macOS Benchmarks

Generated 2026-10-02 from corosio e2de06fda069 — benchmark run. This page is fully generated; do not edit by hand (see the Benchmarks landing page).

Summary

Within noise means the relative difference is within twice the combined run-to-run noise (the root-sum-square of each side’s CV). Differences that small are indistinguishable from measurement jitter. See Methodology for how these figures are computed.

68 faster · 15 within noise · 30 slower of 113 benchmarks — median +58.8% vs Boost.Asio (callbacks) on the same reactor.

Summary — macOS ← slower · within noise · faster → per benchmark, vs Boost.Asio (callbacks) local_socket_throughput 24 benchmarks faster than asio (callbacks) 24 +467.2% local_socket_latency 12 benchmarks faster than asio (callbacks) 12 +95.7% socket_latency 2 benchmarks within noise than asio (callbacks) 10 benchmarks faster than asio (callbacks) 10 +66.3% socket_throughput 3 benchmarks slower than asio (callbacks) 3 7 benchmarks within noise than asio (callbacks) 7 17 benchmarks faster than asio (callbacks) 17 +63.3% http_server 6 benchmarks slower than asio (callbacks) 6 3 benchmarks within noise than asio (callbacks) 3 2 benchmarks faster than asio (callbacks) -4.5% fan_out 16 benchmarks slower than asio (callbacks) 16 2 benchmarks within noise than asio (callbacks) -8.7% accept_churn 5 benchmarks slower than asio (callbacks) 5 1 benchmark within noise than asio (callbacks) 3 benchmarks faster than asio (callbacks) 3 -10.2%

Test Environment

CPU

Apple M1

Cores

8

RAM (GB)

16

OS

Darwin 25.6.0

Kernel/build

Darwin Kernel Version 25.6.0: Tue Aug 18 17:48:35 PDT 2026; root:xnu-12377.161.15.700.19~2/RELEASE_ARM64_T8103

Compiler

Apple clang version 17.0.0 (clang-1700.6.3.2)

CMake

cmake version 4.4.3

liburing

n/a

Boost commit

39fa1e9a3491bd099b570935b3f3422065f91b03

Asio commit

a7dc25b4cb6c49a6946d86ea20664f1027203225

Asio reactor

kqueue

Capy commit

a372a6b054261f29497ac0d19c2a9533f9eaad40

Corosio commit

e2de06fda069607f34bc957cbefcbd9ba12fddca

Corosio branch

pr/benchmark-report

Date (UTC)

2026-10-02

Iterations

7

Duration per benchmark (s)

2.0

Results

accept_churn

Rate of setting up and tearing down short-lived TCP connections: connect, accept, close.

accept_churn — macOS ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -60% -40% -20% +20% 0% burst/10 — N connects are fired at once, then all N are accepted before closing; N is the burst size. burst/10 — N connects are fired at once, then all N are accepted before closing; N is the burst size. burst/10 burst/100 — N connects are fired at once, then all N are accepted before closing; N is the burst size. burst/100 — N connects are fired at once, then all N are accepted before closing; N is the burst size. -41% burst/100 burst_lockless/10 — Same as burst, with the context in single-threaded lockless mode; N is the burst size. burst_lockless/10 — Same as burst, with the context in single-threaded lockless mode; N is the burst size. burst_lockless/10 burst_lockless/100 — Same as burst, with the context in single-threaded lockless mode; N is the burst size. -28% burst_lockless/100 — Same as burst, with the context in single-threaded lockless mode; N is the burst size. burst_lockless/100 concurrent/1 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. concurrent/1 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. concurrent/1 concurrent/16 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. concurrent/16 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. -1% concurrent/16 concurrent/4 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. concurrent/4 — N independent accept loops run on separate listeners at once, each doing a one-way one-byte transfer (client writes, server reads) per connection; N is the number of loops. concurrent/4 sequential — Single connect/accept/close loop with a one-way one-byte transfer (client writes, server reads) per connection, one connection at a time. sequential — Single connect/accept/close loop with a one-way one-byte transfer (client writes, server reads) per connection, one connection at a time. sequential sequential_lockless — Same as sequential, with the context in single-threaded lockless mode. +10% sequential_lockless — Same as sequential, with the context in single-threaded lockless mode. sequential_lockless
Detailed results
Benchmark Implementation Median CV vs asio

burst/10

corosio (kqueue)

4.12K ops/s

4.19%

-10.2%

burst/10

asio (coroutines)

3.18K ops/s

1.54%

-30.6%

burst/10

asio (callbacks)

4.58K ops/s

2.22%

baseline

burst/100

corosio (kqueue)

426.7 ops/s

4.25%

-27.3%

burst/100

asio (coroutines)

343.8 ops/s

2.96%

-41.4%

burst/100

asio (callbacks)

587.1 ops/s

12.84%

baseline

burst_lockless/10

corosio (kqueue)

4.11K ops/s

3.92%

-10.5%

burst_lockless/10

asio (coroutines)

3.24K ops/s

1.91%

-29.4%

burst_lockless/10

asio (callbacks)

4.59K ops/s

9.95%

baseline

burst_lockless/100

corosio (kqueue)

424.0 ops/s

3.50%

-28.0%

burst_lockless/100

asio (coroutines)

347.1 ops/s

2.79%

-41.0%

burst_lockless/100

asio (callbacks)

588.5 ops/s

4.01%

baseline

concurrent/1

corosio (kqueue)

19.67K ops/s

0.52%

+9.7%

concurrent/1

asio (coroutines)

17.79K ops/s

1.75%

-0.7%

concurrent/1

asio (callbacks)

17.93K ops/s

0.74%

baseline

concurrent/16

corosio (kqueue)

34.00K ops/s

3.18%

-20.6%

concurrent/16

asio (coroutines)

42.53K ops/s

2.55%

-0.7%

concurrent/16

asio (callbacks)

