Audio
audio-adpcmCustom adaptive 4-bit differential PCM: saturating predictor and variable quantizer step
2.3× faster
audio-autocorrelationPitch-analysis autocorrelation over 32 positive sample lags
3.5× faster
audio-biquadQ8 second-order low-pass IIR biquad with reset filter history
2.7× faster
audio-fir32-tap fixed-point FIR audio filtering
4× faster
drwavAudio corpus
wago
2.13ms
wazero
17.9ms
8.4× faster
kissfftAudio corpus
wago
1.14ms
wazero
8.30ms
7.2× faster
3D & rendering
bezier-tessellationCubic 3D Bezier tessellation using Bernstein evaluation at 17 samples
2.4× faster
frustum-cullingOrthographic sphere/frustum culling against six clipping planes
2.4× faster
mesh-skinningFixed-point skeletal skinning with four bone influences and 3×4 matrices
2.7× faster
ray-boxFixed-point slab ray/AABB intersection with positive ray directions
2.6× faster
raytracerecursive Whitted ray tracer: n x n frame, 4 spheres + checker plane, depth-4 mirrors (heavy f64 vector math)
7.6× faster
triangle-rasterCPU triangle coverage rasterization into a 128×128 depth buffer
3.8× faster
Bioinformatics
bio-global-alignmentNeedleman-Wunsch global DNA alignment score with linear gap penalty
4.2× faster
bio-kmer-countComplete histogram of overlapping DNA 5-mers
3.7× faster
bio-lcsLongest common subsequence length of two deterministic DNA strings
4.8× faster
bio-local-alignmentSmith-Waterman best local DNA alignment score with linear gap penalty
3.9× faster
bio-reverse-complementReverse complement of a deterministic DNA sequence
2.7× faster
polybench-nussinovPolyBench/C nussinov (small dataset)
5.6× faster
Hashing & cryptography
blake3Hashing & cryptography corpus
wago
1.12ms
wazero
3.56ms
3.2× faster
checksum-crc32IEEE CRC-32 byte-stream checksum using reflected bitwise polynomial reduction
2.5× faster
crypto-aes128AES-128 ECB encryption with key expansion and algebraic S-box setup included
5.2× faster
monocypherHashing & cryptography corpus
5× faster
sha256standards-correct SHA-256 of an n-KiB buffer (i32 rotate/shift/xor/add, 64-round)
3.6× faster
xxhashHashing & cryptography corpus
5.4× faster
Databases & analytics
columnar-queryColumnar range filter with grouped SUM and COUNT over 256 keys
3.6× faster
db-bitmap-indexBitmap filter intersection and set-bit counting
4.5× faster
db-btreeBulk-built eight-way B-tree index and point lookups
6.2× faster
db-hash-joinChained hash equijoin emitting matching left/right row pairs
4.1× faster
db-sort-merge-joinSort-merge equijoin with grouped result multiplicities
4.6× faster
Compilers & development tools
compiler-constant-foldConstant propagation and folding over a mixed literal/variable expression DAG
3.1× faster
compiler-dead-codeBackward reachability elimination of dead nodes in an acyclic IR dependency graph
4.3× faster
compiler-register-allocationLinear-scan register allocation with eight registers and furthest-end spilling
5.5× faster
Compression
compression-lzwLempel–Ziv–Welch dictionary compression with 12-bit code-space limit
3.8× faster
compression-rleByte run-length encoding of mixed repeated and noisy input spans
2.8× faster
lz4Compression corpus
4.2× faster
zlibCompression corpus
wago
1.16ms
wazero
11.3ms
9.7× faster
zstdCompression corpus
wago
3.72ms
wazero
36.6ms
9.8× faster
Integer & graph algorithms
coremarkInteger & graph algorithms corpus
5.2× faster
fannkuchBenchmarks-Game fannkuch-redux: permute + pancake-flip (branch/array churn, no f64)
6.9× faster
graph-dijkstraDense-graph Dijkstra single-source shortest paths
5.4× faster
graph-kruskalKruskal minimum-spanning-tree weight using sorted edges and disjoint sets
4.8× faster
polybench-floyd-warshallPolyBench/C floyd-warshall (small dataset)
4.4× faster
Calls & memory
dispatchcall_indirect through a function table
2.2× faster
fib_recrecursion; call/return heavy
wago
42.1µs
wazero
80.7µs
1.9× faster
linked_listsingly-linked list build + dependent-load pointer chase
3.4× faster
many_funcs300 small functions; per-function compile cost
2× faster
memorylinear-memory load/store loop
2.3× faster
memory_treerecursive call tree with load/store churn at every node
2.3× faster
tinyminimal module; fixed pipeline overhead
wago
30.8µs
wazero
73.6µs
2.4× faster
Documents & layout
document-bidi-reorderBidirectional run reordering from supplied embedding levels (L2 stage only)
4.8× faster
document-edit-distanceTwo-row Levenshtein document-token edit distance
5× faster
document-kerningGlyph positioning with pair-kerning adjustments and proportional advances
2.7× faster
document-layoutProportional glyph advances and greedy word line breaking into 640-unit lines
