Windows Benchmarks

Generated 2026-11-16 from corosio 13bcf933adf8 — 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.

0 faster · 17 within noise · 60 slower of 77 benchmarks — median -4.7% vs Boost.Asio (callbacks) on the same reactor.

Summary — Windows ← slower · within noise · faster → per benchmark, vs Boost.Asio (callbacks) socket_throughput 12 benchmarks slower than asio (callbacks) 12 15 benchmarks within noise than asio (callbacks) 15 -1.6% http_server 10 benchmarks slower than asio (callbacks) 10 1 benchmark within noise than asio (callbacks) -3.1% fan_out 18 benchmarks slower than asio (callbacks) 18 -5.9% accept_churn 9 benchmarks slower than asio (callbacks) 9 -6.7% socket_latency 11 benchmarks slower than asio (callbacks) 11 1 benchmark within noise than asio (callbacks) -9.6%

Test Environment

CPU

AMD Ryzen 5 3600 6-Core Processor

Cores

6

RAM (GB)

64

OS

MSYS_NT-10.0-20348 3.4.7-ea781829.x86_64

Kernel/build

2023-07-05 12:05 UTC

Compiler

c++.exe (x86_64-posix-seh-rev2, Built by MinGW-W64 project) 12.2.0

CMake

cmake version 3.26.3

liburing

n/a

Boost commit

644b3c38a6c52972a1678677951307194ccb04e2

Asio commit

e99d9da6d9b76a222e29dcaf3d2df51c788b7527

Asio reactor

IOCP

Capy commit

25299eb4fd17423b761cc35d47008df7b68976d1

Corosio commit

13bcf933adf8c5ce8576bd3349ae58d2e46e5c53

Corosio branch

develop

Date (UTC)

2026-11-16

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 — Windows ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -15% -10% -5% +5% +10% 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. 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. -4% burst_lockless/100 — Same as burst, with the context in single-threaded lockless mode; N is the burst size. +6% 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. -4% 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. 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. -9% 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. 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 (IOCP)

1.55K ops/s

0.94%

-6.0%

burst/10

asio (coroutines)

1.69K ops/s

0.65%

+2.2%

burst/10

asio (callbacks)

1.65K ops/s

0.64%

baseline

burst/100

corosio (IOCP)

156.7 ops/s

1.41%

-4.3%

burst/100

asio (coroutines)

170.8 ops/s

0.46%

+4.2%

burst/100

asio (callbacks)

163.8 ops/s

0.61%

baseline

burst_lockless/10

corosio (IOCP)

1.57K ops/s

0.16%

-6.1%

burst_lockless/10

asio (coroutines)

1.71K ops/s

0.22%

+2.3%

burst_lockless/10

asio (callbacks)

1.67K ops/s

0.43%

baseline

burst_lockless/100

corosio (IOCP)

159.3 ops/s

0.86%

-4.1%

burst_lockless/100

asio (coroutines)

176.2 ops/s

0.76%

+6.0%

burst_lockless/100

asio (callbacks)

166.2 ops/s

0.60%

baseline

concurrent/1

corosio (IOCP)

12.56K ops/s

0.20%

-8.6%

concurrent/1

asio (coroutines)

13.22K ops/s

0.20%

-3.8%

concurrent/1

asio (callbacks)

13.74K ops/s

0.20%

baseline

concurrent/16

corosio (IOCP)

13.47K ops/s

0.36%

-6.7%

concurrent/16

asio (coroutines)

14.04K ops/s

0.24%

-2.8%

concurrent/16

asio (callbacks)

14.44K ops/s

0.30%

baseline

concurrent/4

corosio (IOCP)

13.18K ops/s

0.19%

-7.6%

concurrent/4

asio (coroutines)

13.76K ops/s

0.38%

-3.4%

concurrent/4

asio (callbacks)

14.25K ops/s

0.20%

baseline

sequential

corosio (IOCP)

12.51K ops/s

0.28%

-8.8%

sequential

asio (coroutines)

13.22K ops/s

1.59%

-3.7%

sequential

asio (callbacks)

13.72K ops/s

0.29%

baseline

sequential_lockless

corosio (IOCP)

12.55K ops/s

0.19%

-8.8%

sequential_lockless

asio (coroutines)

13.25K ops/s

0.26%

-3.7%

sequential_lockless

asio (callbacks)

