Windows 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.

0 faster · 18 within noise · 59 slower of 77 benchmarks — median -6.1% vs Boost.Asio (callbacks) on the same reactor.

Summary — Windows ← slower · within noise · faster → per benchmark, vs Boost.Asio (callbacks) http_server 8 benchmarks slower than asio (callbacks) 8 3 benchmarks within noise than asio (callbacks) 3 -1.7% socket_throughput 13 benchmarks slower than asio (callbacks) 13 14 benchmarks within noise than asio (callbacks) 14 -2.1% socket_latency 11 benchmarks slower than asio (callbacks) 11 1 benchmark within noise than asio (callbacks) -7.3% accept_churn 9 benchmarks slower than asio (callbacks) 9 -8.5% fan_out 18 benchmarks slower than asio (callbacks) 18 -8.8%

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

39fa1e9a3491bd099b570935b3f3422065f91b03

Asio commit

a7dc25b4cb6c49a6946d86ea20664f1027203225

Asio reactor

IOCP

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 — Windows ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -20% -10% +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. -8% 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. -10% 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. 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. -4% 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)

1.54K ops/s

0.38%

-7.8%

burst/10

asio (coroutines)

1.70K ops/s

0.57%

+2.2%

burst/10

asio (callbacks)

1.67K ops/s

0.65%

baseline

burst/100

corosio (IOCP)

149.4 ops/s

0.43%

-8.7%

burst/100

asio (coroutines)

171.8 ops/s

0.54%

+4.9%

burst/100

asio (callbacks)

163.8 ops/s

0.68%

baseline

burst_lockless/10

corosio (IOCP)

1.54K ops/s

0.26%

-7.8%

burst_lockless/10

asio (coroutines)

1.72K ops/s

0.23%

+2.5%

burst_lockless/10

asio (callbacks)

1.67K ops/s

0.23%

baseline

burst_lockless/100

corosio (IOCP)

154.4 ops/s

0.73%

-7.7%

burst_lockless/100

asio (coroutines)

177.1 ops/s

0.53%

+5.9%

burst_lockless/100

asio (callbacks)

167.2 ops/s

0.48%

baseline

concurrent/1

corosio (IOCP)

12.38K ops/s

0.31%

-9.8%

concurrent/1

asio (coroutines)

13.22K ops/s

0.09%

-3.6%

concurrent/1

asio (callbacks)

13.72K ops/s

0.18%

baseline

concurrent/16

corosio (IOCP)

13.30K ops/s

0.24%

-7.9%

concurrent/16

asio (coroutines)

14.00K ops/s

0.22%

-3.0%

concurrent/16

asio (callbacks)

14.44K ops/s

0.17%

baseline

concurrent/4

corosio (IOCP)

13.02K ops/s

0.26%

-8.5%

concurrent/4

asio (coroutines)

13.81K ops/s

0.19%

-3.0%

concurrent/4

asio (callbacks)

14.24K ops/s

0.33%

baseline

sequential

corosio (IOCP)

12.34K ops/s

1.24%

-9.7%

sequential

asio (coroutines)

13.15K ops/s

0.17%

-3.8%

sequential

asio (callbacks)

13.67K ops/s

0.18%

baseline

sequential_lockless

corosio (IOCP)

12.39K ops/s

1.32%

-9.5%

sequential_lockless

asio (coroutines)

13.20K ops/s

0.16%

-3.7%

sequential_lockless

asio (callbacks)

13.70K ops/s

0.15%

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. -7% 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. -11% 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. -9% 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. -7% 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. -11% 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. -9% 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
Detailed results
Benchmark Implementation Median CV vs asio

concurrent_parents/1

corosio (IOCP)

7.16K ops/s

0.30%

-7.4%

concurrent_parents/1

asio (coroutines)

6.85K ops/s

0.38%

-11.3%

concurrent_parents/1

asio (callbacks)

7.73K ops/s

0.21%

baseline

concurrent_parents/16

corosio (IOCP)

6.43K ops/s

0.76%

-10.7%

concurrent_parents/16

asio (coroutines)