3.1× faster
document-optimal-wrapDynamic-programming minimum-raggedness line breaking with quadratic slack cost
3.8× faster
document-page-packingFirst-fit decreasing page packing of document block heights
4.3× faster
Text & parsing
fastfloatText & parsing corpus
wago
1.53ms
wazero
13.9ms
9.1× faster
pcre2Text & parsing corpus
wago
11.7ms
wazero
163.6ms
14× faster
text-substitutionLiteral non-overlapping error and warning substitutions over a deterministic record stream
4.9× faster
utf-as (simd)AssemblyScript utf-as SIMD: UTF-8 length and validation of mixed text
6× faster
utf8procText & parsing corpus
wago
1.14ms
wazero
7.91ms
6.9× faster
Files & directories
files-content-chunkingGear-hash content-defined chunk boundaries with 32-byte minimum and 1024-byte maximum
2.9× faster
files-glob-matchWildcard filename matching with star backtracking and single-character wildcards
2.9× faster
files-merkle-treeBinary Merkle content tree with a non-cryptographic FNV fixture hash
3.7× faster
files-path-normalizeLexical path normalization by dot-segment stack reduction
5.1× faster
files-path-trieDirectory-component trie construction with deduplicated prefixes
4.2× faster
Games & physics
game-cellular-automataEight generations of Conway Life on a toroidal game grid
5.2× faster
game-collision-impulseOne-dimensional elastic collision impulse resolution for contacting bodies
2.5× faster
game-flockingLocal-neighbor boid cohesion and separation steering
3.7× faster
game-spatial-hashUniform-grid broadphase pair enumeration for collision candidates
4.4× faster
grid-pathfindingFour-neighbor breadth-first shortest paths over a seeded obstacle grid
6.4× faster
particle-physics64 simulation steps of particle integration, gravity and reflecting wall collisions
2.5× faster
Geospatial
geo-convex-hullAndrew monotone-chain convex hull of two-dimensional map points
4.8× faster
geo-geohashBinary geohash encoding by interleaving quantized longitude/latitude bits
2.7× faster
geo-point-in-polygonEven-odd point-in-polygon classification against an eight-vertex map polygon
3.5× faster
geo-polygon-clippingInteger Sutherland–Hodgman clipping of map polygons to a tile rectangle
5.6× faster
geo-polyline-simplifyRamer–Douglas–Peucker polyline simplification with squared perpendicular-distance tolerance
3.6× faster
map-point-segmentMap geometry classification by exact squared distance to a finite line segment
2.5× faster
Graphics & images
graphics-bresenhamBresenham integer line rasterization into a 128×128 coverage image
3.8× faster
graphics-png-paethPNG Paeth scanline residual filtering with four-byte pixels
2.3× faster
graphics-reed-solomonGF(256) polynomial multiplication used in Reed–Solomon image-code parity construction
4.4× faster
nanosvgGraphics & images corpus
wago
3.80ms
wazero
19.7ms
5.2× faster
polybench-derichePolyBench/C deriche (small dataset)
5.3× faster
qoiGraphics & images corpus
3.6× faster
Hardware design
hardware-boolean-anfBoolean Möbius transform into algebraic normal form for eight-input logic functions
6.1× faster
hardware-prime-implicantsFour-variable Boolean truth-table prime implicant enumeration
3.9× faster
Image editing
image-blur3×3 Gaussian blur with clamped image borders
4.6× faster
image-compositeAlpha compositing and clamped color adjustment
3.2× faster
image-error-diffusionFloyd–Steinberg monochrome error-diffusion dithering
4.3× faster
image-median3×3 median filtering of interior grayscale pixels
6.4× faster
image-resizeBilinear image downsampling at pixel centers
3.3× faster
image-ycocgReversible integer RGB-to-YCoCg lifting color transform
2.7× faster
JSON & serialization
json-as (simd)AssemblyScript json-as SIMD: serialize/deserialize canonical payload
wago
1.89ms
wazero
14.8ms
7.8× faster
serialization-msgpackMessagePack shortest-form unsigned integer encoding
2.7× faster
serialization-protobufPacked Protobuf unsigned-varint stream decoding, with fixture encoding included
3.6× faster
yyjsonJSON & serialization corpus
wago
6.91ms
wazero
61.4ms
8.9× faster
Language runtimes
language-brainfuckBrainfuck tape interpreter with bracket scanning and loop control
3.7× faster
language-forthTiny Forth-style text interpreter: decimal literals, arithmetic and output words
4× faster
language-register-vmEight-register bytecode interpreter with arithmetic, table loads and conditional skips
4.2× faster
Numerical arithmetic
libtommathNumerical arithmetic corpus
wago
2.79ms
wazero
19.7ms
7.1× faster
nbodyBenchmarks-Game N-body: leapfrog solar-system integration (f64 mul/add/div/sqrt)
4.8× faster