13.75K ops/s

0.20%

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 — Windows, part 1 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -15% -10% -5% +5% 0% concurrent_parents/1 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -6% 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. -8% concurrent_parents/16 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -11% concurrent_parents/16 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 — N independent parents each fan out to 16 sub-requests at once; N is the number of parents. -10% 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. -10% 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. -7% concurrent_parents_lockless/16 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -11% 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. -5% concurrent_parents_lockless/4 — Same as concurrent_parents, with the context in single-threaded lockless mode; N is the number of parents. -10% concurrent_parents_lockless/4
Detailed results
Benchmark Implementation Median CV vs asio

concurrent_parents/1

corosio (IOCP)

7.36K ops/s

0.16%

-5.7%

concurrent_parents/1

asio (coroutines)

6.97K ops/s

0.86%

-10.8%

concurrent_parents/1

asio (callbacks)

7.81K ops/s

0.29%

baseline

concurrent_parents/16

corosio (IOCP)

6.79K ops/s

0.79%

-7.7%

concurrent_parents/16

asio (coroutines)

6.52K ops/s

0.85%

-11.4%

concurrent_parents/16

asio (callbacks)

7.36K ops/s

0.62%

baseline

concurrent_parents/4

corosio (IOCP)

7.29K ops/s

0.59%

-5.1%

concurrent_parents/4

asio (coroutines)

6.92K ops/s

0.66%

-10.0%

concurrent_parents/4

asio (callbacks)

7.68K ops/s

0.29%

baseline

concurrent_parents_lockless/1

corosio (IOCP)

7.38K ops/s

0.30%

-5.5%

concurrent_parents_lockless/1

asio (coroutines)

7.06K ops/s

0.30%

-9.7%

concurrent_parents_lockless/1

asio (callbacks)

7.81K ops/s

0.27%

baseline

concurrent_parents_lockless/16

corosio (IOCP)

6.87K ops/s

1.05%

-6.9%

concurrent_parents_lockless/16

asio (coroutines)

6.58K ops/s

0.59%

-10.9%

concurrent_parents_lockless/16

asio (callbacks)

7.38K ops/s

0.51%

baseline

concurrent_parents_lockless/4

corosio (IOCP)

7.29K ops/s

0.58%

-5.1%

concurrent_parents_lockless/4

asio (coroutines)

6.93K ops/s

0.36%

-9.9%

concurrent_parents_lockless/4

asio (callbacks)

7.68K ops/s

0.33%

baseline

fan_out — Windows, part 2 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -20% -15% -10% -5% +5% 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. 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. 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. -11% fork_join_lockless/1 — Same as fork_join, with the context in single-threaded lockless mode; N is the fan-out width. 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. -10% 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. -5% 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. -14% nested_lockless/4
Detailed results
Benchmark Implementation Median CV vs asio

fork_join/1

corosio (IOCP)

112.5K ops/s

0.19%

-10.4%

fork_join/1

asio (coroutines)

111.5K ops/s

0.47%

-11.2%

fork_join/1

asio (callbacks)

125.6K ops/s

0.45%

baseline

fork_join/16

corosio (IOCP)

7.37K ops/s

0.22%

-5.7%

fork_join/16

asio (coroutines)

6.98K ops/s

0.39%

-10.6%

fork_join/16

asio (callbacks)

7.81K ops/s

0.19%

baseline

fork_join/4

corosio (IOCP)

29.16K ops/s

0.34%

-6.7%

fork_join/4

asio (coroutines)

28.14K ops/s

0.44%

-9.9%

fork_join/4

asio (callbacks)

31.24K ops/s

0.32%

baseline

fork_join/64

corosio (IOCP)

1.83K ops/s

1.17%

-4.8%

fork_join/64

asio (coroutines)

1.73K ops/s

1.14%

-9.9%

fork_join/64

asio (callbacks)

1.92K ops/s

0.30%

baseline

fork_join_lockless/1

corosio (IOCP)

112.1K ops/s

0.52%

-10.6%

fork_join_lockless/1

asio (coroutines)

111.7K ops/s

0.27%

-11.0%

fork_join_lockless/1

asio (callbacks)

125.5K ops/s

0.24%

baseline

fork_join_lockless/16

corosio (IOCP)

7.36K ops/s

0.56%

-5.8%

fork_join_lockless/16

asio (coroutines)

