
15 threads: Avg: 0.0692ms Range: [0.0458ms, 0.0738ms] Ops/s: 216.64k Ops/s/t: 14.44k
Operations per second per thread (weighted average): 3.15M
> tbb::concurrent_queue
1 thread: Avg: 0.0286us Range: [0.0286us, 0.0287us] Ops/s: 34.91M Ops/s/t: 34.91M
3 threads: Avg: 0.9964us Range: [0.9581us, 1.0265us] Ops/s: 3.01M Ops/s/t: 1.00M
7 threads: Avg: 5.4370us Range: [5.3420us, 5.4884us] Ops/s: 1.29M Ops/s/t: 183.93k
15 threads: Avg: 0.1103ms Range: [0.1003ms, 0.1155ms] Ops/s: 135.94k Ops/s/t: 9.06k
Operations per second per thread (weighted average): 4.02M
> SimpleLockFreeQueue
1 thread: Avg: 0.0275us Range: [0.0275us, 0.0275us] Ops/s: 36.31M Ops/s/t: 36.31M
3 threads: Avg: 3.3246us Range: [2.6774us, 3.5321us] Ops/s: 902.37k Ops/s/t: 300.79k
7 threads: Avg: 0.0204ms Range: [6.8770us, 0.0235ms] Ops/s: 342.89k Ops/s/t: 48.98k
15 threads: Avg: 0.0959ms Range: [0.0412ms, 0.1095ms] Ops/s: 156.35k Ops/s/t: 10.42k
Operations per second per thread (weighted average): 4.00M
> LockBasedQueue
1 thread: Avg: 0.0616us Range: [0.0615us, 0.0616us] Ops/s: 16.24M Ops/s/t: 16.24M
3 threads: Avg: 2.7538us Range: [2.6907us, 2.8152us] Ops/s: 1.09M Ops/s/t: 363.14k
7 threads: Avg: 0.0215ms Range: [0.0190ms, 0.0225ms] Ops/s: 325.43k Ops/s/t: 46.49k
15 threads: Avg: 0.0868ms Range: [0.0813ms, 0.0919ms] Ops/s: 172.89k Ops/s/t: 11.53k
Operations per second per thread (weighted average): 1.84M
> std::queue
(skipping, benchmark not supported...)
multi-producer, single-consumer (measuring 1 thread):
(Measures the average speed of dequeueing with only one consumer, but multiple producers)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.0625us Range: [0.0624us, 0.0626us] Ops/s: 15.99M Ops/s/t: 15.99M
4 threads: Avg: 0.0639us Range: [0.0625us, 0.0644us] Ops/s: 15.66M Ops/s/t: 15.66M
8 threads: Avg: 0.0671us Range: [0.0628us, 0.0715us] Ops/s: 14.89M Ops/s/t: 14.89M
16 threads: Avg: 0.0916us Range: [0.0869us, 0.0946us] Ops/s: 10.92M Ops/s/t: 10.92M
Operations per second per thread (weighted average): 14.37M
With tokens
2 threads: Avg: 0.0556us Range: [0.0555us, 0.0557us] Ops/s: 17.99M Ops/s/t: 17.99M
4 threads: Avg: 0.0454us Range: [0.0452us, 0.0455us] Ops/s: 22.03M Ops/s/t: 22.03M
8 threads: Avg: 0.0410us Range: [0.0410us, 0.0411us] Ops/s: 24.37M Ops/s/t: 24.37M
16 threads: Avg: 0.0418us Range: [0.0412us, 0.0421us] Ops/s: 23.93M Ops/s/t: 23.93M
Operations per second per thread (weighted average): 22.08M
> boost::lockfree::queue
2 threads: Avg: 0.0444us Range: [0.0420us, 0.0516us] Ops/s: 22.53M Ops/s/t: 22.53M
4 threads: Avg: 0.2854us Range: [0.2487us, 0.4188us] Ops/s: 3.50M Ops/s/t: 3.50M
8 threads: Avg: 0.5135us Range: [0.5043us, 0.5177us] Ops/s: 1.95M Ops/s/t: 1.95M
16 threads: Avg: 0.5293us Range: [0.5162us, 0.5350us] Ops/s: 1.89M Ops/s/t: 1.89M