numeric-cordicQ16 CORDIC sine/cosine rotation over angles in the convergence interval
3.7× faster
numeric-euclidean-gcdEuclidean greatest-common-divisor reduction of integer pairs
3.4× faster
numeric-integer-sqrtInteger Newton square-root iteration with floor-result correction
3.1× faster
numeric-simpsonComposite Simpson quadrature of a quadratic polynomial over integer sample points
2.5× faster
Machine learning
ml-convolutionValid 3×3 int8 convolution followed by ReLU activation
3.5× faster
ml-decision-treeDepth-eight decision-tree classifier over eight input features
2.5× faster
ml-inferenceInt8 dense inference: 64 input channels, 32 outputs, ReLU activation
3.7× faster
ml-kmeansFive Lloyd iterations of eight-cluster two-dimensional k-means
5.3× faster
ml-knnFive-nearest-neighbor classification against 32 eight-dimensional training vectors
3.7× faster
ml-max-pooling2×2 maximum-pooling reduction of activation planes
3.4× faster
Stencils
polybench-adiPolyBench/C adi (small dataset)
4.8× faster
polybench-fdtd-2dPolyBench/C fdtd-2d (small dataset)
5× faster
polybench-jacobi-2dPolyBench/C jacobi-2d (small dataset)
4.3× faster
polybench-seidel-2dPolyBench/C seidel-2d (small dataset)
3.7× faster
stencil-lattice-boltzmannD2Q9 lattice-Boltzmann fluid collision/streaming with BGK relaxation
6.3× faster
stencil-wave-equationSecond-order two-dimensional wave-equation time stepping on a periodic grid
4.7× faster
Linear algebra
polybench-choleskyPolyBench/C cholesky (small dataset)
7.6× faster
polybench-durbinPolyBench/C durbin (small dataset)
5.8× faster
polybench-gemmPolyBench/C gemm (small dataset)
5.3× faster
polybench-gramschmidtPolyBench/C gramschmidt (small dataset)
5.8× faster
polybench-luPolyBench/C lu (small dataset)
7.2× faster
polybench-trisolvPolyBench/C trisolv (small dataset)
4.7× faster
Statistics
polybench-correlationPolyBench/C correlation (small dataset)
5.4× faster
stats-bootstrapDeterministic bootstrap resampling of the mean, 64 resample replicates
3.4× faster
stats-giniQ16 empirical categorical Gini impurity from class-frequency counts
3× faster
stats-linear-regressionQ8 ordinary least-squares slope and intercept from sufficient statistics
2.4× faster
stats-quickselectLomuto quickselect median without fully sorting observations
3.2× faster
stats-welfordQ8 online Welford mean and variance accumulator
2.5× faster
Search & indexing
search-aho-corasickAho–Corasick multi-pattern automaton with suffix-failure links
5.8× faster
search-bk-treeBK-tree fuzzy lookup using Hamming distance over 16-bit symbol vectors
5.4× faster
search-horspoolBoyer–Moore–Horspool substring search with bad-character shifts
6× faster
search-kmpKnuth–Morris–Pratt substring matching with prefix failure links
5.7× faster
search-rabin-karpRabin–Karp rolling polynomial hash with exact match verification
7.4× faster
Video processing
video-dctSeparable Q10 8×8 cosine transform (unscaled DCT-II coefficients)
4.4× faster
video-deblockingThresholded vertical block-boundary deblocking with clamped pair updates
4.5× faster
video-motion8×8 block motion estimation over a 5×5 displacement search
5.2× faster
video-optical-flowQ6 Lucas–Kanade optical flow from 3×3 patch normal equations
3.2× faster
video-temporal-denoiseTen-frame exponential temporal denoising with integer 3:1 history weighting
5.9× faster
video-yuv420Integer BT.601 RGB-to-YUV420 conversion with 2×2 chroma averaging
3.4× faster
Computer vision
vision-componentsFour-connected binary component labeling by union-find with canonical minimum labels
5.5× faster
vision-dilationBinary 3×3 morphological dilation with cropped image borders
5.6× faster
vision-distance-transformTwo-pass city-block distance transform of a binary image
6.5× faster
vision-fast-cornersFAST-9 corner detection on the 16-sample radius-three ring
4× faster
vision-otsuOtsu histogram-based global threshold selection for grayscale segmentation
3× faster
vision-sobelSobel horizontal/vertical edges and gradient magnitude
3.8× faster
Audio
audio-adpcmCustom adaptive 4-bit differential PCM: saturating predictor and variable quantizer step
wago
9.9 MB
wazero
9.2 MB
1.1× more
audio-autocorrelationPitch-analysis autocorrelation over 32 positive sample lags
wago
10.2 MB
wazero
11.2 MB
1.1× less
audio-biquadQ8 second-order low-pass IIR biquad with reset filter history
wago
9.9 MB
wazero
9.2 MB
1.1× more
audio-fir32-tap fixed-point FIR audio filtering
wago
10.2 MB
wazero
12.4 MB
1.2× less
drwavAudio corpus
wago
12.6 MB
wazero
24.6 MB
2× less
kissfftAudio corpus
wago
11.7 MB
wazero
16.5 MB
1.4× less
3D & rendering
