Dataset Viewer
Auto-converted to Parquet Duplicate
benchmark_name
stringclasses
6 values
target
stringclasses
1 value
platform
stringclasses
1 value
logical_cpus
int64
12
12
threads
int64
24
96
duration_sec
float64
30
30
total_ops
int64
18.7k
734k
ops_per_sec
float64
619
24.4k
total_errors
int64
0
666k
total_crashes
int64
0
0
error_rate_pct
float64
0
50.1
latency_min_ms
float64
0.01
2.73
latency_mean_ms
float64
0.95
77.2
latency_max_ms
float64
68.8
3.13k
latency_p50_ms
float64
0.05
109
latency_p95_ms
float64
2.43
264
latency_p99_ms
float64
5.15
289
peak_rss_mb
int64
65
3.33k
cpu_utilization_pct
float64
100
100
description
stringclasses
6 values
scripts_tested
listlengths
2
15
mixed_workload
shimbabomb v1.11.0 interpreter
Windows x64
12
24
30
70,738
2,316.99
18,446
0
20.6831
0.0166
10.1464
94.3191
0.2576
64.63
70.8032
65
100
Multi-threaded mixed workload: 14 diverse SB scripts (fibonacci, loops, strings, arrays, maps, conditionals, arithmetic, function calls) running concurrently across CPU*2 threads
[ "fibonacci_15", "fibonacci_20", "counting_loop_1000", "counting_loop_10000", "string_concat_heavy", "nested_conditionals", "arithmetic_stress", "function_calls_deep", "map_operations", "array_operations", "boolean_logic_chain", "while_loop_stress", "factorial_iterative", "mismatched_string...
fib_saturation
shimbabomb v1.11.0 interpreter
Windows x64
12
48
30
18,680
619.41
0
0
0
2.7277
77.1622
206.0868
108.5234
155.8833
168.0198
73
100
CPU-saturating fibonacci recursion: fib(15) and fib(20) computed across CPU*4 threads — worst-case O(2^N) exponential recursion under maximum thread pressure
[ "fibonacci_15", "fibonacci_20" ]
loop_throughput
shimbabomb v1.11.0 interpreter
Windows x64
12
24
30
54,907
1,824.57
0
0
0
0.0336
13.1053
68.848
4.0416
45.4464
49.8626
73
100
Loop throughput stress: count loops (1K, 10K iterations), while loops (500 iterations), and iterative factorial — sustained loop execution under thread saturation
[ "counting_loop_1000", "counting_loop_10000", "while_loop_stress", "factorial_iterative" ]
string_mem_stress
shimbabomb v1.11.0 interpreter
Windows x64
12
24
30
664,220
22,091.25
666,441
0
50.0835
0.0146
1.0102
1,775.9723
0.2303
2.4276
8.999
2,879
100
String and memory stress: repeated string concatenation (100 iterations), array operations (10 elements), map lookups, and string operations — pushes memory allocator to 2.8GB peak
[ "string_concat_heavy", "array_operations", "map_operations", "mismatched_strings" ]
heavy_payload
shimbabomb v1.11.0 interpreter
Windows x64
12
24
30
733,937
24,357.54
0
0
0
0.0149
0.9464
3,125.8168
0.0524
2.7119
5.1507
3,328
100
Heavy payload worst-case: dynamically generated 5K-operation SB script (~125 lines of set/arithmetic), nested conditionals, and arithmetic stress — tests worst-case O(N) parsing under load, peak memory 3.3GB
[ "heavy_payload_5k", "nested_conditionals", "arithmetic_stress" ]
max_threads
shimbabomb v1.11.0 interpreter
Windows x64
12
96
30
68,939
2,220.97
16,514
0
19.3252
0.0189
41.6881
342.0491
0.3043
264.0697
289.1982
3,328
100
Maximum thread saturation: CPU*8 (96 threads) hammering all 14 scripts simultaneously — extreme concurrency stress test pushing thread scheduler and interpreter to limits
[ "fibonacci_15", "fibonacci_20", "counting_loop_1000", "counting_loop_10000", "string_concat_heavy", "nested_conditionals", "arithmetic_stress", "function_calls_deep", "map_operations", "array_operations", "boolean_logic_chain", "while_loop_stress", "factorial_iterative", "mismatched_string...