Operations per second per thread (weighted average): 7.47M
> tbb::concurrent_queue
2 threads: Avg: 0.1757us Range: [0.1333us, 0.2080us] Ops/s: 5.69M Ops/s/t: 5.69M
4 threads: Avg: 0.1487us Range: [0.1463us, 0.1507us] Ops/s: 6.72M Ops/s/t: 6.72M
8 threads: Avg: 0.1561us Range: [0.1542us, 0.1565us] Ops/s: 6.41M Ops/s/t: 6.41M
16 threads: Avg: 0.6278us Range: [0.5630us, 0.6839us] Ops/s: 1.59M Ops/s/t: 1.59M
Operations per second per thread (weighted average): 5.10M
> SimpleLockFreeQueue
2 threads: Avg: 0.0464us Range: [0.0392us, 0.0543us] Ops/s: 21.55M Ops/s/t: 21.55M
4 threads: Avg: 0.4172us Range: [0.3421us, 0.4385us] Ops/s: 2.40M Ops/s/t: 2.40M
8 threads: Avg: 0.2543us Range: [0.2397us, 0.2625us] Ops/s: 3.93M Ops/s/t: 3.93M
16 threads: Avg: 0.2392us Range: [0.2348us, 0.2422us] Ops/s: 4.18M Ops/s/t: 4.18M
Operations per second per thread (weighted average): 8.01M
> LockBasedQueue
2 threads: Avg: 0.1546us Range: [0.1125us, 0.1981us] Ops/s: 6.47M Ops/s/t: 6.47M
4 threads: Avg: 0.6662us Range: [0.5840us, 0.7603us] Ops/s: 1.50M Ops/s/t: 1.50M
8 threads: Avg: 0.7431us Range: [0.6830us, 0.7976us] Ops/s: 1.35M Ops/s/t: 1.35M
16 threads: Avg: 0.7052us Range: [0.6063us, 0.7579us] Ops/s: 1.42M Ops/s/t: 1.42M
Operations per second per thread (weighted average): 2.68M
> std::queue
(skipping, benchmark not supported...)
dequeue from empty:
(Measures the average speed of attempting to dequeue from an empty queue
(that eight separate threads had at one point enqueued to))
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0409us Range: [0.0406us, 0.0416us] Ops/s: 24.45M Ops/s/t: 24.45M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.0905us Range: [0.0825us, 0.0962us] Ops/s: 22.10M Ops/s/t: 11.05M
8 threads: Avg: 0.3867us Range: [0.3854us, 0.3899us] Ops/s: 20.69M Ops/s/t: 2.59M
Operations per second per thread (weighted average): 9.04M
With tokens
1 thread: Avg: 0.0237us Range: [0.0200us, 0.0249us] Ops/s: 42.21M Ops/s/t: 42.21M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.0524us Range: [0.0500us, 0.0539us] Ops/s: 38.19M Ops/s/t: 19.09M
8 threads: Avg: 0.2339us Range: [0.1980us, 0.2642us] Ops/s: 34.21M Ops/s/t: 4.28M
Operations per second per thread (weighted average): 15.51M
> boost::lockfree::queue
1 thread: Avg: 6.5831ns Range: [6.5815ns, 6.5847ns] Ops/s: 151.90M Ops/s/t: 151.90M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.0122us Range: [0.0122us, 0.0122us] Ops/s: 163.52M Ops/s/t: 81.76M
8 threads: Avg: 0.0488us Range: [0.0488us, 0.0489us] Ops/s: 163.91M Ops/s/t: 20.49M
Operations per second per thread (weighted average): 62.08M
> tbb::concurrent_queue
1 thread: Avg: 5.1712ns Range: [5.1697ns, 5.1721ns] Ops/s: 193.38M Ops/s/t: 193.38M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.0104us Range: [0.0103us, 0.0104us] Ops/s: 193.18M Ops/s/t: 96.59M