bezier-tessellationCubic 3D Bezier tessellation using Bernstein evaluation at 17 samples
wago
10.1 MB
wazero
10 MB
parity
frustum-cullingOrthographic sphere/frustum culling against six clipping planes
wago
9.9 MB
wazero
9.3 MB
1.1× more
mesh-skinningFixed-point skeletal skinning with four bone influences and 3×4 matrices
parity
ray-boxFixed-point slab ray/AABB intersection with positive ray directions
wago
9.9 MB
wazero
9.7 MB
parity
raytracerecursive Whitted ray tracer: n x n frame, 4 spheres + checker plane, depth-4 mirrors (heavy f64 vector math)
wago
10.3 MB
wazero
16.4 MB
1.6× less
triangle-rasterCPU triangle coverage rasterization into a 128×128 depth buffer
wago
10.2 MB
wazero
11.1 MB
1.1× less
Bioinformatics
bio-global-alignmentNeedleman-Wunsch global DNA alignment score with linear gap penalty
wago
10.1 MB
wazero
11.5 MB
1.1× less
bio-kmer-countComplete histogram of overlapping DNA 5-mers
wago
10.2 MB
wazero
11 MB
1.1× less
bio-lcsLongest common subsequence length of two deterministic DNA strings
wago
10.5 MB
wazero
11.3 MB
1.1× less
bio-local-alignmentSmith-Waterman best local DNA alignment score with linear gap penalty
wago
10.1 MB
wazero
10.8 MB
1.1× less
bio-reverse-complementReverse complement of a deterministic DNA sequence
wago
9.9 MB
wazero
9.2 MB
1.1× more
polybench-nussinovPolyBench/C nussinov (small dataset)
wago
10.3 MB
wazero
12.6 MB
1.2× less
Hashing & cryptography
blake3Hashing & cryptography corpus
wago
11.4 MB
wazero
14.9 MB
1.3× less
checksum-crc32IEEE CRC-32 byte-stream checksum using reflected bitwise polynomial reduction
wago
9.9 MB
wazero
9.2 MB
1.1× more
crypto-aes128AES-128 ECB encryption with key expansion and algebraic S-box setup included
wago
10.2 MB
wazero
13.7 MB
1.3× less
monocypherHashing & cryptography corpus
wago
13 MB
wazero
13.8 MB
1.1× less
sha256standards-correct SHA-256 of an n-KiB buffer (i32 rotate/shift/xor/add, 64-round)
wago
10.2 MB
wazero
13.6 MB
1.3× less
xxhashHashing & cryptography corpus
wago
10.2 MB
wazero
12.1 MB
1.2× less
Databases & analytics
columnar-queryColumnar range filter with grouped SUM and COUNT over 256 keys
wago
10.2 MB
wazero
11.3 MB
1.1× less
db-bitmap-indexBitmap filter intersection and set-bit counting
1.2× less
db-btreeBulk-built eight-way B-tree index and point lookups
wago
10.1 MB
wazero
12.3 MB
1.2× less
db-hash-joinChained hash equijoin emitting matching left/right row pairs
wago
9.9 MB
wazero
11.7 MB
1.2× less
db-sort-merge-joinSort-merge equijoin with grouped result multiplicities
wago
9.9 MB
wazero
12.1 MB
1.2× less
Compilers & development tools
compiler-constant-foldConstant propagation and folding over a mixed literal/variable expression DAG
wago
9.9 MB
wazero
11.6 MB
1.2× less
compiler-dead-codeBackward reachability elimination of dead nodes in an acyclic IR dependency graph
wago
10.2 MB
wazero
11.5 MB
1.1× less
compiler-register-allocationLinear-scan register allocation with eight registers and furthest-end spilling
wago
10.1 MB
wazero
12.4 MB
1.2× less
Compression
compression-lzwLempel–Ziv–Welch dictionary compression with 12-bit code-space limit
wago
9.9 MB
wazero
11.7 MB
1.2× less
compression-rleByte run-length encoding of mixed repeated and noisy input spans
wago
9.9 MB
wazero
9.5 MB
parity
lz4Compression corpus
wago
10.3 MB
wazero
14.2 MB
1.4× less
zlibCompression corpus
wago
10.7 MB
wazero
32.3 MB
3× less
zstdCompression corpus
wago
14 MB
wazero
54.8 MB
3.9× less
Integer & graph algorithms
coremarkInteger & graph algorithms corpus
wago
11.1 MB
wazero
12.2 MB
1.1× less
fannkuchBenchmarks-Game fannkuch-redux: permute + pancake-flip (branch/array churn, no f64)
wago
10.2 MB
wazero
12 MB
1.2× less
graph-dijkstraDense-graph Dijkstra single-source shortest paths
wago
10.5 MB
wazero
12.4 MB
1.2× less
graph-kruskalKruskal minimum-spanning-tree weight using sorted edges and disjoint sets
wago
10.5 MB
wazero
11.9 MB
1.1× less
polybench-floyd-warshallPolyBench/C floyd-warshall (small dataset)
wago
10.7 MB
wazero
11.7 MB
1.1× less
Calls & memory
dispatchcall_indirect through a function table
1.2× more
fib_recrecursion; call/return heavy
wago
9.7 MB
wazero
7.9 MB
1.2× more
linked_listsingly-linked list build + dependent-load pointer chase
wago
10 MB
wazero
10.4 MB
parity
many_funcs300 small functions; per-function compile cost
wago
11.2 MB
wazero
9.9 MB
1.1× more
memorylinear-memory load/store loop
wago
9.8 MB
wazero
8.8 MB
1.1× more
memory_treerecursive call tree with load/store churn at every node
wago
9.7 MB
wazero
9.2 MB
1.1× more
tinyminimal module; fixed pipeline overhead
wago
9.6 MB
wazero
7.8 MB
1.2× more
Documents & layout