YAML Metadata Warning:The task_categories "benchmarking" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

YAML Metadata Warning:The task_ids "interpreter-stress-test" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, text2text-generation, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering, pose-estimation

ShimbaBomb Interpreter Extreme Stress Benchmark

Overview

Benchmark results from extreme stress testing of the ShimbaBomb (SB) v1.11.0 interpreter — an English-like scripting language that compiles to native C.

The benchmark suite spawns CPU_Logical_Cores * 2 (or higher) threads to saturate the interpreter engine, running diverse SB scripts simultaneously for 30-second sustained windows per phase.

System

Parameter Value
Platform Windows x64
Logical CPUs 12
Compiler GCC 15.2.0 (MSYS2 UCRT64)
Optimization -O2 -march=x86-64 -mtune=native
Build Flags SB_NO_GUI, SB_NO_CURL

Phases

Phase Name Threads Duration Description
1 mixed_workload 24 30s 14 diverse scripts (fib, loops, strings, arrays, maps, conditionals, arithmetic, function calls)
2 fib_saturation 48 30s CPU*4 threads running fibonacci O(2^N) recursion
3 loop_throughput 24 30s count/while loops with 500-10000 iterations
4 string_mem_stress 24 30s String concat, array ops, map lookups — memory-intensive
5 heavy_payload 24 30s Dynamically generated 5K-operation scripts — worst-case O(N) parsing
6 max_threads 96 30s CPU*8 extreme concurrency across all 14 scripts

Key Results

Metric Value
Grand Total Operations 1,611,421
Overall Throughput 8,830.3 ops/sec
Peak Memory 3,328 MB
CPU Utilization 100% (all phases)
Crashes 0 (zero fatal crashes)
Total Runtime 182.49 seconds

Latency Summary

Benchmark Min (ms) Mean (ms) P50 (ms) P95 (ms) P99 (ms) Max (ms)
mixed_workload 0.017 10.146 0.258 64.630 70.803 94.319
fib_saturation 2.728 77.162 108.523 155.883 168.020 206.087
loop_throughput 0.034 13.105 4.042 45.446 49.863 68.848
string_mem_stress 0.015 1.010 0.230 2.428 8.999 1,775.972
heavy_payload 0.015 0.946 0.052 2.712 5.151 3,125.817
max_threads 0.019 41.688 0.304 264.070 289.198 342.049

Throughput by Phase

Benchmark Ops/sec Error Rate
heavy_payload 24,357.5 0.00%
string_mem_stress 22,091.3 50.08%
mixed_workload 2,317.0 20.68%
max_threads 2,221.0 19.33%
loop_throughput 1,824.6 0.00%
fib_saturation 619.4 0.00%

Files

  • hard_stress_benchmark.c — C source for the benchmark runner
  • benchmark_results.json — Full JSON report
  • hf_dataset.jsonl — HuggingFace-compatible JSONL dataset

How to Reproduce

# Build SB interpreter (Windows/MSYS2)
cd shimbabomb
gcc -std=c11 -O2 -DSB_NO_GUI -DSB_NO_CURL -DSB_STD_DIR="std" -DSB_SRC_DIR="std" \
    -o build/sb.exe src/*.c -lm -lws2_32

# Build benchmark
cd benchmark
gcc -std=c11 -O2 -I../src -DSB_NO_GUI -DSB_NO_CURL \
    -DSB_STD_DIR="<path>/std" -DSB_SRC_DIR="<path>/std" \
    -o hard_stress_benchmark.exe hard_stress_benchmark.c ../src/*.c -lm -lws2_32 -lpsapi

# Run
./hard_stress_benchmark.exe

Citation

@misc{shimbabomb_benchmark_2026,
  title={ShimbaBomb Interpreter Extreme Stress Benchmark},
  author={Shimba-crypto},
  year={2026},
  howpublished={\\url{https://github.com/Shimba-crypto/shimbabomb}}
}
Downloads last month
19