fork: nascheme
ref: gh_129201_gc_mark_pr
machine: linux-x86_64
commit hash: 1b4e8c3
commit date: 2025-01-22
overall geometric mean: 1.009x slower
HPT reliability: 91.45%
HPT 99th percentile: 1.00x faster
Memory change: 1.00x
Benchmarks with tag 'apps':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
2to3
306 ms
307 ms: 1.00x slower
html5lib
69.2 ms
71.3 ms: 1.03x slower
Geometric mean
(ref)
1.01x slower
Benchmark hidden because not significant (2): docutils, sphinx
Benchmarks with tag 'asyncio':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
async_tree_none_tg
244 ms
240 ms: 1.01x faster
async_tree_io
615 ms
608 ms: 1.01x faster
async_generators
374 ms
370 ms: 1.01x faster
async_tree_memoization
354 ms
350 ms: 1.01x faster
coroutines
23.7 ms
24.7 ms: 1.04x slower
async_tree_cpu_io_mixed
536 ms
621 ms: 1.16x slower
async_tree_cpu_io_mixed_tg
499 ms
583 ms: 1.17x slower
Geometric mean
(ref)
1.03x slower
Benchmark hidden because not significant (4): async_tree_none, async_tree_memoization_tg, async_tree_io_tg, asyncio_websockets
Benchmarks with tag 'math':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
float
77.8 ms
75.3 ms: 1.03x faster
nbody
132 ms
135 ms: 1.02x slower
pidigits
215 ms
221 ms: 1.03x slower
Geometric mean
(ref)
1.01x slower
Benchmarks with tag 'regex':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
regex_compile
151 ms
150 ms: 1.01x faster
regex_v8
24.2 ms
24.1 ms: 1.00x faster
regex_dna
182 ms
186 ms: 1.03x slower
regex_effbot
2.85 ms
2.95 ms: 1.04x slower
Geometric mean
(ref)
1.01x slower
Benchmarks with tag 'serialize':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
unpickle_pure_python
254 us
248 us: 1.02x faster
pickle_pure_python
380 us
373 us: 1.02x faster
tomli_loads
2.33 sec
2.31 sec: 1.01x faster
json_loads
30.4 us
30.5 us: 1.00x slower
json_dumps
12.9 ms
13.0 ms: 1.00x slower
Geometric mean
(ref)
1.01x faster
Benchmarks with tag 'startup':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
python_startup
15.2 ms
15.5 ms: 1.02x slower
python_startup_no_site
9.57 ms
9.78 ms: 1.02x slower
Geometric mean
(ref)
1.02x slower
Benchmarks with tag 'template':
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
django_template
44.1 ms
43.2 ms: 1.02x faster
genshi_text
28.4 ms
28.1 ms: 1.01x faster
genshi_xml
60.6 ms
60.0 ms: 1.01x faster
mako
15.9 ms
15.8 ms: 1.01x faster
Geometric mean
(ref)
1.01x faster
Benchmark
bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104
bm-20250122-vultr-x86_64-nascheme-gh_129201_gc_mark_pr-3.14.0a4+-1b4e8c3
deepcopy_reduce
3.38 us
3.25 us: 1.04x faster
sqlglot_parse
1.62 ms
1.57 ms: 1.03x faster
deepcopy_memo
39.0 us
37.7 us: 1.03x faster
float
77.8 ms
75.3 ms: 1.03x faster
logging_silent
118 ns
114 ns: 1.03x faster
deepcopy
319 us
312 us: 1.02x faster
chaos
70.1 ms
68.6 ms: 1.02x faster
unpickle_pure_python
254 us
248 us: 1.02x faster
mdp
2.77 sec
2.72 sec: 1.02x faster
django_template
44.1 ms
43.2 ms: 1.02x faster
pickle_pure_python
380 us
373 us: 1.02x faster
sqlglot_transpile
1.95 ms
1.91 ms: 1.02x faster
thrift
933 us
916 us: 1.02x faster
pycparser
1.20 sec
1.18 sec: 1.02x faster
richards
56.6 ms