42.81K ops/s

6.50%

baseline

concurrent/4

corosio (kqueue)

28.71K ops/s

2.41%

-8.2%

concurrent/4

asio (coroutines)

30.05K ops/s

1.48%

-3.9%

concurrent/4

asio (callbacks)

31.27K ops/s

3.26%

baseline

sequential

corosio (kqueue)

19.68K ops/s

0.50%

+9.1%

sequential

asio (coroutines)

17.81K ops/s

0.97%

-1.3%

sequential

asio (callbacks)

18.05K ops/s

0.54%

baseline

sequential_lockless

corosio (kqueue)

19.73K ops/s

1.83%

+9.8%

sequential_lockless

asio (coroutines)

17.79K ops/s

1.50%

-1.0%

sequential_lockless

asio (callbacks)

17.97K ops/s

0.50%

baseline

fan_out

Fan-out/fan-in coroutine coordination: a parent starts concurrent sub-requests against echo servers and awaits their completion via a shared latch.

fan_out — macOS, part 1 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -40% -20% +20% 0% concurrent_parents/1 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -5% concurrent_parents/1 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -11% concurrent_parents/1 concurrent_parents/16 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -7% concurrent_parents/16 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -30% concurrent_parents/16 concurrent_parents/4 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -8% concurrent_parents/4 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -5% concurrent_parents/4 concurrent_parents_lockless/1 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -5% concurrent_parents_lockless/1 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -11% concurrent_parents_lockless/1 concurrent_parents_lockless/16 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -10% concurrent_parents_lockless/16 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -31% concurrent_parents_lockless/16 concurrent_parents_lockless/4 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -11% concurrent_parents_lockless/4 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -4% concurrent_parents_lockless/4
Detailed results
Benchmark Implementation Median CV vs asio

concurrent_parents/1

corosio (kqueue)

7.72K ops/s

0.42%

-4.7%

concurrent_parents/1

asio (coroutines)

7.20K ops/s

0.22%

-11.1%

concurrent_parents/1

asio (callbacks)

8.10K ops/s

0.65%

baseline

concurrent_parents/16

corosio (kqueue)

8.45K ops/s

19.25%

-7.3%

concurrent_parents/16

asio (coroutines)

6.35K ops/s

19.09%

-30.3%

concurrent_parents/16

asio (callbacks)

9.11K ops/s

17.43%

baseline

concurrent_parents/4

corosio (kqueue)

8.25K ops/s

1.64%

-7.8%

concurrent_parents/4

asio (coroutines)

8.47K ops/s

0.83%

-5.4%

concurrent_parents/4

asio (callbacks)

8.95K ops/s

0.17%

baseline

concurrent_parents_lockless/1

corosio (kqueue)

7.76K ops/s

0.94%

-4.8%

concurrent_parents_lockless/1

asio (coroutines)

7.25K ops/s

0.21%

-11.1%

concurrent_parents_lockless/1

asio (callbacks)

8.16K ops/s

0.69%

baseline

concurrent_parents_lockless/16

corosio (kqueue)

8.29K ops/s

14.13%

-10.3%

concurrent_parents_lockless/16

asio (coroutines)

6.34K ops/s

18.94%

-31.4%

concurrent_parents_lockless/16

asio (callbacks)

9.24K ops/s

10.75%

baseline

concurrent_parents_lockless/4

corosio (kqueue)

8.10K ops/s

1.89%

-10.5%

concurrent_parents_lockless/4

asio (coroutines)

8.65K ops/s

0.83%

-4.4%

concurrent_parents_lockless/4

asio (callbacks)

9.05K ops/s

0.31%

baseline
fan_out — macOS, part 2 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -80% -60% -40% -20% +20% 0% fork_join/1 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/1 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/1 fork_join/16 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. -4% fork_join/16 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/16 fork_join/4 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/4 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/4 fork_join/64 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. fork_join/64 — One parent starts N sub-requests to echo servers and waits for all to finish before repeating; N is the fan-out width. -7% fork_join/64 fork_join_lockless/1 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. -27% fork_join_lockless/1 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. -53% fork_join_lockless/1 fork_join_lockless/16 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/16 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/16 fork_join_lockless/4 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/4 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/4 fork_join_lockless/64 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/64 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. fork_join_lockless/64 nested/16 — Two-level fan-out: the parent starts N groups of 4 sub-requests each, each group awaited via its own latch; N is the number of groups. nested/16 — Two-level fan-out: the parent starts N groups of 4 sub-requests each, each group awaited via its own latch; N is the number of groups. nested/16 nested/4 — Two-level fan-out: the parent starts N groups of 4 sub-requests each, each group awaited via its own latch; N is the number of groups. nested/4 — Two-level fan-out: the parent starts N groups of 4 sub-requests each, each group awaited via its own latch; N is the number of groups. nested/4 nested_lockless/16 — Same as nested, with the context in single-threaded lockless mode; N is the number of groups. nested_lockless/16 — Same as nested, with the context in single-threaded lockless mode; N is the number of groups. nested_lockless/16 nested_lockless/4 — Same as nested, with the context in single-threaded lockless mode; N is the number of groups. nested_lockless/4 — Same as nested, with the context in single-threaded lockless mode; N is the number of groups. nested_lockless/4
Detailed results
Benchmark Implementation Median CV vs asio

fork_join/1

corosio (kqueue)

46.39K ops/s

0.98%

-23.4%

fork_join/1

asio (coroutines)

29.26K ops/s

1.31%

-51.7%

fork_join/1

asio (callbacks)

60.56K ops/s

1.56%

baseline

fork_join/16

corosio (kqueue)

7.70K ops/s

0.41%

-4.1%

fork_join/16

asio (coroutines)

7.21K ops/s

0.18%

-10.2%

fork_join/16

asio (callbacks)