8 threads: Avg: 0.0413us Range: [0.0413us, 0.0413us] Ops/s: 193.79M Ops/s/t: 24.22M
Operations per second per thread (weighted average): 76.01M
> SimpleLockFreeQueue
1 thread: Avg: 7.0505ns Range: [7.0486ns, 7.0525ns] Ops/s: 141.83M Ops/s/t: 141.83M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.0160us Range: [0.0160us, 0.0160us] Ops/s: 125.02M Ops/s/t: 62.51M
8 threads: Avg: 0.0638us Range: [0.0638us, 0.0638us] Ops/s: 125.33M Ops/s/t: 15.67M
Operations per second per thread (weighted average): 52.37M
> LockBasedQueue
1 thread: Avg: 0.0291us Range: [0.0291us, 0.0291us] Ops/s: 34.32M Ops/s/t: 34.32M
^ Note: No contention -- measures raw failed dequeue speed on empty queue
2 threads: Avg: 0.2276us Range: [0.1948us, 0.2441us] Ops/s: 8.79M Ops/s/t: 4.39M
8 threads: Avg: 5.2936us Range: [2.8883us, 6.2920us] Ops/s: 1.51M Ops/s/t: 188.91k
Operations per second per thread (weighted average): 7.83M
> std::queue
1 thread: Avg: 0.9395ns Range: [0.9392ns, 0.9396ns] Ops/s: 1.06G Ops/s/t: 1.06G
^ Note: No contention -- measures raw failed dequeue speed on empty queue
Operations per second per thread (weighted average): 1.06G
enqueue-dequeue pairs:
(Measures the average operation speed with each thread doing an enqueue
followed by a dequeue)
> moodycamel::ConcurrentQueue
Without tokens
1 thread: Avg: 0.0292us Range: [0.0292us, 0.0292us] Ops/s: 34.23M Ops/s/t: 34.23M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.2037us Range: [0.1940us, 0.2153us] Ops/s: 9.82M Ops/s/t: 4.91M
4 threads: Avg: 1.7694us Range: [1.7308us, 1.7856us] Ops/s: 2.26M Ops/s/t: 565.17k
8 threads: Avg: 6.5195us Range: [6.3181us, 6.6066us] Ops/s: 1.23M Ops/s/t: 153.39k
Operations per second per thread (weighted average): 5.90M
With tokens
1 thread: Avg: 0.0226us Range: [0.0226us, 0.0226us] Ops/s: 44.19M Ops/s/t: 44.19M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.0487us Range: [0.0485us, 0.0488us] Ops/s: 41.10M Ops/s/t: 20.55M
4 threads: Avg: 0.1205us Range: [0.1195us, 0.1214us] Ops/s: 33.20M Ops/s/t: 8.30M
8 threads: Avg: 0.3268us Range: [0.3128us, 0.3348us] Ops/s: 24.48M Ops/s/t: 3.06M
Operations per second per thread (weighted average): 13.60M
> boost::lockfree::queue
1 thread: Avg: 0.0437us Range: [0.0436us, 0.0437us] Ops/s: 22.90M Ops/s/t: 22.90M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.3969us Range: [0.3245us, 0.4216us] Ops/s: 5.04M Ops/s/t: 2.52M
4 threads: Avg: 3.3639us Range: [3.2929us, 3.4045us] Ops/s: 1.19M Ops/s/t: 297.27k
8 threads: Avg: 0.0146ms Range: [0.0139ms, 0.0148ms] Ops/s: 548.28k Ops/s/t: 68.54k
Operations per second per thread (weighted average): 3.76M
> tbb::concurrent_queue
1 thread: Avg: 0.0288us Range: [0.0287us, 0.0289us] Ops/s: 34.72M Ops/s/t: 34.72M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.3518us Range: [0.3293us, 0.3811us] Ops/s: 5.69M Ops/s/t: 2.84M