6.37K ops/s

0.73%

-11.5%

concurrent_parents/16

asio (callbacks)

7.20K ops/s

0.43%

baseline

concurrent_parents/4

corosio (IOCP)

6.90K ops/s

0.36%

-9.1%

concurrent_parents/4

asio (coroutines)

6.81K ops/s

0.34%

-10.2%

concurrent_parents/4

asio (callbacks)

7.59K ops/s

0.29%

baseline

concurrent_parents_lockless/1

corosio (IOCP)

7.15K ops/s

0.22%

-7.5%

concurrent_parents_lockless/1

asio (coroutines)

6.95K ops/s

0.23%

-10.1%

concurrent_parents_lockless/1

asio (callbacks)

7.73K ops/s

0.16%

baseline

concurrent_parents_lockless/16

corosio (IOCP)

6.44K ops/s

0.63%

-11.1%

concurrent_parents_lockless/16

asio (coroutines)

6.42K ops/s

0.34%

-11.3%

concurrent_parents_lockless/16

asio (callbacks)

7.24K ops/s

0.65%

baseline

concurrent_parents_lockless/4

corosio (IOCP)

6.92K ops/s

0.34%

-9.1%

concurrent_parents_lockless/4

asio (coroutines)

6.80K ops/s

0.23%

-10.5%

concurrent_parents_lockless/4

asio (callbacks)

7.61K ops/s

0.16%

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. -5% 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. 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. 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. -11% 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. -13% nested_lockless/4
Detailed results
Benchmark Implementation Median CV vs asio

fork_join/1

corosio (IOCP)

112.4K ops/s

0.37%

-8.9%

fork_join/1

asio (coroutines)

109.9K ops/s

0.96%

-11.0%

fork_join/1

asio (callbacks)

123.5K ops/s

0.27%

baseline

fork_join/16

corosio (IOCP)

7.15K ops/s

0.14%

-7.6%

fork_join/16

asio (coroutines)

6.85K ops/s

0.87%

-11.5%

fork_join/16

asio (callbacks)

7.74K ops/s

0.18%

baseline

fork_join/4

corosio (IOCP)

29.33K ops/s

0.26%

-4.8%

fork_join/4

asio (coroutines)

27.75K ops/s

0.28%

-9.9%

fork_join/4

asio (callbacks)

30.82K ops/s

0.17%

baseline

fork_join/64

corosio (IOCP)

1.73K ops/s

0.83%

-8.7%

fork_join/64

asio (coroutines)

1.70K ops/s

0.17%

-9.9%

fork_join/64

asio (callbacks)

1.89K ops/s

0.46%

baseline

fork_join_lockless/1

corosio (IOCP)

112.9K ops/s

0.19%

-8.8%

fork_join_lockless/1

asio (coroutines)

109.8K ops/s

0.53%

-11.3%

fork_join_lockless/1

asio (callbacks)

123.7K ops/s

0.39%

baseline

fork_join_lockless/16

corosio (IOCP)

7.16K ops/s

0.14%

-7.4%

fork_join_lockless/16

asio (coroutines)

6.86K ops/s

0.51%

-11.3%

fork_join_lockless/16

asio (callbacks)

7.73K ops/s

0.26%

baseline

fork_join_lockless/4

corosio (IOCP)

29.24K ops/s

0.33%

-5.1%

fork_join_lockless/4

asio (coroutines)

27.80K ops/s

0.19%

-9.8%

fork_join_lockless/4

asio (callbacks)

30.81K ops/s

0.16%

baseline

fork_join_lockless/64

corosio (IOCP)

1.73K ops/s

0.52%

-8.8%

fork_join_lockless/64

asio (coroutines)

1.71K ops/s

0.37%

-9.8%

fork_join_lockless/64

asio (callbacks)

1.90K ops/s

0.42%

baseline

nested/16

corosio (IOCP)

1.70K ops/s

0.56%

-10.5%

nested/16

asio (coroutines)