8.03K ops/s

0.70%

baseline

fork_join/4

corosio (kqueue)

21.70K ops/s

0.41%

-6.7%

fork_join/4

asio (coroutines)

16.80K ops/s

0.33%

-27.8%

fork_join/4

asio (callbacks)

23.27K ops/s

0.31%

baseline

fork_join/64

corosio (kqueue)

1.99K ops/s

2.25%

-11.2%

fork_join/64

asio (coroutines)

2.08K ops/s

0.97%

-7.4%

fork_join/64

asio (callbacks)

2.25K ops/s

0.16%

baseline

fork_join_lockless/1

corosio (kqueue)

46.88K ops/s

0.48%

-26.6%

fork_join_lockless/1

asio (coroutines)

29.74K ops/s

0.48%

-53.4%

fork_join_lockless/1

asio (callbacks)

63.85K ops/s

0.50%

baseline

fork_join_lockless/16

corosio (kqueue)

7.80K ops/s

0.71%

-4.3%

fork_join_lockless/16

asio (coroutines)

7.27K ops/s

0.24%

-10.8%

fork_join_lockless/16

asio (callbacks)

8.15K ops/s

0.69%

baseline

fork_join_lockless/4

corosio (kqueue)

21.77K ops/s

0.50%

-7.2%

fork_join_lockless/4

asio (coroutines)

16.89K ops/s

0.22%

-28.0%

fork_join_lockless/4

asio (callbacks)

23.47K ops/s

0.19%

baseline

fork_join_lockless/64

corosio (kqueue)

2.08K ops/s

2.38%

-8.5%

fork_join_lockless/64

asio (coroutines)

2.10K ops/s

1.02%

-7.4%

fork_join_lockless/64

asio (callbacks)

2.27K ops/s

0.13%

baseline

nested/16

corosio (kqueue)

2.04K ops/s

1.78%

-8.9%

nested/16

asio (coroutines)

2.01K ops/s

0.50%

-10.2%

nested/16

asio (callbacks)

2.24K ops/s

0.29%

baseline

nested/4

corosio (kqueue)

7.07K ops/s

0.18%

-12.5%

nested/4

asio (coroutines)

6.49K ops/s

0.25%

-19.7%

nested/4

asio (callbacks)

8.08K ops/s

0.84%

baseline

nested_lockless/16

corosio (kqueue)

2.00K ops/s

2.14%

-11.6%

nested_lockless/16

asio (coroutines)

2.03K ops/s

0.73%

-10.4%

nested_lockless/16

asio (callbacks)

2.26K ops/s

0.25%

baseline

nested_lockless/4

corosio (kqueue)

7.19K ops/s

0.41%

-11.8%

nested_lockless/4

asio (coroutines)

6.56K ops/s

0.32%

-19.5%

nested_lockless/4

asio (callbacks)

8.15K ops/s

0.69%

baseline

http_server

Request/response throughput of a minimal HTTP/1.1 server exchanging a fixed small request and canned response over persistent TCP loopback connections.

http_server — macOS ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -40% -20% +20% 0% concurrent/1 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/1 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/1 concurrent/16 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/16 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. 0% concurrent/16 concurrent/32 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/32 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/32 concurrent/4 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/4 — N client/server pairs run the single_conn request/response loop concurrently; N is the number of connections. concurrent/4 multithread/1 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/1 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/1 multithread/16 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/16 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/16 multithread/2 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/2 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/2 multithread/4 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. +12% multithread/4 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/4 multithread/8 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/8 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. multithread/8 single_conn — One client repeatedly sends a fixed small HTTP request to one server and reads the response. single_conn — One client repeatedly sends a fixed small HTTP request to one server and reads the response. -6% single_conn single_conn_lockless — Same as single_conn, with the context in single-threaded lockless mode. -21% single_conn_lockless — Same as single_conn, with the context in single-threaded lockless mode. single_conn_lockless
Detailed results
Benchmark Implementation Median CV vs asio

concurrent/1

corosio (kqueue)

49.55K ops/s

0.29%

-17.9%

concurrent/1

asio (coroutines)

57.16K ops/s

1.17%

-5.3%

concurrent/1

asio (callbacks)

60.35K ops/s

0.94%

baseline

concurrent/16

corosio (kqueue)

126.9K ops/s

0.37%

-0.2%

concurrent/16

asio (coroutines)

126.8K ops/s

0.63%

-0.3%

concurrent/16

asio (callbacks)

127.2K ops/s

0.41%

baseline

concurrent/32

corosio (kqueue)

127.9K ops/s

0.92%

-8.4%

concurrent/32

asio (coroutines)

135.2K ops/s

0.21%

-3.1%

concurrent/32

asio (callbacks)

139.6K ops/s

0.44%

baseline

concurrent/4

corosio (kqueue)

88.07K ops/s

0.45%

-4.5%

concurrent/4

asio (coroutines)

91.01K ops/s

0.16%

-1.3%

concurrent/4

asio (callbacks)

92.17K ops/s

0.35%

baseline

multithread/1

corosio (kqueue)

127.5K ops/s

0.90%

-8.6%

multithread/1

asio (coroutines)

134.9K ops/s

0.18%

-3.3%

multithread/1

asio (callbacks)

139.5K ops/s

0.43%

baseline

multithread/16

corosio (kqueue)

126.8K ops/s

0.72%

+9.8%

multithread/16

asio (coroutines)

114.2K ops/s

0.49%

-1.1%

multithread/16

asio (callbacks)

115.5K ops/s

1.09%

baseline

multithread/2

corosio (kqueue)

174.1K ops/s

1.89%

+2.7%

multithread/2

asio (coroutines)

167.9K ops/s

0.20%

-1.0%

multithread/2

asio (callbacks)

169.5K ops/s

0.13%

baseline

multithread/4

corosio (kqueue)

162.0K ops/s

1.02%

+11.7%

multithread/4

asio (coroutines)