4 threads: Avg: 2.4316us Range: [2.3749us, 2.5254us] Ops/s: 1.64M Ops/s/t: 411.24k
8 threads: Avg: 0.0119ms Range: [0.0109ms, 0.0127ms] Ops/s: 670.40k Ops/s/t: 83.80k
Operations per second per thread (weighted average): 5.49M
> SimpleLockFreeQueue
1 thread: Avg: 0.0477us Range: [0.0476us, 0.0477us] Ops/s: 20.99M Ops/s/t: 20.99M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.4059us Range: [0.3067us, 0.4286us] Ops/s: 4.93M Ops/s/t: 2.46M
4 threads: Avg: 3.6382us Range: [3.5906us, 3.6695us] Ops/s: 1.10M Ops/s/t: 274.86k
8 threads: Avg: 0.0152ms Range: [0.0140ms, 0.0157ms] Ops/s: 526.95k Ops/s/t: 65.87k
Operations per second per thread (weighted average): 3.48M
> LockBasedQueue
1 thread: Avg: 0.0646us Range: [0.0646us, 0.0646us] Ops/s: 15.48M Ops/s/t: 15.48M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
2 threads: Avg: 0.6880us Range: [0.6445us, 0.7316us] Ops/s: 2.91M Ops/s/t: 1.45M
4 threads: Avg: 6.0035us Range: [5.9227us, 6.0622us] Ops/s: 666.28k Ops/s/t: 166.57k
8 threads: Avg: 0.0200ms Range: [0.0106ms, 0.0247ms] Ops/s: 400.14k Ops/s/t: 50.02k
Operations per second per thread (weighted average): 2.49M
> std::queue
1 thread: Avg: 4.4183ns Range: [4.4094ns, 4.4228ns] Ops/s: 226.33M Ops/s/t: 226.33M
^ Note: No contention -- measures speed of immediately dequeueing the item that was just enqueued
Operations per second per thread (weighted average): 226.33M
heavy concurrent:
(Measures the average operation speed with many threads under heavy load)
> moodycamel::ConcurrentQueue
Without tokens
2 threads: Avg: 0.2189us Range: [0.2185us, 0.2194us] Ops/s: 9.14M Ops/s/t: 4.57M
3 threads: Avg: 0.3700us Range: [0.3134us, 0.4042us] Ops/s: 8.11M Ops/s/t: 2.70M
4 threads: Avg: 1.1026us Range: [1.0439us, 1.1272us] Ops/s: 3.63M Ops/s/t: 906.94k
8 threads: Avg: 3.1887us Range: [3.0584us, 3.2292us] Ops/s: 2.51M Ops/s/t: 313.61k
12 threads: Avg: 6.0866us Range: [5.9009us, 6.2746us] Ops/s: 1.97M Ops/s/t: 164.30k
16 threads: Avg: 9.9339us Range: [9.5659us, 0.0102ms] Ops/s: 1.61M Ops/s/t: 100.67k
Operations per second per thread (weighted average): 959.54k
With tokens
2 threads: Avg: 0.0502us Range: [0.0475us, 0.0523us] Ops/s: 39.86M Ops/s/t: 19.93M
3 threads: Avg: 0.1121us Range: [0.1088us, 0.1140us] Ops/s: 26.76M Ops/s/t: 8.92M
4 threads: Avg: 0.4054us Range: [0.3912us, 0.4140us] Ops/s: 9.87M Ops/s/t: 2.47M
8 threads: Avg: 0.7169us Range: [0.5758us, 0.8319us] Ops/s: 11.16M Ops/s/t: 1.39M
12 threads: Avg: 1.2992us Range: [1.0963us, 1.5459us] Ops/s: 9.24M Ops/s/t: 769.70k
16 threads: Avg: 1.8166us Range: [1.4060us, 1.9616us] Ops/s: 8.81M Ops/s/t: 550.47k
Operations per second per thread (weighted average): 3.72M
> boost::lockfree::queue
2 threads: Avg: 0.3558us Range: [0.3235us, 0.3937us] Ops/s: 5.62M Ops/s/t: 2.81M