1.66K ops/s

0.39%

-12.3%

nested/16

asio (callbacks)

1.89K ops/s

0.26%

baseline

nested/4

corosio (IOCP)

6.99K ops/s

0.24%

-8.9%

nested/4

asio (coroutines)

6.65K ops/s

0.53%

-13.3%

nested/4

asio (callbacks)

7.67K ops/s

0.52%

baseline

nested_lockless/16

corosio (IOCP)

1.69K ops/s

0.40%

-10.7%

nested_lockless/16

asio (coroutines)

1.66K ops/s

0.49%

-12.7%

nested_lockless/16

asio (callbacks)

1.90K ops/s

0.34%

baseline

nested_lockless/4

corosio (IOCP)

7.00K ops/s

0.30%

-9.0%

nested_lockless/4

asio (coroutines)

6.66K ops/s

0.42%

-13.4%

nested_lockless/4

asio (callbacks)

7.69K ops/s

0.63%

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. -2% 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. 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. single_conn single_conn_lockless — Same as single_conn, with the context in single-threaded lockless mode. 0% single_conn_lockless — Same as single_conn, with the context in single-threaded lockless mode. -5% single_conn_lockless
Detailed results
Benchmark Implementation Median CV vs asio

concurrent/1

corosio (IOCP)

111.9K ops/s

0.42%

-0.8%

concurrent/1

asio (coroutines)

107.8K ops/s

0.33%

-4.5%

concurrent/1

asio (callbacks)

112.8K ops/s

0.37%

baseline

concurrent/16

corosio (IOCP)

108.6K ops/s

0.33%

-1.9%

concurrent/16

asio (coroutines)

106.2K ops/s

0.39%

-4.0%

concurrent/16

asio (callbacks)

110.7K ops/s

0.23%

baseline

concurrent/32

corosio (IOCP)

108.1K ops/s

0.20%

-1.7%

concurrent/32

asio (coroutines)

105.6K ops/s

0.15%

-4.0%

concurrent/32

asio (callbacks)

109.9K ops/s

0.24%

baseline

concurrent/4

corosio (IOCP)

109.6K ops/s

0.24%

-1.1%

concurrent/4

asio (coroutines)

106.4K ops/s

0.26%

-4.0%

concurrent/4

asio (callbacks)

110.8K ops/s

0.35%

baseline

multithread/1

corosio (IOCP)

108.0K ops/s

0.17%

-1.6%

multithread/1

asio (coroutines)

105.6K ops/s

0.24%

-3.7%

multithread/1

asio (callbacks)

109.7K ops/s

0.26%

baseline

multithread/16

corosio (IOCP)

268.9K ops/s

0.21%

-6.1%

multithread/16

asio (coroutines)

279.3K ops/s

0.43%

-2.4%

multithread/16

asio (callbacks)

286.3K ops/s

0.52%

baseline

multithread/2

corosio (IOCP)

181.9K ops/s

0.68%

-2.9%

multithread/2

asio (coroutines)

180.1K ops/s

0.51%

-3.9%

multithread/2

asio (callbacks)

187.4K ops/s

0.47%

baseline

multithread/4

corosio (IOCP)

225.9K ops/s

0.37%

-4.5%

multithread/4

asio (coroutines)

228.7K ops/s

0.65%

-3.3%

multithread/4

asio (callbacks)

236.6K ops/s

0.55%

baseline

multithread/8

corosio (IOCP)

268.0K ops/s

0.42%

-5.8%

multithread/8

asio (coroutines)

277.0K ops/s

0.56%

-2.6%

multithread/8

asio (callbacks)

284.6K ops/s

0.68%

baseline

single_conn

corosio (IOCP)

111.8K ops/s

0.37%

-0.7%

single_conn

asio (coroutines)

107.8K ops/s

0.05%

-4.2%

single_conn

asio (callbacks)

112.5K ops/s

0.26%

baseline

single_conn_lockless

corosio (IOCP)

112.2K ops/s

0.32%

-0.5%

single_conn_lockless

asio (coroutines)