143.3K ops/s

0.19%

-1.2%

multithread/4

asio (callbacks)

145.0K ops/s

0.21%

baseline

multithread/8

corosio (kqueue)

132.4K ops/s

1.15%

+0.1%

multithread/8

asio (coroutines)

128.8K ops/s

0.57%

-2.6%

multithread/8

asio (callbacks)

132.2K ops/s

0.55%

baseline

single_conn

corosio (kqueue)

49.49K ops/s

0.20%

-18.2%

single_conn

asio (coroutines)

56.88K ops/s

0.89%

-6.0%

single_conn

asio (callbacks)

60.49K ops/s

0.80%

baseline

single_conn_lockless

corosio (kqueue)

49.61K ops/s

0.34%

-21.0%

single_conn_lockless

asio (coroutines)

60.26K ops/s

1.28%

-4.1%

single_conn_lockless

asio (callbacks)

62.82K ops/s

0.70%

baseline

local_socket_latency

Round-trip latency of a single Unix domain stream socket write/read exchange across message sizes and concurrent pair counts.

local_socket_latency — macOS ↑ faster (lower mean latency) · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -50% +50% +100% +150% 0% concurrent/1 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/1 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/1 concurrent/16 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/16 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/16 concurrent/4 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/4 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/4 concurrent_lockless/1 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/1 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. +0% concurrent_lockless/1 concurrent_lockless/16 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. +59% concurrent_lockless/16 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. -4% concurrent_lockless/16 concurrent_lockless/4 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/4 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/4 pingpong/1 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1 pingpong/1024 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1024 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1024 pingpong/64 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/64 — One connected Unix domain socket pair ping-pongs a message client->server->client; N is the message size in bytes. pingpong/64 pingpong_lockless/1 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. +96% pingpong_lockless/1 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/1 pingpong_lockless/1024 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/1024 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/1024 pingpong_lockless/64 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/64 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/64
Detailed results
Benchmark Implementation Median CV vs asio

concurrent/1

corosio (kqueue)

2.06 µs

0.14%

+95.8%

concurrent/1

asio (coroutines)

48.60 µs

0.08%

+0.0%

concurrent/1

asio (callbacks)

48.62 µs

0.09%

baseline

concurrent/16

corosio (kqueue)

30.67 µs

0.18%

+58.8%

concurrent/16

asio (coroutines)

77.22 µs

0.51%

-3.6%

concurrent/16

asio (callbacks)

74.53 µs

0.68%

baseline

concurrent/4

corosio (kqueue)

7.69 µs

0.28%

+86.1%

concurrent/4

asio (coroutines)

55.63 µs

0.21%

-0.5%

concurrent/4

asio (callbacks)

55.38 µs

0.19%

baseline

concurrent_lockless/1

corosio (kqueue)

2.03 µs

0.17%

+95.8%

concurrent_lockless/1

asio (coroutines)

48.60 µs

0.13%

+0.0%

concurrent_lockless/1

asio (callbacks)

48.61 µs

0.04%

baseline

concurrent_lockless/16

corosio (kqueue)

30.25 µs

0.31%

+58.6%

concurrent_lockless/16

asio (coroutines)

75.99 µs

0.92%

-4.1%

concurrent_lockless/16

asio (callbacks)

73.02 µs

0.32%

baseline

concurrent_lockless/4

corosio (kqueue)

7.54 µs

0.14%

+86.4%

concurrent_lockless/4

asio (coroutines)

55.41 µs

0.25%

-0.1%

concurrent_lockless/4

asio (callbacks)

55.33 µs

0.49%

baseline

pingpong/1

corosio (kqueue)

2.06 µs

0.19%

+95.8%

pingpong/1

asio (coroutines)

48.62 µs

0.03%

-0.0%

pingpong/1

asio (callbacks)

48.61 µs

0.10%

baseline

pingpong/1024

corosio (kqueue)

2.16 µs

0.24%

+95.6%

pingpong/1024

asio (coroutines)

48.64 µs

0.15%

-0.1%

pingpong/1024

asio (callbacks)

48.60 µs

0.12%

baseline

pingpong/64

corosio (kqueue)

2.06 µs

0.36%

+95.8%

pingpong/64

asio (coroutines)

48.58 µs

0.03%

+0.0%

pingpong/64

asio (callbacks)

48.59 µs

0.14%

baseline

pingpong_lockless/1

corosio (kqueue)

2.03 µs

0.15%

+95.8%

pingpong_lockless/1

asio (coroutines)

48.60 µs

0.06%

-0.0%

pingpong_lockless/1

asio (callbacks)

48.58 µs

0.07%

baseline

pingpong_lockless/1024

corosio (kqueue)

2.13 µs

0.28%

+95.6%

pingpong_lockless/1024

asio (coroutines)

48.60 µs

0.06%

+0.0%

pingpong_lockless/1024

asio (callbacks)

48.61 µs

0.08%

baseline

pingpong_lockless/64

corosio (kqueue)

2.03 µs

0.27%

+95.8%

pingpong_lockless/64

asio (coroutines)

48.58 µs

0.04%

+0.0%

pingpong_lockless/64

asio (callbacks)

48.58 µs

0.04%

baseline

local_socket_throughput

Sustained byte throughput of a Unix domain stream socket pair under continuous streaming across chunk sizes and directions.