6.99K ops/s

0.88%

-10.5%

fork_join_lockless/16

asio (callbacks)

7.81K ops/s

0.21%

baseline

fork_join_lockless/4

corosio (IOCP)

29.21K ops/s

0.24%

-6.4%

fork_join_lockless/4

asio (coroutines)

28.24K ops/s

0.52%

-9.6%

fork_join_lockless/4

asio (callbacks)

31.23K ops/s

0.19%

baseline

fork_join_lockless/64

corosio (IOCP)

1.83K ops/s

1.15%

-4.7%

fork_join_lockless/64

asio (coroutines)

1.73K ops/s

0.80%

-9.8%

fork_join_lockless/64

asio (callbacks)

1.92K ops/s

0.18%

baseline

nested/16

corosio (IOCP)

1.80K ops/s

0.94%

-5.9%

nested/16

asio (coroutines)

1.67K ops/s

0.63%

-12.5%

nested/16

asio (callbacks)

1.91K ops/s

0.33%

baseline

nested/4

corosio (IOCP)

7.23K ops/s

0.21%

-7.1%

nested/4

asio (coroutines)

6.70K ops/s

0.50%

-13.9%

nested/4

asio (callbacks)

7.78K ops/s

0.35%

baseline

nested_lockless/16

corosio (IOCP)

1.80K ops/s

0.72%

-5.9%

nested_lockless/16

asio (coroutines)

1.69K ops/s

0.68%

-11.9%

nested_lockless/16

asio (callbacks)

1.91K ops/s

0.28%

baseline

nested_lockless/4

corosio (IOCP)

7.23K ops/s

0.34%

-7.1%

nested_lockless/4

asio (coroutines)

6.68K ops/s

0.86%

-14.1%

nested_lockless/4

asio (callbacks)

7.78K ops/s

0.49%

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 — Windows ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -7.5% -5% -2.5% +2.5% 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. 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. -6% 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. multithread/4 — 32 client/server pairs share one context serviced by N threads running the context; N is the thread count. -3% 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. -5% single_conn single_conn_lockless — Same as single_conn, with the context in single-threaded lockless mode. -2% 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 (IOCP)

111.8K ops/s

0.33%

-2.3%

concurrent/1

asio (coroutines)

109.1K ops/s

0.26%

-4.6%

concurrent/1

asio (callbacks)

114.4K ops/s

0.33%

baseline

concurrent/16

corosio (IOCP)

108.9K ops/s

0.29%

-3.1%

concurrent/16

asio (coroutines)

107.2K ops/s

0.27%

-4.6%

concurrent/16

asio (callbacks)

112.4K ops/s

0.24%

baseline

concurrent/32

corosio (IOCP)

108.1K ops/s

1.58%

-3.4%

concurrent/32

asio (coroutines)

106.8K ops/s

0.23%

-4.6%

concurrent/32

asio (callbacks)

111.9K ops/s

0.24%

baseline

concurrent/4

corosio (IOCP)

109.6K ops/s

0.32%

-2.7%

concurrent/4

asio (coroutines)

107.2K ops/s

0.47%

-4.8%

concurrent/4

asio (callbacks)

112.6K ops/s

0.26%

baseline

multithread/1

corosio (IOCP)

108.3K ops/s

1.56%

-3.1%

multithread/1

asio (coroutines)

106.7K ops/s

0.21%

-4.5%

multithread/1

asio (callbacks)

111.7K ops/s

0.37%

baseline

multithread/16

corosio (IOCP)

264.4K ops/s

0.44%

-6.5%

multithread/16

asio (coroutines)

274.3K ops/s

0.33%

-3.0%

multithread/16

asio (callbacks)

282.7K ops/s

0.45%

baseline

multithread/2

corosio (IOCP)

181.3K ops/s

0.47%

-3.3%

multithread/2

asio (coroutines)

180.9K ops/s

0.40%

-3.6%

multithread/2

asio (callbacks)

187.6K ops/s

0.44%

baseline

multithread/4

corosio (IOCP)

223.3K ops/s

0.78%

-4.2%

multithread/4

asio (coroutines)

226.8K ops/s

0.38%

-2.8%

multithread/4

asio (callbacks)

233.2K ops/s

0.29%

baseline

multithread/8

corosio (IOCP)

264.2K ops/s

0.65%

-5.8%

multithread/8

asio (coroutines)