document-bidi-reorderBidirectional run reordering from supplied embedding levels (L2 stage only)
wago
10.6 MB
wazero
12.7 MB
1.2× less
document-edit-distanceTwo-row Levenshtein document-token edit distance
wago
10.5 MB
wazero
12.1 MB
1.2× less
document-kerningGlyph positioning with pair-kerning adjustments and proportional advances
wago
9.9 MB
wazero
9.2 MB
1.1× more
document-layoutProportional glyph advances and greedy word line breaking into 640-unit lines
wago
9.9 MB
wazero
9.9 MB
parity
document-optimal-wrapDynamic-programming minimum-raggedness line breaking with quadratic slack cost
wago
9.9 MB
wazero
12.1 MB
1.2× less
document-page-packingFirst-fit decreasing page packing of document block heights
wago
10 MB
wazero
11.8 MB
1.2× less
Text & parsing
fastfloatText & parsing corpus
wago
10.8 MB
wazero
28.4 MB
2.6× less
pcre2Text & parsing corpus
wago
16.2 MB
wazero
136.5 MB
8.4× less
text-substitutionLiteral non-overlapping error and warning substitutions over a deterministic record stream
wago
10 MB
wazero
10.9 MB
1.1× less
utf-as (simd)AssemblyScript utf-as SIMD: UTF-8 length and validation of mixed text
wago
11.5 MB
wazero
12.8 MB
1.1× less
utf8procText & parsing corpus
wago
14.7 MB
wazero
21 MB
1.4× less
Files & directories
files-content-chunkingGear-hash content-defined chunk boundaries with 32-byte minimum and 1024-byte maximum
wago
9.9 MB
wazero
9.3 MB
1.1× more
files-glob-matchWildcard filename matching with star backtracking and single-character wildcards
parity
files-merkle-treeBinary Merkle content tree with a non-cryptographic FNV fixture hash
wago
10.5 MB
wazero
11.9 MB
1.1× less
files-path-normalizeLexical path normalization by dot-segment stack reduction
wago
10.1 MB
wazero
12 MB
1.2× less
files-path-trieDirectory-component trie construction with deduplicated prefixes
wago
10.4 MB
wazero
11.8 MB
1.1× less
Games & physics
game-cellular-automataEight generations of Conway Life on a toroidal game grid
wago
10.1 MB
wazero
12.1 MB
1.2× less
game-collision-impulseOne-dimensional elastic collision impulse resolution for contacting bodies
wago
9.9 MB
wazero
9.5 MB
parity
game-flockingLocal-neighbor boid cohesion and separation steering
wago
10.2 MB
wazero
11 MB
1.1× less
game-spatial-hashUniform-grid broadphase pair enumeration for collision candidates
wago
9.9 MB
wazero
11.7 MB
1.2× less
grid-pathfindingFour-neighbor breadth-first shortest paths over a seeded obstacle grid
wago
10.5 MB
wazero
12.1 MB
1.1× less
particle-physics64 simulation steps of particle integration, gravity and reflecting wall collisions
1.1× more
Geospatial
geo-convex-hullAndrew monotone-chain convex hull of two-dimensional map points
wago
10 MB
wazero
11.8 MB
1.2× less
geo-geohashBinary geohash encoding by interleaving quantized longitude/latitude bits
parity
geo-point-in-polygonEven-odd point-in-polygon classification against an eight-vertex map polygon
wago
10 MB
wazero
10.1 MB
parity
geo-polygon-clippingInteger Sutherland–Hodgman clipping of map polygons to a tile rectangle
wago
10.1 MB
wazero
12.1 MB
1.2× less
geo-polyline-simplifyRamer–Douglas–Peucker polyline simplification with squared perpendicular-distance tolerance
wago
10 MB
wazero
12.3 MB
1.2× less
map-point-segmentMap geometry classification by exact squared distance to a finite line segment
wago
9.9 MB
wazero
9.5 MB
parity
Graphics & images
graphics-bresenhamBresenham integer line rasterization into a 128×128 coverage image
wago
10.2 MB
wazero
11.5 MB
1.1× less
graphics-png-paethPNG Paeth scanline residual filtering with four-byte pixels
1.1× more
graphics-reed-solomonGF(256) polynomial multiplication used in Reed–Solomon image-code parity construction
wago
10.2 MB
wazero
12 MB
1.2× less
nanosvgGraphics & images corpus
wago
14.1 MB
wazero
17.1 MB
1.2× less
polybench-derichePolyBench/C deriche (small dataset)
wago
10.5 MB
wazero
14.5 MB
1.4× less
qoiGraphics & images corpus
wago
10.4 MB
wazero
13.7 MB
1.3× less
Hardware design
hardware-boolean-anfBoolean Möbius transform into algebraic normal form for eight-input logic functions
wago
10.9 MB
wazero
12.1 MB
1.1× less
hardware-prime-implicantsFour-variable Boolean truth-table prime implicant enumeration
wago
10.1 MB
wazero
10.9 MB
1.1× less
Image editing
image-blur3×3 Gaussian blur with clamped image borders
wago
10.1 MB
wazero
12.1 MB
1.2× less
image-compositeAlpha compositing and clamped color adjustment
wago
10.2 MB
wazero
11.3 MB
1.1× less
image-error-diffusionFloyd–Steinberg monochrome error-diffusion dithering
wago
10.8 MB
wazero
11.7 MB
1.1× less
image-median3×3 median filtering of interior grayscale pixels
wago
10.5 MB
wazero
12.7 MB
1.2× less