local_socket_throughput — macOS, part 1 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -250% +250% +500% +750% +1000% 0% bidirectional/1024 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/1024 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/1024 bidirectional/1048576 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/1048576 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/1048576 bidirectional/16384 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/16384 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/16384 bidirectional/262144 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/262144 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/262144 bidirectional/4096 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/4096 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/4096 bidirectional/65536 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/65536 — Both ends of the socket pair write and read simultaneously; N is the chunk size in bytes. bidirectional/65536 bidirectional_lockless/1024 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +709% bidirectional_lockless/1024 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. 0% bidirectional_lockless/1024 bidirectional_lockless/1048576 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/1048576 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. -3% bidirectional_lockless/1048576 bidirectional_lockless/16384 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +241% bidirectional_lockless/16384 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/16384 bidirectional_lockless/262144 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/262144 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/262144 bidirectional_lockless/4096 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/4096 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/4096 bidirectional_lockless/65536 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/65536 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/65536
Detailed results
Benchmark Implementation Median CV vs asio

bidirectional/1024

corosio (kqueue)

1.22 GB/s

0.26%

+703.2%

bidirectional/1024

asio (coroutines)

151.7 MB/s

0.18%

-0.1%

bidirectional/1024

asio (callbacks)

151.9 MB/s

0.04%

baseline

bidirectional/1048576

corosio (kqueue)

3.85 GB/s

0.45%

+246.9%

bidirectional/1048576

asio (coroutines)

1.11 GB/s

0.05%

-0.3%

bidirectional/1048576

asio (callbacks)

1.11 GB/s

0.07%

baseline

bidirectional/16384

corosio (kqueue)

3.85 GB/s

0.63%

+246.6%

bidirectional/16384

asio (coroutines)

1.11 GB/s

0.05%

-0.2%

bidirectional/16384

asio (callbacks)

1.11 GB/s

0.38%

baseline

bidirectional/262144

corosio (kqueue)

3.86 GB/s

0.72%

+247.3%

bidirectional/262144

asio (coroutines)

1.11 GB/s

0.02%

-0.3%

bidirectional/262144

asio (callbacks)

1.11 GB/s

0.23%

baseline

bidirectional/4096

corosio (kqueue)

2.90 GB/s

0.37%

+377.7%

bidirectional/4096

asio (coroutines)

605.1 MB/s

0.30%

-0.3%

bidirectional/4096

asio (callbacks)

606.7 MB/s

0.13%

baseline

bidirectional/65536

corosio (kqueue)

3.85 GB/s

0.32%

+247.0%

bidirectional/65536

asio (coroutines)

1.11 GB/s

0.03%

-0.3%

bidirectional/65536

asio (callbacks)

1.11 GB/s

0.31%

baseline

bidirectional_lockless/1024

corosio (kqueue)

1.23 GB/s

0.14%

+708.6%

bidirectional_lockless/1024

asio (coroutines)

151.8 MB/s

0.05%

-0.1%

bidirectional_lockless/1024

asio (callbacks)

151.9 MB/s

0.04%

baseline

bidirectional_lockless/1048576

corosio (kqueue)

3.92 GB/s

0.32%

+242.0%

bidirectional_lockless/1048576

asio (coroutines)

1.11 GB/s

0.03%

-3.3%

bidirectional_lockless/1048576

asio (callbacks)

1.14 GB/s

1.59%

baseline

bidirectional_lockless/16384

corosio (kqueue)

3.90 GB/s

0.54%

+240.9%

bidirectional_lockless/16384

asio (coroutines)

1.11 GB/s

0.03%

-3.1%

bidirectional_lockless/16384

asio (callbacks)

1.14 GB/s

1.35%

baseline

bidirectional_lockless/262144

corosio (kqueue)

3.90 GB/s

0.43%

+241.4%

bidirectional_lockless/262144

asio (coroutines)

1.11 GB/s

0.02%

-3.1%

bidirectional_lockless/262144

asio (callbacks)

1.14 GB/s

1.55%

baseline

bidirectional_lockless/4096

corosio (kqueue)

2.95 GB/s

0.38%

+384.9%

bidirectional_lockless/4096

asio (coroutines)

605.8 MB/s

0.16%

-0.3%

bidirectional_lockless/4096

asio (callbacks)

607.4 MB/s

0.05%

baseline

bidirectional_lockless/65536

corosio (kqueue)

3.89 GB/s

0.47%

+242.3%

bidirectional_lockless/65536

asio (coroutines)

1.11 GB/s

0.02%

-2.6%

bidirectional_lockless/65536

asio (callbacks)

1.14 GB/s

1.77%

baseline
local_socket_throughput — macOS, part 2 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -500% +500% +1000% +1500% +2000% 0% unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. +1402% unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1024 unidirectional/1048576 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1048576 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1048576 unidirectional/16384 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/16384 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/16384 unidirectional/262144 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/262144 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. 0% unidirectional/262144 unidirectional/4096 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/4096 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/4096 unidirectional/65536 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. +466% unidirectional/65536 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/65536 unidirectional_lockless/1024 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/1024 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. -7% unidirectional_lockless/1024 unidirectional_lockless/1048576 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/1048576 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/1048576 unidirectional_lockless/16384 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/16384 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/16384 unidirectional_lockless/262144 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/262144 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/262144 unidirectional_lockless/4096 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/4096 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/4096 unidirectional_lockless/65536 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/65536 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/65536
Detailed results
Benchmark Implementation Median CV vs asio

unidirectional/1024

corosio (kqueue)

1.19 GB/s

0.36%

+1402.1%

unidirectional/1024

asio (coroutines)

76.19 MB/s

0.04%

-4.0%

unidirectional/1024

asio (callbacks)

79.38 MB/s

0.75%

baseline

unidirectional/1048576

corosio (kqueue)

3.46 GB/s

0.31%

+467.8%

unidirectional/1048576

asio (coroutines)

608.5 MB/s

0.08%

-0.1%

unidirectional/1048576

asio (callbacks)

608.9 MB/s

0.18%

baseline

unidirectional/16384

corosio (kqueue)

3.46 GB/s

0.38%

+467.9%

unidirectional/16384

asio (coroutines)

608.6 MB/s

0.03%

-0.0%

unidirectional/16384

asio (callbacks)

608.7 MB/s

0.03%

baseline

unidirectional/262144

corosio (kqueue)