107.6K ops/s

0.56%

-4.5%

single_conn_lockless

asio (callbacks)

112.7K ops/s

0.20%

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. -8% 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. -6% 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. 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. 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. -7% 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)

7.57 µs

0.40%

-7.2%

concurrent/1

asio (coroutines)

7.48 µs

0.45%

-5.9%

concurrent/1

asio (callbacks)

7.06 µs

0.17%

baseline

concurrent/16

corosio (IOCP)

130.96 µs

6.14%

-11.2%

concurrent/16

asio (coroutines)

127.44 µs

2.57%

-8.2%

concurrent/16

asio (callbacks)

117.75 µs

3.93%

baseline

concurrent/4

corosio (IOCP)

30.40 µs

0.21%

-7.4%

concurrent/4

asio (coroutines)

29.95 µs

0.34%

-5.9%

concurrent/4

asio (callbacks)

28.30 µs

0.28%

baseline

concurrent_lockless/1

corosio (IOCP)

7.57 µs

0.56%

-7.3%

concurrent_lockless/1

asio (coroutines)

7.48 µs

0.22%

-6.0%

concurrent_lockless/1

asio (callbacks)

7.06 µs

0.17%

baseline

concurrent_lockless/16

corosio (IOCP)

121.42 µs

0.27%

-8.0%

concurrent_lockless/16

asio (coroutines)

120.27 µs

0.37%

-7.0%

concurrent_lockless/16

asio (callbacks)

112.43 µs

0.28%

baseline

concurrent_lockless/4

corosio (IOCP)

30.43 µs

0.42%

-7.6%

concurrent_lockless/4

asio (coroutines)

29.95 µs

0.41%

-5.9%

concurrent_lockless/4

asio (callbacks)

28.28 µs

0.37%

baseline

pingpong/1

corosio (IOCP)

7.56 µs

0.44%

-7.0%

pingpong/1

asio (coroutines)

7.45 µs

0.77%

-5.5%

pingpong/1

asio (callbacks)

7.06 µs

0.28%

baseline

pingpong/1024

corosio (IOCP)

7.67 µs

0.48%

-7.2%

pingpong/1024

asio (coroutines)

7.56 µs

0.37%

-5.7%

pingpong/1024

asio (callbacks)

7.16 µs

0.49%

baseline

pingpong/64

corosio (IOCP)

7.60 µs

0.41%

-7.5%

pingpong/64

asio (coroutines)

7.46 µs

0.22%

-5.5%

pingpong/64

asio (callbacks)

7.07 µs

0.22%

baseline

pingpong_lockless/1

corosio (IOCP)

7.53 µs

0.25%

-7.0%

pingpong_lockless/1

asio (coroutines)

7.43 µs

0.27%

-5.6%

pingpong_lockless/1

asio (callbacks)

7.04 µs

0.35%

baseline

pingpong_lockless/1024

corosio (IOCP)

7.69 µs

0.25%

-7.5%

pingpong_lockless/1024

asio (coroutines)

7.56 µs

0.36%

-5.6%

pingpong_lockless/1024

asio (callbacks)

7.15 µs

0.35%

baseline

pingpong_lockless/64

corosio (IOCP)

7.56 µs

0.17%

-7.0%

pingpong_lockless/64

asio (coroutines)

7.47 µs

0.29%

-5.7%

pingpong_lockless/64

asio (callbacks)

7.07 µs

0.18%

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% 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. -5% bidirectional/1024 bidirectional/1048576 — Both ends of the connection write and read simultaneously; N is the chunk size in bytes. -9% 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. +0% 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. 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. -1% 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)

279.9 MB/s

0.32%

-3.0%

bidirectional/1024

asio (coroutines)

274.1 MB/s

0.34%

-5.0%

bidirectional/1024

asio (callbacks)

288.6 MB/s

0.51%

baseline

bidirectional/1048576

corosio (IOCP)

4.42 GB/s

7.80%

-9.0%

bidirectional/1048576

asio (coroutines)