3 threads: Avg: 1.7224us Range: [1.5520us, 1.8241us] Ops/s: 1.74M Ops/s/t: 580.60k
4 threads: Avg: 2.2439us Range: [1.8891us, 2.4835us] Ops/s: 1.78M Ops/s/t: 445.65k
8 threads: Avg: 8.8162us Range: [8.5410us, 9.0435us] Ops/s: 907.42k Ops/s/t: 113.43k
12 threads: Avg: 0.0184ms Range: [0.0179ms, 0.0187ms] Ops/s: 653.30k Ops/s/t: 54.44k
16 threads: Avg: 0.0289ms Range: [0.0236ms, 0.0301ms] Ops/s: 554.55k Ops/s/t: 34.66k
Operations per second per thread (weighted average): 422.31k
> tbb::concurrent_queue
2 threads: Avg: 0.3223us Range: [0.3021us, 0.3596us] Ops/s: 6.20M Ops/s/t: 3.10M
3 threads: Avg: 0.8995us Range: [0.8900us, 0.9129us] Ops/s: 3.34M Ops/s/t: 1.11M
4 threads: Avg: 1.7776us Range: [1.6889us, 1.8523us] Ops/s: 2.25M Ops/s/t: 562.55k
8 threads: Avg: 6.1223us Range: [5.5484us, 6.5143us] Ops/s: 1.31M Ops/s/t: 163.34k
12 threads: Avg: 0.0175ms Range: [0.0137ms, 0.0198ms] Ops/s: 686.07k Ops/s/t: 57.17k
16 threads: Avg: 0.0461ms Range: [0.0382ms, 0.0519ms] Ops/s: 347.11k Ops/s/t: 21.69k
Operations per second per thread (weighted average): 530.15k
> SimpleLockFreeQueue
2 threads: Avg: 0.4770us Range: [0.3797us, 0.5594us] Ops/s: 4.19M Ops/s/t: 2.10M
3 threads: Avg: 1.4174us Range: [1.1771us, 1.6083us] Ops/s: 2.12M Ops/s/t: 705.51k
4 threads: Avg: 3.1814us Range: [2.8453us, 3.4126us] Ops/s: 1.26M Ops/s/t: 314.33k
8 threads: Avg: 0.0115ms Range: [7.8236us, 0.0142ms] Ops/s: 693.66k Ops/s/t: 86.71k
12 threads: Avg: 0.0136ms Range: [9.8911us, 0.0154ms] Ops/s: 879.37k Ops/s/t: 73.28k
16 threads: Avg: 0.0194ms Range: [0.0178ms, 0.0210ms] Ops/s: 823.59k Ops/s/t: 51.47k
Operations per second per thread (weighted average): 357.56k
> LockBasedQueue
2 threads: Avg: 0.5884us Range: [0.5593us, 0.6079us] Ops/s: 3.40M Ops/s/t: 1.70M
3 threads: Avg: 2.7292us Range: [2.5104us, 2.8136us] Ops/s: 1.10M Ops/s/t: 366.41k
4 threads: Avg: 6.0285us Range: [5.7452us, 6.1933us] Ops/s: 663.52k Ops/s/t: 165.88k
8 threads: Avg: 0.0233ms Range: [8.6898us, 0.0286ms] Ops/s: 342.92k Ops/s/t: 42.87k
12 threads: Avg: 0.0594ms Range: [0.0545ms, 0.0609ms] Ops/s: 202.13k Ops/s/t: 16.84k
16 threads: Avg: 0.1016ms Range: [0.0870ms, 0.1082ms] Ops/s: 157.46k Ops/s/t: 9.84k
Operations per second per thread (weighted average): 232.46k
> std::queue
(skipping, benchmark not supported...)
Overall average operations per second per thread (where higher-concurrency runs have more weight):
(Take this summary with a grain of salt -- look at the individual benchmark results for a much
better idea of how the queues measure up to each other):
moodycamel::ConcurrentQueue (including bulk): 19.04M
boost::lockfree::queue: 4.07M
tbb::concurrent_queue: 5.09M
SimpleLockFreeQueue: 3.76M
LockBasedQueue: 1.36M
std::queue (single thread only): 297.43M
View Git Blame Copy Permalink