image-resizeBilinear image downsampling at pixel centers
wago
10.3 MB
wazero
11.2 MB
1.1× less
image-ycocgReversible integer RGB-to-YCoCg lifting color transform
wago
9.9 MB
wazero
9.1 MB
1.1× more
JSON & serialization
json-as (simd)AssemblyScript json-as SIMD: serialize/deserialize canonical payload
wago
13.3 MB
wazero
19.3 MB
1.5× less
serialization-msgpackMessagePack shortest-form unsigned integer encoding
wago
9.9 MB
wazero
9.2 MB
1.1× more
serialization-protobufPacked Protobuf unsigned-varint stream decoding, with fixture encoding included
wago
10.2 MB
wazero
11.3 MB
1.1× less
yyjsonJSON & serialization corpus
wago
15.6 MB
wazero
86.4 MB
5.5× less
Language runtimes
language-brainfuckBrainfuck tape interpreter with bracket scanning and loop control
wago
10 MB
wazero
10.6 MB
1.1× less
language-forthTiny Forth-style text interpreter: decimal literals, arithmetic and output words
wago
10.3 MB
wazero
11.6 MB
1.1× less
language-register-vmEight-register bytecode interpreter with arithmetic, table loads and conditional skips
wago
10.4 MB
wazero
11.9 MB
1.1× less
Numerical arithmetic
libtommathNumerical arithmetic corpus
wago
12.8 MB
wazero
15.2 MB
1.2× less
nbodyBenchmarks-Game N-body: leapfrog solar-system integration (f64 mul/add/div/sqrt)
wago
10.1 MB
wazero
12.2 MB
1.2× less
numeric-cordicQ16 CORDIC sine/cosine rotation over angles in the convergence interval
wago
10.1 MB
wazero
10.6 MB
1.1× less
numeric-euclidean-gcdEuclidean greatest-common-divisor reduction of integer pairs
wago
9.9 MB
wazero
9.2 MB
1.1× more
numeric-integer-sqrtInteger Newton square-root iteration with floor-result correction
wago
9.9 MB
wazero
9.1 MB
1.1× more
numeric-simpsonComposite Simpson quadrature of a quadratic polynomial over integer sample points
wago
9.9 MB
wazero
9.3 MB
1.1× more
Machine learning
ml-convolutionValid 3×3 int8 convolution followed by ReLU activation
wago
10.2 MB
wazero
11.5 MB
1.1× less
ml-decision-treeDepth-eight decision-tree classifier over eight input features
wago
10 MB
wazero
10.1 MB
parity
ml-inferenceInt8 dense inference: 64 input channels, 32 outputs, ReLU activation
wago
9.9 MB
wazero
11.8 MB
1.2× less
ml-kmeansFive Lloyd iterations of eight-cluster two-dimensional k-means
wago
10.1 MB
wazero
12.9 MB
1.3× less
ml-knnFive-nearest-neighbor classification against 32 eight-dimensional training vectors
wago
10.2 MB
wazero
11.5 MB
1.1× less
ml-max-pooling2×2 maximum-pooling reduction of activation planes
wago
10.2 MB
wazero
11.4 MB
1.1× less
Stencils
polybench-adiPolyBench/C adi (small dataset)
wago
10.7 MB
wazero
13.8 MB
1.3× less
polybench-fdtd-2dPolyBench/C fdtd-2d (small dataset)
wago
10.5 MB
wazero
13.6 MB
1.3× less
polybench-jacobi-2dPolyBench/C jacobi-2d (small dataset)
wago
10.2 MB
wazero
12 MB
1.2× less
polybench-seidel-2dPolyBench/C seidel-2d (small dataset)
wago
10.1 MB
wazero
11.6 MB
1.1× less
stencil-lattice-boltzmannD2Q9 lattice-Boltzmann fluid collision/streaming with BGK relaxation
wago
10.4 MB
wazero
12.6 MB
1.2× less
stencil-wave-equationSecond-order two-dimensional wave-equation time stepping on a periodic grid
wago
10.1 MB
wazero
12.2 MB
1.2× less
Linear algebra
polybench-choleskyPolyBench/C cholesky (small dataset)
wago
10.5 MB
wazero
12.7 MB
1.2× less
polybench-durbinPolyBench/C durbin (small dataset)
wago
10.2 MB
wazero
13.2 MB
1.3× less
polybench-gemmPolyBench/C gemm (small dataset)
wago
10.3 MB
wazero
12.4 MB
1.2× less
polybench-gramschmidtPolyBench/C gramschmidt (small dataset)
wago
10.3 MB
wazero
12.7 MB
1.2× less
polybench-luPolyBench/C lu (small dataset)
wago
10.6 MB
wazero
12.9 MB
1.2× less
polybench-trisolvPolyBench/C trisolv (small dataset)
wago
10.2 MB
wazero
12.3 MB
1.2× less
Statistics
polybench-correlationPolyBench/C correlation (small dataset)
wago
10.3 MB
wazero
12.9 MB
1.3× less
stats-bootstrapDeterministic bootstrap resampling of the mean, 64 resample replicates
wago
10.2 MB
wazero
11.4 MB
1.1× less
stats-giniQ16 empirical categorical Gini impurity from class-frequency counts
wago
10.1 MB
wazero
10.6 MB
1.1× less
stats-linear-regressionQ8 ordinary least-squares slope and intercept from sufficient statistics
wago
9.9 MB
wazero
9.4 MB
1.1× more
stats-quickselectLomuto quickselect median without fully sorting observations
wago
10.2 MB
wazero
11.4 MB
1.1× less
stats-welfordQ8 online Welford mean and variance accumulator
wago
9.9 MB
wazero
9.1 MB
1.1× more
Search & indexing
search-aho-corasickAho–Corasick multi-pattern automaton with suffix-failure links
wago
10.6 MB
wazero
14.7 MB
1.4× less
search-bk-treeBK-tree fuzzy lookup using Hamming distance over 16-bit symbol vectors