3.45 GB/s

0.25%

+466.7%

unidirectional/262144

asio (coroutines)

608.6 MB/s

0.03%

-0.0%

unidirectional/262144

asio (callbacks)

608.8 MB/s

0.03%

baseline

unidirectional/4096

corosio (kqueue)

2.68 GB/s

0.21%

+778.7%

unidirectional/4096

asio (coroutines)

304.5 MB/s

0.08%

-0.1%

unidirectional/4096

asio (callbacks)

304.8 MB/s

0.02%

baseline

unidirectional/65536

corosio (kqueue)

3.44 GB/s

0.42%

+465.7%

unidirectional/65536

asio (coroutines)

608.7 MB/s

0.02%

-0.0%

unidirectional/65536

asio (callbacks)

608.8 MB/s

0.09%

baseline

unidirectional_lockless/1024

corosio (kqueue)

1.20 GB/s

0.26%

+1335.5%

unidirectional_lockless/1024

asio (coroutines)

77.56 MB/s

0.32%

-7.0%

unidirectional_lockless/1024

asio (callbacks)

83.40 MB/s

0.26%

baseline

unidirectional_lockless/1048576

corosio (kqueue)

3.50 GB/s

0.43%

+474.9%

unidirectional_lockless/1048576

asio (coroutines)

608.6 MB/s

0.04%

-0.1%

unidirectional_lockless/1048576

asio (callbacks)

609.0 MB/s

0.03%

baseline

unidirectional_lockless/16384

corosio (kqueue)

3.50 GB/s

0.34%

+475.4%

unidirectional_lockless/16384

asio (coroutines)

608.7 MB/s

0.01%

-0.0%

unidirectional_lockless/16384

asio (callbacks)

608.9 MB/s

0.03%

baseline

unidirectional_lockless/262144

corosio (kqueue)

3.50 GB/s

0.34%

+475.1%

unidirectional_lockless/262144

asio (coroutines)

608.6 MB/s

0.05%

-0.1%

unidirectional_lockless/262144

asio (callbacks)

608.9 MB/s

0.02%

baseline

unidirectional_lockless/4096

corosio (kqueue)

2.71 GB/s

0.22%

+776.8%

unidirectional_lockless/4096

asio (coroutines)

304.6 MB/s

0.03%

-1.6%

unidirectional_lockless/4096

asio (callbacks)

309.5 MB/s

0.58%

baseline

unidirectional_lockless/65536

corosio (kqueue)

3.50 GB/s

0.42%

+474.0%

unidirectional_lockless/65536

asio (coroutines)

608.7 MB/s

0.02%

-0.0%

unidirectional_lockless/65536

asio (callbacks)

608.9 MB/s

0.03%

baseline

socket_latency

Round-trip latency of a TCP loopback connection. Each sample is one full round trip (request out, reply back) across message sizes and concurrent pair counts.

socket_latency — macOS ↑ faster (lower mean latency) · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -50% +50% +100% 0% concurrent/1 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/1 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/1 concurrent/16 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/16 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/16 concurrent/4 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/4 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. concurrent/4 concurrent_lockless/1 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/1 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/1 concurrent_lockless/16 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. +8% concurrent_lockless/16 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/16 concurrent_lockless/4 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/4 — Same as concurrent, with the context in single-threaded lockless mode; N is the number of connection pairs. concurrent_lockless/4 pingpong/1 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1 pingpong/1024 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. pingpong/1024 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. +5% pingpong/1024 pingpong/64 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. pingpong/64 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. pingpong/64 pingpong_lockless/1 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. +70% pingpong_lockless/1 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. -8% pingpong_lockless/1 pingpong_lockless/1024 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/1024 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/1024 pingpong_lockless/64 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/64 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. pingpong_lockless/64
Detailed results
Benchmark Implementation Median CV vs asio

concurrent/1

corosio (kqueue)

18.82 µs

0.21%

+64.8%

concurrent/1

asio (coroutines)

55.96 µs

2.40%

-4.6%

concurrent/1

asio (callbacks)

53.52 µs

4.64%

baseline

concurrent/16

corosio (kqueue)

120.69 µs

0.53%

+9.5%

concurrent/16

asio (coroutines)

142.60 µs

5.34%

-6.9%

concurrent/16

asio (callbacks)

133.43 µs

5.65%

baseline

concurrent/4

corosio (kqueue)

44.09 µs

0.77%

+40.3%

concurrent/4

asio (coroutines)

74.88 µs

7.32%

-1.3%

concurrent/4

asio (callbacks)

73.91 µs

0.80%

baseline

concurrent_lockless/1

corosio (kqueue)

18.85 µs

0.49%

+66.7%

concurrent_lockless/1

asio (coroutines)

55.13 µs

3.89%

+2.6%

concurrent_lockless/1

asio (callbacks)

56.61 µs

6.54%

baseline

concurrent_lockless/16

corosio (kqueue)

120.26 µs

0.39%

+8.3%

concurrent_lockless/16

asio (coroutines)

136.84 µs

5.71%

-4.3%

concurrent_lockless/16

asio (callbacks)

131.16 µs

7.19%

baseline

concurrent_lockless/4

corosio (kqueue)

43.39 µs

0.67%

+40.9%

concurrent_lockless/4

asio (coroutines)

74.37 µs

13.79%

-1.2%

concurrent_lockless/4

asio (callbacks)

73.46 µs

0.93%

baseline

pingpong/1

corosio (kqueue)

15.56 µs

0.60%

+68.5%

pingpong/1

asio (coroutines)

50.73 µs

1.93%

-2.6%

pingpong/1

asio (callbacks)

49.42 µs

2.41%

baseline

pingpong/1024

corosio (kqueue)

19.02 µs

0.25%

+66.2%

pingpong/1024

asio (coroutines)