270.8K ops/s

0.74%

-3.4%

multithread/8

asio (callbacks)

280.3K ops/s

0.34%

baseline

single_conn

corosio (IOCP)

111.9K ops/s

0.28%

-2.6%

single_conn

asio (coroutines)

108.9K ops/s

0.44%

-5.1%

single_conn

asio (callbacks)

114.8K ops/s

0.49%

baseline

single_conn_lockless

corosio (IOCP)

111.9K ops/s

0.49%

-2.2%

single_conn_lockless

asio (coroutines)

109.1K ops/s

0.42%

-4.6%

single_conn_lockless

asio (callbacks)

114.4K ops/s

0.38%

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 — Windows ↑ faster (lower mean latency) · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -15% -10% -5% +5% 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. -11% concurrent/16 — N independent 64-byte pingpong pairs run concurrently on one context; N is the number of connection pairs. -9% 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. 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. pingpong/1024 pingpong/64 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. -9% pingpong/64 — One TCP connection ping-pongs a message client->server->client; N is the message size in bytes. -6% pingpong/64 pingpong_lockless/1 — Same as pingpong, with the context in single-threaded lockless mode; N is the message size in bytes. 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 (IOCP)

7.59 µs

0.37%

-9.3%

concurrent/1

asio (coroutines)

7.37 µs

0.73%

-6.0%

concurrent/1

asio (callbacks)

6.95 µs

0.38%

baseline

concurrent/16

corosio (IOCP)

125.43 µs

5.94%

-10.7%

concurrent/16

asio (coroutines)

123.54 µs

2.36%

-9.0%

concurrent/16

asio (callbacks)

113.34 µs

4.32%

baseline

concurrent/4

corosio (IOCP)

30.54 µs

0.38%

-9.6%

concurrent/4

asio (coroutines)

29.67 µs

0.38%

-6.5%

concurrent/4

asio (callbacks)

27.86 µs

0.39%

baseline

concurrent_lockless/1

corosio (IOCP)

7.61 µs

0.13%

-9.7%

concurrent_lockless/1

asio (coroutines)

7.35 µs

0.50%

-6.0%

concurrent_lockless/1

asio (callbacks)

6.93 µs

0.34%

baseline

concurrent_lockless/16

corosio (IOCP)

121.62 µs

0.40%

-9.8%

concurrent_lockless/16

asio (coroutines)

118.41 µs

4.37%

-6.9%

concurrent_lockless/16

asio (callbacks)

110.79 µs

0.61%

baseline

concurrent_lockless/4

corosio (IOCP)

30.51 µs

0.16%

-9.7%

concurrent_lockless/4

asio (coroutines)

29.54 µs

0.33%

-6.2%

concurrent_lockless/4

asio (callbacks)

27.82 µs

0.39%

baseline

pingpong/1

corosio (IOCP)

7.60 µs

0.58%

-9.4%

pingpong/1

asio (coroutines)

7.33 µs

0.56%

-5.6%

pingpong/1

asio (callbacks)

6.94 µs

0.46%

baseline

pingpong/1024

corosio (IOCP)

7.74 µs

0.76%

-10.1%

pingpong/1024

asio (coroutines)

7.46 µs

0.48%

-6.1%

pingpong/1024

asio (callbacks)

7.03 µs

0.29%

baseline

pingpong/64

corosio (IOCP)

7.60 µs

0.33%

-9.2%

pingpong/64

asio (coroutines)

7.34 µs

0.51%

-5.5%

pingpong/64

asio (callbacks)

6.96 µs

0.21%

baseline

pingpong_lockless/1

corosio (IOCP)

7.56 µs

0.28%

-9.5%

pingpong_lockless/1

asio (coroutines)

7.33 µs

0.53%

-6.2%

pingpong_lockless/1

asio (callbacks)

6.90 µs

0.48%

baseline

pingpong_lockless/1024

corosio (IOCP)

7.71 µs

0.33%

-9.6%

pingpong_lockless/1024

asio (coroutines)

7.45 µs

0.68%

-5.9%

pingpong_lockless/1024

asio (callbacks)

7.04 µs

0.33%

baseline

pingpong_lockless/64

corosio (IOCP)

7.60 µs

0.30%

-9.3%

pingpong_lockless/64

asio (coroutines)

7.37 µs

0.41%

-5.9%

pingpong_lockless/64

asio (callbacks)