wago
10 MB
wazero
12.6 MB
1.3× less
search-horspoolBoyer–Moore–Horspool substring search with bad-character shifts
wago
10.1 MB
wazero
12.2 MB
1.2× less
search-kmpKnuth–Morris–Pratt substring matching with prefix failure links
wago
10.1 MB
wazero
12.2 MB
1.2× less
search-rabin-karpRabin–Karp rolling polynomial hash with exact match verification
wago
10.1 MB
wazero
12.2 MB
1.2× less
Video processing
video-dctSeparable Q10 8×8 cosine transform (unscaled DCT-II coefficients)
wago
10.1 MB
wazero
12.4 MB
1.2× less
video-deblockingThresholded vertical block-boundary deblocking with clamped pair updates
wago
10 MB
wazero
11.9 MB
1.2× less
video-motion8×8 block motion estimation over a 5×5 displacement search
wago
10.2 MB
wazero
11.7 MB
1.1× less
video-optical-flowQ6 Lucas–Kanade optical flow from 3×3 patch normal equations
wago
10.1 MB
wazero
12.2 MB
1.2× less
video-temporal-denoiseTen-frame exponential temporal denoising with integer 3:1 history weighting
wago
10.9 MB
wazero
12.7 MB
1.2× less
video-yuv420Integer BT.601 RGB-to-YUV420 conversion with 2×2 chroma averaging
wago
9.9 MB
wazero
11.7 MB
1.2× less
Computer vision
vision-componentsFour-connected binary component labeling by union-find with canonical minimum labels
wago
10.1 MB
wazero
11.6 MB
1.1× less
vision-dilationBinary 3×3 morphological dilation with cropped image borders
wago
10.3 MB
wazero
11.8 MB
1.1× less
vision-distance-transformTwo-pass city-block distance transform of a binary image
wago
10.4 MB
wazero
12.4 MB
1.2× less
vision-fast-cornersFAST-9 corner detection on the 16-sample radius-three ring
wago
10.1 MB
wazero
11.9 MB
1.2× less
vision-otsuOtsu histogram-based global threshold selection for grayscale segmentation
wago
10 MB
wazero
10.2 MB
parity
vision-sobelSobel horizontal/vertical edges and gradient magnitude
wago
10.2 MB
wazero
11.4 MB
1.1× less
Audio
audio-adpcmCustom adaptive 4-bit differential PCM: saturating predictor and variable quantizer step
10.7× faster
audio-autocorrelationPitch-analysis autocorrelation over 32 positive sample lags
4.1× faster
audio-biquadQ8 second-order low-pass IIR biquad with reset filter history
6.5× faster
audio-fir32-tap fixed-point FIR audio filtering
11.5× faster
drwavAudio corpus
2.7× faster
kissfftAudio corpus
4.5× faster
3D & rendering
bezier-tessellationCubic 3D Bezier tessellation using Bernstein evaluation at 17 samples
11.3× faster
frustum-cullingOrthographic sphere/frustum culling against six clipping planes
7.3× faster
mesh-skinningFixed-point skeletal skinning with four bone influences and 3×4 matrices
13× faster
ray-boxFixed-point slab ray/AABB intersection with positive ray directions
5.6× faster
raytracerecursive Whitted ray tracer: n x n frame, 4 spheres + checker plane, depth-4 mirrors (heavy f64 vector math)
22.4× faster
triangle-rasterCPU triangle coverage rasterization into a 128×128 depth buffer
8.8× faster
Bioinformatics
bio-global-alignmentNeedleman-Wunsch global DNA alignment score with linear gap penalty
6.2× faster
bio-kmer-countComplete histogram of overlapping DNA 5-mers
12.1× faster
bio-lcsLongest common subsequence length of two deterministic DNA strings
3.4× faster
bio-local-alignmentSmith-Waterman best local DNA alignment score with linear gap penalty
5.8× faster
bio-reverse-complementReverse complement of a deterministic DNA sequence
4.5× faster
polybench-nussinovPolyBench/C nussinov (small dataset)
5.9× faster
Hashing & cryptography
blake3Hashing & cryptography corpus
26.5× faster
checksum-crc32IEEE CRC-32 byte-stream checksum using reflected bitwise polynomial reduction
6.9× faster
crypto-aes128AES-128 ECB encryption with key expansion and algebraic S-box setup included
4.3× faster
monocypherHashing & cryptography corpus
5.4× faster
sha256standards-correct SHA-256 of an n-KiB buffer (i32 rotate/shift/xor/add, 64-round)
15.7× faster
xxhashHashing & cryptography corpus
19.7× faster
Databases & analytics
columnar-queryColumnar range filter with grouped SUM and COUNT over 256 keys
8.5× faster
db-bitmap-indexBitmap filter intersection and set-bit counting
12.3× faster
db-btreeBulk-built eight-way B-tree index and point lookups
12.9× faster
db-hash-joinChained hash equijoin emitting matching left/right row pairs
13.2× faster
db-sort-merge-joinSort-merge equijoin with grouped result multiplicities
10.1× faster
Compilers & development tools
compiler-constant-foldConstant propagation and folding over a mixed literal/variable expression DAG
7.9× faster
compiler-dead-codeBackward reachability elimination of dead nodes in an acyclic IR dependency graph
9.4× faster