53.34 µs

6.26%

+5.1%

pingpong/1024

asio (callbacks)

56.18 µs

5.76%

baseline

pingpong/64

corosio (kqueue)

18.81 µs

0.51%

+66.8%

pingpong/64

asio (coroutines)

56.64 µs

3.02%

-0.0%

pingpong/64

asio (callbacks)

56.62 µs

4.45%

baseline

pingpong_lockless/1

corosio (kqueue)

15.47 µs

0.43%

+69.7%

pingpong_lockless/1

asio (coroutines)

55.05 µs

5.28%

-7.8%

pingpong_lockless/1

asio (callbacks)

51.09 µs

5.04%

baseline

pingpong_lockless/1024

corosio (kqueue)

19.03 µs

0.36%

+66.4%

pingpong_lockless/1024

asio (coroutines)

56.89 µs

2.43%

-0.4%

pingpong_lockless/1024

asio (callbacks)

56.67 µs

2.05%

baseline

pingpong_lockless/64

corosio (kqueue)

18.85 µs

0.41%

+66.8%

pingpong_lockless/64

asio (coroutines)

56.61 µs

3.37%

+0.4%

pingpong_lockless/64

asio (callbacks)

56.84 µs

4.40%

baseline

socket_throughput

Sustained byte throughput of a TCP loopback connection under continuous streaming, varying chunk size, direction, and concurrency.

socket_throughput — macOS, part 1 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -50% +50% +100% +150% 0% bidirectional/1024 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/1024 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/1024 bidirectional/1048576 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. -12% bidirectional/1048576 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/1048576 bidirectional/16384 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/16384 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/16384 bidirectional/262144 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/262144 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/262144 bidirectional/4096 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/4096 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/4096 bidirectional/65536 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/65536 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. bidirectional/65536 bidirectional_lockless/1024 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +104% bidirectional_lockless/1024 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/1024 bidirectional_lockless/1048576 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/1048576 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/1048576 bidirectional_lockless/16384 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/16384 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/16384 bidirectional_lockless/262144 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/262144 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. -20% bidirectional_lockless/262144 bidirectional_lockless/4096 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/4096 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/4096 bidirectional_lockless/65536 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. bidirectional_lockless/65536 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. 0% bidirectional_lockless/65536
Detailed results
Benchmark Implementation Median CV vs asio

bidirectional/1024

corosio (kqueue)

410.3 MB/s

0.65%

+100.0%

bidirectional/1024

asio (coroutines)

198.8 MB/s

0.87%

-3.1%

bidirectional/1024

asio (callbacks)

205.1 MB/s

0.67%

baseline

bidirectional/1048576

corosio (kqueue)

7.75 GB/s

0.71%

-11.6%

bidirectional/1048576

asio (coroutines)

8.57 GB/s

1.62%

-2.2%

bidirectional/1048576

asio (callbacks)

8.77 GB/s

2.18%

baseline

bidirectional/16384

corosio (kqueue)

4.01 GB/s

0.58%

+64.9%

bidirectional/16384

asio (coroutines)

2.39 GB/s

0.37%

-1.8%

bidirectional/16384

asio (callbacks)

2.43 GB/s

0.31%

baseline

bidirectional/262144

corosio (kqueue)

9.02 GB/s

1.59%

-1.9%

bidirectional/262144

asio (coroutines)

7.37 GB/s

8.49%

-19.9%

bidirectional/262144

asio (callbacks)

9.20 GB/s

0.56%

baseline

bidirectional/4096

corosio (kqueue)

1.49 GB/s

0.30%

+88.6%

bidirectional/4096

asio (coroutines)

756.1 MB/s

0.87%

-4.2%

bidirectional/4096

asio (callbacks)

789.4 MB/s

0.68%

baseline

bidirectional/65536

corosio (kqueue)

7.59 GB/s

1.46%

+32.1%

bidirectional/65536

asio (coroutines)

5.74 GB/s

1.55%

-0.2%

bidirectional/65536

asio (callbacks)

5.75 GB/s

0.86%

baseline

bidirectional_lockless/1024

corosio (kqueue)

411.7 MB/s

0.57%

+104.4%

bidirectional_lockless/1024

asio (coroutines)

201.3 MB/s

0.67%

-0.1%

bidirectional_lockless/1024

asio (callbacks)

201.4 MB/s

1.34%

baseline

bidirectional_lockless/1048576

corosio (kqueue)

7.76 GB/s

2.00%

-11.3%

bidirectional_lockless/1048576

asio (coroutines)

8.72 GB/s

2.31%

-0.3%

bidirectional_lockless/1048576

asio (callbacks)

8.75 GB/s

2.38%

baseline

bidirectional_lockless/16384

corosio (kqueue)

4.00 GB/s

0.47%

+63.3%

bidirectional_lockless/16384

asio (coroutines)

2.40 GB/s

0.51%

-1.9%

bidirectional_lockless/16384

asio (callbacks)

2.45 GB/s

0.22%

baseline

bidirectional_lockless/262144

corosio (kqueue)

8.95 GB/s

1.93%

-3.8%

bidirectional_lockless/262144

asio (coroutines)

7.43 GB/s

10.71%

-20.2%

bidirectional_lockless/262144

asio (callbacks)

9.31 GB/s

7.51%

baseline

bidirectional_lockless/4096

corosio (kqueue)

1.50 GB/s

0.51%

+86.4%

bidirectional_lockless/4096

asio (coroutines)

773.4 MB/s

0.50%

-3.6%

bidirectional_lockless/4096

asio (callbacks)

802.7 MB/s

1.08%

baseline

bidirectional_lockless/65536

corosio (kqueue)

7.82 GB/s

1.25%

+36.1%

bidirectional_lockless/65536

asio (coroutines)

5.74 GB/s

1.34%

-0.0%

bidirectional_lockless/65536

asio (callbacks)