6.96 µs

0.38%

baseline


socket_throughput

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

socket_throughput — Windows, part 1 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -15% -10% -5% +5% +10% 0% bidirectional/1024 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. -3% 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. bidirectional/1048576 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. -9% 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. 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. +3% bidirectional_lockless/1048576 — Same as bidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. +2% 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. 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 (IOCP)

283.6 MB/s

0.42%

-3.3%

bidirectional/1024

asio (coroutines)

279.3 MB/s

0.54%

-4.8%

bidirectional/1024

asio (callbacks)

293.4 MB/s

0.44%

baseline

bidirectional/1048576

corosio (IOCP)

5.01 GB/s

14.71%

+0.5%

bidirectional/1048576

asio (coroutines)

4.56 GB/s

9.69%

-8.6%

bidirectional/1048576

asio (callbacks)

4.99 GB/s

10.56%

baseline

bidirectional/16384

corosio (IOCP)

3.77 GB/s

0.71%

-2.7%

bidirectional/16384

asio (coroutines)

3.73 GB/s

1.12%

-3.9%

bidirectional/16384

asio (callbacks)

3.88 GB/s

0.73%

baseline

bidirectional/262144

corosio (IOCP)

9.65 GB/s

0.71%

+1.7%

bidirectional/262144

asio (coroutines)

9.44 GB/s

1.69%

-0.4%

bidirectional/262144

asio (callbacks)

9.48 GB/s

6.74%

baseline

bidirectional/4096

corosio (IOCP)

1.09 GB/s

0.49%

-2.9%

bidirectional/4096

asio (coroutines)

1.07 GB/s

0.33%

-4.0%

bidirectional/4096

asio (callbacks)

1.12 GB/s

0.61%

baseline

bidirectional/65536

corosio (IOCP)

9.16 GB/s

1.14%

-1.6%

bidirectional/65536

asio (coroutines)

9.08 GB/s

2.01%

-2.5%

bidirectional/65536

asio (callbacks)

9.31 GB/s

1.47%

baseline

bidirectional_lockless/1024

corosio (IOCP)

283.5 MB/s

0.51%

-3.0%

bidirectional_lockless/1024

asio (coroutines)

278.6 MB/s

0.47%

-4.7%

bidirectional_lockless/1024

asio (callbacks)

292.2 MB/s

0.32%

baseline

bidirectional_lockless/1048576

corosio (IOCP)

5.37 GB/s

13.09%

+3.5%

bidirectional_lockless/1048576

asio (coroutines)

5.29 GB/s

10.08%

+1.9%

bidirectional_lockless/1048576

asio (callbacks)

5.19 GB/s

11.01%

baseline

bidirectional_lockless/16384

corosio (IOCP)

3.73 GB/s

0.51%

-2.9%

bidirectional_lockless/16384

asio (coroutines)

3.67 GB/s

0.23%

-4.4%

bidirectional_lockless/16384

asio (callbacks)

3.85 GB/s

0.30%

baseline

bidirectional_lockless/262144

corosio (IOCP)

9.76 GB/s

1.66%

-1.2%

bidirectional_lockless/262144

asio (coroutines)

9.78 GB/s

1.24%

-1.0%

bidirectional_lockless/262144

asio (callbacks)

9.88 GB/s

1.09%

baseline

bidirectional_lockless/4096

corosio (IOCP)

1.08 GB/s

0.36%

-3.2%

bidirectional_lockless/4096

asio (coroutines)

1.07 GB/s

0.47%

-4.5%

bidirectional_lockless/4096

asio (callbacks)

1.12 GB/s

0.38%

baseline

bidirectional_lockless/65536

corosio (IOCP)

8.97 GB/s

0.70%

-2.3%

bidirectional_lockless/65536

asio (coroutines)

8.88 GB/s

0.45%

-3.3%

bidirectional_lockless/65536

asio (callbacks)

9.18 GB/s

0.72%

baseline

socket_throughput — Windows, part 2 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -6% -4% -2% +2% 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. -1% 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. -1% 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. -3% 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. -3% 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. -1% 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. -1% multithread/8
Detailed results
Benchmark Implementation Median CV vs asio

multithread/2

corosio (IOCP)

12.64 GB/s

0.32%

-1.4%

multithread/2

asio (coroutines)

12.67 GB/s

1.12%

-1.2%

multithread/2

asio (callbacks)