4.66 GB/s

6.82%

-4.2%

bidirectional/1048576

asio (callbacks)

4.86 GB/s

9.63%

baseline

bidirectional/16384

corosio (IOCP)

3.75 GB/s

0.61%

-1.9%

bidirectional/16384

asio (coroutines)

3.68 GB/s

0.34%

-3.8%

bidirectional/16384

asio (callbacks)

3.82 GB/s

0.55%

baseline

bidirectional/262144

corosio (IOCP)

9.55 GB/s

0.83%

+0.2%

bidirectional/262144

asio (coroutines)

9.47 GB/s

6.89%

-0.7%

bidirectional/262144

asio (callbacks)

9.54 GB/s

0.42%

baseline

bidirectional/4096

corosio (IOCP)

1.07 GB/s

0.27%

-2.8%

bidirectional/4096

asio (coroutines)

1.06 GB/s

0.45%

-4.3%

bidirectional/4096

asio (callbacks)

1.10 GB/s

0.36%

baseline

bidirectional/65536

corosio (IOCP)

9.04 GB/s

0.94%

-2.6%

bidirectional/65536

asio (coroutines)

9.00 GB/s

2.37%

-3.0%

bidirectional/65536

asio (callbacks)

9.28 GB/s

0.96%

baseline

bidirectional_lockless/1024

corosio (IOCP)

279.6 MB/s

0.44%

-2.9%

bidirectional_lockless/1024

asio (coroutines)

274.6 MB/s

0.65%

-4.6%

bidirectional_lockless/1024

asio (callbacks)

287.9 MB/s

0.49%

baseline

bidirectional_lockless/1048576

corosio (IOCP)

4.69 GB/s

7.57%

-1.5%

bidirectional_lockless/1048576

asio (coroutines)

4.61 GB/s

7.67%

-3.1%

bidirectional_lockless/1048576

asio (callbacks)

4.76 GB/s

8.69%

baseline

bidirectional_lockless/16384

corosio (IOCP)

3.72 GB/s

0.85%

-1.8%

bidirectional_lockless/16384

asio (coroutines)

3.68 GB/s

0.61%

-3.0%

bidirectional_lockless/16384

asio (callbacks)

3.79 GB/s

0.23%

baseline

bidirectional_lockless/262144

corosio (IOCP)

9.76 GB/s

1.34%

-0.1%

bidirectional_lockless/262144

asio (coroutines)

9.71 GB/s

0.99%

-0.6%

bidirectional_lockless/262144

asio (callbacks)

9.76 GB/s

1.25%

baseline

bidirectional_lockless/4096

corosio (IOCP)

1.07 GB/s

0.29%

-3.0%

bidirectional_lockless/4096

asio (coroutines)

1.06 GB/s

0.52%

-4.6%

bidirectional_lockless/4096

asio (callbacks)

1.11 GB/s

0.47%

baseline

bidirectional_lockless/65536

corosio (IOCP)

8.91 GB/s

1.43%

-2.0%

bidirectional_lockless/65536

asio (coroutines)

8.85 GB/s

0.38%

-2.7%

bidirectional_lockless/65536

asio (callbacks)

9.09 GB/s

0.19%

baseline
socket_throughput — Windows, part 2 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -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. -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. -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. 0% 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)

12.68 GB/s

0.54%

-2.1%

multithread/2

asio (coroutines)

12.73 GB/s

0.28%

-1.8%

multithread/2

asio (callbacks)

12.96 GB/s

0.37%

baseline

multithread/4

corosio (IOCP)

17.09 GB/s

0.54%

-2.7%

multithread/4

asio (coroutines)

17.06 GB/s

0.77%

-2.8%

multithread/4

asio (callbacks)

17.56 GB/s

0.45%

baseline

multithread/8

corosio (IOCP)

18.40 GB/s

0.37%

-0.4%

multithread/8

asio (coroutines)

18.38 GB/s

0.98%

-0.6%

multithread/8

asio (callbacks)

