labeled_counter:churn | otel=3,721,890.36

Cache Benchmarks

fast/otel speedup by case

fast vs otel throughput by case (dumbbell, log scale)

Left dot = OTel, right dot = fast-telemetry (log scale).

fast median total ops/sec by case

case	run_dir	fast_total	otel_total	atomic_total	fast/otel	fast/atomic
counter:uniform	cache_20260406_193818	2,050,476,161.57	10,885,493.13	34,082,756.34	188.37x	60.16x
counter:hotspot	cache_20260406_193836	3,706,335,201.70	14,588,221.03	37,012,841.61	254.06x	100.14x
counter:churn	cache_20260406_193838	2,278,531,313.40	15,649,356.34	35,477,919.40	145.60x	64.22x
distribution:uniform	cache_20260406_193839	2,438,033,121.66	20,745,770.67	0.00	117.52x	0.00x
distribution:hotspot	cache_20260406_193840	3,151,426,682.37	20,751,537.69	0.00	151.86x	0.00x
dynamic_counter:uniform	cache_20260406_193841	6,075,946,905.95	3,485,689.68	0.00	1743.11x	0.00x
dynamic_counter:hotspot	cache_20260406_193845	2,424,242,424.24	3,456,894.06	0.00	701.28x	0.00x
dynamic_counter:churn	cache_20260406_193848	2,738,287,863.61	3,835,149.06	0.00	714.00x	0.00x
dynamic_distribution:uniform	cache_20260406_193851	1,483,202,875.75	3,232,687.95	0.00	458.81x	0.00x
dynamic_distribution:hotspot	cache_20260406_193854	1,519,997,543.68	4,291,505.21	0.00	354.19x	0.00x
dynamic_distribution:churn	cache_20260406_193857	907,976,574.20	3,552,769.13	0.00	255.57x	0.00x
dynamic_gauge:uniform	cache_20260406_193900	404,188,067.41	3,377,212.33	0.00	119.68x	0.00x
dynamic_gauge:hotspot	cache_20260406_193904	1,311,489,685.72	3,704,848.95	0.00	353.99x	0.00x
dynamic_gauge:churn	cache_20260406_193907	446,248,438.47	4,014,185.87	0.00	111.17x	0.00x
dynamic_gauge_i64:uniform	cache_20260406_193910	32,229,521.43	3,570,375.20	0.00	9.03x	0.00x
dynamic_gauge_i64:hotspot	cache_20260406_193913	32,416,473.89	3,855,810.12	0.00	8.41x	0.00x
dynamic_gauge_i64:churn	cache_20260406_193916	30,543,801.91	4,716,215.58	0.00	6.48x	0.00x
dynamic_histogram:uniform	cache_20260406_193919	372,507,681.48	3,625,841.01	0.00	102.74x	0.00x
dynamic_histogram:hotspot	cache_20260406_193922	417,283,916.51	4,548,587.82	0.00	91.74x	0.00x
dynamic_histogram:churn	cache_20260406_193925	197,546,796.71	3,905,704.89	0.00	50.58x	0.00x
labeled_counter:uniform	cache_20260406_193928	5,269,395,194.63	3,892,642.48	0.00	1353.68x	0.00x
labeled_counter:hotspot	cache_20260406_193931	6,054,338,902.52	4,063,163.57	0.00	1490.06x	0.00x
labeled_counter:churn	cache_20260406_193934	8,798,611,227.20	3,721,890.36	0.00	2364.02x	0.00x
labeled_gauge:uniform	cache_20260406_193937	520,787,004.99	4,409,525.25	0.00	118.11x	0.00x
labeled_gauge:hotspot	cache_20260406_193939	1,072,616,110.69	8,258,500.70	0.00	129.88x	0.00x
labeled_gauge:churn	cache_20260406_193942	506,152,921.45	7,702,837.37	0.00	65.71x	0.00x
labeled_histogram:uniform	cache_20260406_193944	275,726,107.47	5,758,604.34	0.00	47.88x	0.00x
labeled_histogram:hotspot	cache_20260406_193947	241,498,743.13	6,188,399.48	0.00	39.02x	0.00x
labeled_histogram:churn	cache_20260406_193950	240,726,267.30	4,464,335.54	0.00	53.92x	0.00x

fast-telemetry harness benchmarks

Metric-path resultsMeasures sustained multi-threaded metric write throughput across counter, distribution, labeled, and dynamic-series contention profiles.fast-telemetry speedup2,364xBest throughput gain versus OTel in this section, from labeled_counter:churn. Peak fast throughput: 8.8B ops/sec.

Metric-path results

Span resultsMeasures end-to-end span throughput across root, lifecycle, and pipeline scenarios from the same local threaded harness run.fast-telemetry speedup324xBest throughput gain versus OTel in this section, from pipeline. Peak fast throughput: 267.3M ops/sec.

Span results