55.7 ms: 1.02x faster
logging_format
8.43 us
8.30 us: 1.02x faster
scimark_lu
137 ms
135 ms: 1.02x faster
nqueens
97.4 ms
95.9 ms: 1.02x faster
logging_simple
7.33 us
7.22 us: 1.02x faster
scimark_fft
387 ms
382 ms: 1.01x faster
async_tree_none_tg
244 ms
240 ms: 1.01x faster
async_tree_io
615 ms
608 ms: 1.01x faster
regex_compile
151 ms
150 ms: 1.01x faster
genshi_text
28.4 ms
28.1 ms: 1.01x faster
genshi_xml
60.6 ms
60.0 ms: 1.01x faster
async_generators
374 ms
370 ms: 1.01x faster
typing_runtime_protocols
204 us
202 us: 1.01x faster
sqlite_synth
2.06 us
2.04 us: 1.01x faster
async_tree_memoization
354 ms
350 ms: 1.01x faster
sqlglot_optimize
62.2 ms
61.7 ms: 1.01x faster
pyflate
498 ms
494 ms: 1.01x faster
pprint_safe_repr
832 ms
825 ms: 1.01x faster
tomli_loads
2.33 sec
2.31 sec: 1.01x faster
pathlib
18.9 ms
18.8 ms: 1.01x faster
sqlglot_normalize
122 ms
121 ms: 1.01x faster
bpe_tokeniser
4.72 sec
4.69 sec: 1.01x faster
go
140 ms
139 ms: 1.01x faster
scimark_sparse_mat_mult
5.43 ms
5.40 ms: 1.01x faster
scimark_monte_carlo
82.3 ms
81.8 ms: 1.01x faster
mako
15.9 ms
15.8 ms: 1.01x faster
k_core
2.32 sec
2.31 sec: 1.00x faster
pprint_pformat
1.71 sec
1.70 sec: 1.00x faster
regex_v8
24.2 ms
24.1 ms: 1.00x faster
raytrace
331 ms
331 ms: 1.00x faster
json_loads
30.4 us
30.5 us: 1.00x slower
sqlalchemy_declarative
163 ms
163 ms: 1.00x slower
json_dumps
12.9 ms
13.0 ms: 1.00x slower
2to3
306 ms
307 ms: 1.00x slower
bench_thread_pool
3.30 ms
3.32 ms: 1.00x slower
sympy_str
334 ms
336 ms: 1.01x slower
sympy_sum
186 ms
188 ms: 1.01x slower
crypto_pyaes
87.2 ms
88.1 ms: 1.01x slower
bench_mp_pool
94.3 ms
95.2 ms: 1.01x slower
telco
8.54 ms
8.63 ms: 1.01x slower
spectral_norm
106 ms
108 ms: 1.01x slower
hexiom
7.31 ms
7.43 ms: 1.02x slower
python_startup
15.2 ms
15.5 ms: 1.02x slower
python_startup_no_site
9.57 ms
9.78 ms: 1.02x slower
nbody
132 ms
135 ms: 1.02x slower
regex_dna
182 ms
186 ms: 1.03x slower
pidigits
215 ms
221 ms: 1.03x slower
html5lib
69.2 ms
71.3 ms: 1.03x slower
regex_effbot
2.85 ms
2.95 ms: 1.04x slower
coroutines
23.7 ms
24.7 ms: 1.04x slower
async_tree_cpu_io_mixed
536 ms
621 ms: 1.16x slower
async_tree_cpu_io_mixed_tg
499 ms
583 ms: 1.17x slower
create_gc_cycles
1.33 ms
1.71 ms: 1.28x slower
gc_traversal
3.24 ms
5.43 ms: 1.68x slower
Geometric mean
(ref)
1.01x slower
Benchmark hidden because not significant (24): sqlalchemy_imperative, generators, async_tree_none, scimark_sor, async_tree_memoization_tg, many_optionals, async_tree_io_tg, deltablue, coverage, fannkuch, sympy_integrate, comprehensions, dulwich_log, docutils, subparsers, richards_super, asyncio_websockets, sphinx, pylint, json, shortest_path, sympy_expand, connected_components, meteor_contest
Ignored benchmarks (4) of results/bm-20250117-3.14.0a4+-3829104-NOGIL/bm-20250117-vultr-x86_64-python-3829104ab412a47bf3f3-3.14.0a4+-3829104.json: xml_etree_generate, xml_etree_iterparse, xml_etree_parse, xml_etree_process
Geometric mean (including insignificant results): 1.009x slower
Reliability score: 91.45% likely to be faster
90% likely to have a speedup of 1.00x
95% likely to have a speedup of 1.00x
99% likely to have a speedup of 1.00x