18.48 GB/s

1.82%

baseline
socket_throughput — Windows, part 3 ↑ faster · dotted line = asio (callbacks) · backend = IOCP corosio asio (coroutines) -6% -4% -2% +2% +4% 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. unidirectional/1048576 — One writer streams to one reader as fast as possible; N is the write/read chunk size in bytes. +1% 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. 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. +2% 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. -3% unidirectional_lockless/16384 — Same as unidirectional, with the context in single-threaded lockless mode; N is the chunk size in bytes. -4% 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)

283.5 MB/s

0.40%

-2.7%

unidirectional/1024

asio (coroutines)

279.4 MB/s

0.56%

-4.1%

unidirectional/1024

asio (callbacks)

291.4 MB/s

0.32%

baseline

unidirectional/1048576

corosio (IOCP)

3.93 GB/s

3.90%

+0.5%

unidirectional/1048576

asio (coroutines)

3.94 GB/s

2.24%

+0.6%

unidirectional/1048576

asio (callbacks)

3.91 GB/s

5.15%

baseline

unidirectional/16384

corosio (IOCP)

3.78 GB/s

1.35%

-2.3%

unidirectional/16384

asio (coroutines)

3.73 GB/s

0.73%

-3.5%

unidirectional/16384

asio (callbacks)

3.87 GB/s

0.66%

baseline

unidirectional/262144

corosio (IOCP)

8.75 GB/s

7.22%

-1.2%

unidirectional/262144

asio (coroutines)

8.54 GB/s

4.50%

-3.6%

unidirectional/262144

asio (callbacks)

8.86 GB/s

3.82%

baseline

unidirectional/4096

corosio (IOCP)

1.09 GB/s

0.43%

-2.5%

unidirectional/4096

asio (coroutines)

1.07 GB/s

0.12%

-4.0%

unidirectional/4096

asio (callbacks)

1.12 GB/s

0.44%

baseline

unidirectional/65536

corosio (IOCP)

9.27 GB/s

3.59%

-1.9%

unidirectional/65536

asio (coroutines)

9.23 GB/s

1.69%

-2.4%

unidirectional/65536

asio (callbacks)

9.45 GB/s

1.78%

baseline

unidirectional_lockless/1024

corosio (IOCP)

282.6 MB/s

0.33%

-2.6%

unidirectional_lockless/1024

asio (coroutines)

279.3 MB/s

0.50%

-3.8%

unidirectional_lockless/1024

asio (callbacks)

290.2 MB/s

0.42%

baseline

unidirectional_lockless/1048576

corosio (IOCP)

4.37 GB/s

4.83%

+1.7%

unidirectional_lockless/1048576

asio (coroutines)

4.29 GB/s

3.21%

-0.1%

unidirectional_lockless/1048576

asio (callbacks)

4.30 GB/s

7.83%

baseline

unidirectional_lockless/16384

corosio (IOCP)

3.78 GB/s

0.69%

-2.7%

unidirectional_lockless/16384

asio (coroutines)

3.72 GB/s

0.87%

-4.1%

unidirectional_lockless/16384

asio (callbacks)

3.88 GB/s

0.75%

baseline

unidirectional_lockless/262144

corosio (IOCP)

9.77 GB/s

1.43%

-0.7%

unidirectional_lockless/262144

asio (coroutines)

9.76 GB/s

0.85%

-0.8%

unidirectional_lockless/262144

asio (callbacks)

9.84 GB/s

1.23%

baseline

unidirectional_lockless/4096

corosio (IOCP)

1.09 GB/s

0.35%

-2.4%

unidirectional_lockless/4096

asio (coroutines)

1.07 GB/s

0.30%

-3.9%

unidirectional_lockless/4096

asio (callbacks)

1.12 GB/s

0.24%

baseline

unidirectional_lockless/65536

corosio (IOCP)

9.35 GB/s

2.30%

-1.8%

unidirectional_lockless/65536

asio (coroutines)

9.26 GB/s

1.66%

-2.8%

unidirectional_lockless/65536

asio (callbacks)

9.52 GB/s

0.82%

baseline
Edit this Page