12.83 GB/s

0.50%

baseline

multithread/4

corosio (IOCP)

16.84 GB/s

0.65%

-2.5%

multithread/4

asio (coroutines)

16.73 GB/s

0.57%

-3.2%

multithread/4

asio (callbacks)

17.28 GB/s

0.38%

baseline

multithread/8

corosio (IOCP)

18.45 GB/s

0.99%

-0.7%

multithread/8

asio (coroutines)

18.43 GB/s

0.49%

-0.8%

multithread/8

asio (callbacks)

18.58 GB/s

1.17%

baseline

socket_throughput — Windows, part 3 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -10% -5% +5% +10% 0% unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. -3% unidirectional/1024 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. -5% 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. +4% unidirectional/262144 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. +4% 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. unidirectional_lockless/1024 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. 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 (IOCP)

286.3 MB/s

0.35%

-3.4%

unidirectional/1024

asio (coroutines)

281.8 MB/s

0.40%

-4.9%

unidirectional/1024

asio (callbacks)

296.4 MB/s

0.49%

baseline

unidirectional/1048576

corosio (IOCP)

4.18 GB/s

3.73%

+0.6%

unidirectional/1048576

asio (coroutines)

4.27 GB/s

3.87%

+2.9%

unidirectional/1048576

asio (callbacks)

4.15 GB/s

4.79%

baseline

unidirectional/16384

corosio (IOCP)

3.82 GB/s

0.76%

-1.3%

unidirectional/16384

asio (coroutines)

3.75 GB/s

1.23%

-3.1%

unidirectional/16384

asio (callbacks)

3.86 GB/s

4.53%

baseline

unidirectional/262144

corosio (IOCP)

8.94 GB/s

5.14%

+4.1%

unidirectional/262144

asio (coroutines)

8.91 GB/s

4.94%

+3.8%

unidirectional/262144

asio (callbacks)

8.59 GB/s

10.55%

baseline

unidirectional/4096

corosio (IOCP)

1.10 GB/s

0.52%

-2.5%

unidirectional/4096

asio (coroutines)

1.08 GB/s

0.66%

-4.2%

unidirectional/4096

asio (callbacks)

1.13 GB/s

1.39%

baseline

unidirectional/65536

corosio (IOCP)

9.35 GB/s

1.55%

+1.8%

unidirectional/65536

asio (coroutines)

9.12 GB/s

3.50%

-0.7%

unidirectional/65536

asio (callbacks)

9.18 GB/s

9.84%

baseline

unidirectional_lockless/1024

corosio (IOCP)

286.1 MB/s

0.57%

-3.3%

unidirectional_lockless/1024

asio (coroutines)

281.6 MB/s

0.33%

-4.8%

unidirectional_lockless/1024

asio (callbacks)

295.9 MB/s

0.56%

baseline

unidirectional_lockless/1048576

corosio (IOCP)

4.71 GB/s

8.16%

+2.4%

unidirectional_lockless/1048576

asio (coroutines)

4.49 GB/s

6.02%

-2.3%

unidirectional_lockless/1048576

asio (callbacks)

4.60 GB/s

3.24%

baseline

unidirectional_lockless/16384

corosio (IOCP)

3.80 GB/s

0.93%

-1.7%

unidirectional_lockless/16384

asio (coroutines)

3.75 GB/s

1.31%

-3.0%

unidirectional_lockless/16384

asio (callbacks)

3.87 GB/s

4.52%

baseline

unidirectional_lockless/262144

corosio (IOCP)

9.84 GB/s

1.25%

-0.8%

unidirectional_lockless/262144

asio (coroutines)

9.73 GB/s

1.03%

-1.8%

unidirectional_lockless/262144

asio (callbacks)

9.91 GB/s

7.29%

baseline

unidirectional_lockless/4096

corosio (IOCP)

1.09 GB/s

0.38%

-2.8%

unidirectional_lockless/4096

asio (coroutines)

1.08 GB/s

0.72%

-4.4%

unidirectional_lockless/4096

asio (callbacks)

1.13 GB/s

1.24%

baseline

unidirectional_lockless/65536

corosio (IOCP)

9.41 GB/s

2.12%

-1.3%

unidirectional_lockless/65536

asio (coroutines)

9.26 GB/s

3.68%

-2.8%

unidirectional_lockless/65536

asio (callbacks)

9.53 GB/s

10.33%

baseline