5.75 GB/s

1.66%

baseline
socket_throughput — macOS, part 2 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -20% -10% +10% 0% multithread/2 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. +2% multithread/2 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. +2% multithread/2 multithread/4 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. +1% multithread/4 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. +5% multithread/4 multithread/8 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. -11% multithread/8 — 32 bidirectional connection pairs share one context serviced by N threads running the context, with 64 KiB chunks; N is the thread count. +0% multithread/8
Detailed results
Benchmark Implementation Median CV vs asio

multithread/2

corosio (kqueue)

7.75 GB/s

1.43%

+2.4%

multithread/2

asio (coroutines)

7.76 GB/s

1.90%

+2.4%

multithread/2

asio (callbacks)

7.57 GB/s

2.22%

baseline

multithread/4

corosio (kqueue)

3.92 GB/s

2.78%

+0.6%

multithread/4

asio (coroutines)

4.08 GB/s

1.87%

+4.8%

multithread/4

asio (callbacks)

3.90 GB/s

22.01%

baseline

multithread/8

corosio (kqueue)

3.48 GB/s

3.73%

-10.8%

multithread/8

asio (coroutines)

3.91 GB/s

3.00%

+0.2%

multithread/8

asio (callbacks)

3.90 GB/s

2.02%

baseline
socket_throughput — macOS, part 3 ↑ faster · dotted line = asio (callbacks) · backend = kqueue corosio asio (coroutines) -50% +50% +100% +150% 0% unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1024 unidirectional/1048576 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. 0% unidirectional/1048576 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/1048576 unidirectional/16384 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/16384 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/16384 unidirectional/262144 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/262144 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. -23% unidirectional/262144 unidirectional/4096 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/4096 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/4096 unidirectional/65536 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/65536 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. unidirectional/65536 unidirectional_lockless/1024 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +125% unidirectional_lockless/1024 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +10% unidirectional_lockless/1024 unidirectional_lockless/1048576 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/1048576 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/1048576 unidirectional_lockless/16384 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/16384 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/16384 unidirectional_lockless/262144 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/262144 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/262144 unidirectional_lockless/4096 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/4096 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/4096 unidirectional_lockless/65536 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/65536 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. unidirectional_lockless/65536
Detailed results
Benchmark Implementation Median CV vs asio

unidirectional/1024

corosio (kqueue)

321.5 MB/s

1.01%

+94.3%

unidirectional/1024

asio (coroutines)

157.8 MB/s

0.52%

-4.7%

unidirectional/1024

asio (callbacks)

165.5 MB/s

2.47%

baseline

unidirectional/1048576

corosio (kqueue)

9.05 GB/s

1.04%

-0.1%

unidirectional/1048576

asio (coroutines)

8.34 GB/s

5.41%

-8.0%

unidirectional/1048576

asio (callbacks)

9.07 GB/s

1.31%

baseline

unidirectional/16384

corosio (kqueue)

3.75 GB/s

1.25%

+96.0%

unidirectional/16384

asio (coroutines)

1.90 GB/s

0.43%

-0.8%

unidirectional/16384

asio (callbacks)

1.92 GB/s

1.18%

baseline

unidirectional/262144

corosio (kqueue)

8.96 GB/s

2.20%

+14.0%

unidirectional/262144

asio (coroutines)

6.07 GB/s

8.74%

-22.7%

unidirectional/262144

asio (callbacks)

7.86 GB/s

2.31%

baseline

unidirectional/4096

corosio (kqueue)

1.24 GB/s

1.29%

+97.7%

unidirectional/4096

asio (coroutines)

602.1 MB/s

0.54%

-4.4%

unidirectional/4096

asio (callbacks)

629.7 MB/s

1.12%

baseline

unidirectional/65536

corosio (kqueue)

7.32 GB/s

1.44%

+67.2%

unidirectional/65536

asio (coroutines)

4.26 GB/s

1.87%

-2.6%

unidirectional/65536

asio (callbacks)

4.37 GB/s

2.43%

baseline

unidirectional_lockless/1024

corosio (kqueue)

328.2 MB/s

0.51%

+125.2%

unidirectional_lockless/1024

asio (coroutines)

159.8 MB/s

0.58%

+9.6%

unidirectional_lockless/1024

asio (callbacks)

145.7 MB/s

0.59%

baseline

unidirectional_lockless/1048576

corosio (kqueue)

9.12 GB/s

0.81%

+1.2%

unidirectional_lockless/1048576

asio (coroutines)

8.25 GB/s

5.89%

-8.5%

unidirectional_lockless/1048576

asio (callbacks)

9.02 GB/s

1.13%

baseline

unidirectional_lockless/16384

corosio (kqueue)

3.76 GB/s

0.54%

+94.8%

unidirectional_lockless/16384

asio (coroutines)

1.90 GB/s

0.94%

-1.8%

unidirectional_lockless/16384

asio (callbacks)

1.93 GB/s

1.53%

baseline

unidirectional_lockless/262144

corosio (kqueue)

9.14 GB/s

1.74%

+20.4%

unidirectional_lockless/262144

asio (coroutines)

6.07 GB/s

8.88%

-20.0%

unidirectional_lockless/262144

asio (callbacks)

7.59 GB/s

10.28%

baseline

unidirectional_lockless/4096

corosio (kqueue)

1.26 GB/s

0.74%

+96.5%

unidirectional_lockless/4096

asio (coroutines)

613.2 MB/s

0.64%

-4.1%

unidirectional_lockless/4096

asio (callbacks)

639.1 MB/s

0.40%

baseline

unidirectional_lockless/65536

corosio (kqueue)

7.39 GB/s

0.88%

+74.6%

unidirectional_lockless/65536

asio (coroutines)

4.20 GB/s

1.66%

-0.7%

unidirectional_lockless/65536

asio (callbacks)

4.23 GB/s

2.96%

baseline
Edit this Page