compiler-register-allocationLinear-scan register allocation with eight registers and furthest-end spilling
11.2× faster
Compression
compression-lzwLempel–Ziv–Welch dictionary compression with 12-bit code-space limit
14.7× faster
compression-rleByte run-length encoding of mixed repeated and noisy input spans
8.8× faster
lz4Compression corpus
20.3× faster
zlibCompression corpus
12.4× faster
zstdCompression corpus
15.8× faster
Integer & graph algorithms
coremarkInteger & graph algorithms corpus
3.2× faster
fannkuchBenchmarks-Game fannkuch-redux: permute + pancake-flip (branch/array churn, no f64)
9.9× faster
graph-dijkstraDense-graph Dijkstra single-source shortest paths
11.8× faster
graph-kruskalKruskal minimum-spanning-tree weight using sorted edges and disjoint sets
8.7× faster
polybench-floyd-warshallPolyBench/C floyd-warshall (small dataset)
4.5× faster
Calls & memory
dispatchcall_indirect through a function table
4× faster
fib_recrecursion; call/return heavy
3.8× faster
linked_listsingly-linked list build + dependent-load pointer chase
11.3× faster
many_funcs300 small functions; per-function compile cost
3.8× faster
memorylinear-memory load/store loop
5.4× faster
memory_treerecursive call tree with load/store churn at every node
13× faster
tinyminimal module; fixed pipeline overhead
2.2× faster
Documents & layout
document-bidi-reorderBidirectional run reordering from supplied embedding levels (L2 stage only)
9.1× faster
document-edit-distanceTwo-row Levenshtein document-token edit distance
12× faster
document-kerningGlyph positioning with pair-kerning adjustments and proportional advances
10.1× faster
document-layoutProportional glyph advances and greedy word line breaking into 640-unit lines
7.9× faster
document-optimal-wrapDynamic-programming minimum-raggedness line breaking with quadratic slack cost
11.4× faster
document-page-packingFirst-fit decreasing page packing of document block heights
11.7× faster
Text & parsing
fastfloatText & parsing corpus
2.8× faster
pcre2Text & parsing corpus
2.9× faster
text-substitutionLiteral non-overlapping error and warning substitutions over a deterministic record stream
8× faster
utf-as (simd)AssemblyScript utf-as SIMD: UTF-8 length and validation of mixed text
wago
28.5µs
wazero
19.1µs
1.5× slower
utf8procText & parsing corpus
1.9× slower
Files & directories
files-content-chunkingGear-hash content-defined chunk boundaries with 32-byte minimum and 1024-byte maximum
13.9× faster
files-glob-matchWildcard filename matching with star backtracking and single-character wildcards
8× faster
files-merkle-treeBinary Merkle content tree with a non-cryptographic FNV fixture hash
10.1× faster
files-path-normalizeLexical path normalization by dot-segment stack reduction
4.4× faster
files-path-trieDirectory-component trie construction with deduplicated prefixes
14× faster
Games & physics
game-cellular-automataEight generations of Conway Life on a toroidal game grid
17.7× faster
game-collision-impulseOne-dimensional elastic collision impulse resolution for contacting bodies
12.7× faster
game-flockingLocal-neighbor boid cohesion and separation steering
8.9× faster
game-spatial-hashUniform-grid broadphase pair enumeration for collision candidates
9.3× faster
grid-pathfindingFour-neighbor breadth-first shortest paths over a seeded obstacle grid
10.9× faster
particle-physics64 simulation steps of particle integration, gravity and reflecting wall collisions
11.4× faster
Geospatial
geo-convex-hullAndrew monotone-chain convex hull of two-dimensional map points
17.9× faster
geo-geohashBinary geohash encoding by interleaving quantized longitude/latitude bits
4.4× faster
geo-point-in-polygonEven-odd point-in-polygon classification against an eight-vertex map polygon
5.1× faster
geo-polygon-clippingInteger Sutherland–Hodgman clipping of map polygons to a tile rectangle
9.4× faster
geo-polyline-simplifyRamer–Douglas–Peucker polyline simplification with squared perpendicular-distance tolerance
8.3× faster
map-point-segmentMap geometry classification by exact squared distance to a finite line segment
5.7× faster
Graphics & images
graphics-bresenhamBresenham integer line rasterization into a 128×128 coverage image
11.9× faster
graphics-png-paethPNG Paeth scanline residual filtering with four-byte pixels
9× faster
graphics-reed-solomonGF(256) polynomial multiplication used in Reed–Solomon image-code parity construction
4.4× faster
nanosvgGraphics & images corpus
3.4× faster
polybench-derichePolyBench/C deriche (small dataset)
3.5× faster
qoiGraphics & images corpus
22.7× faster
Hardware design
hardware-boolean-anfBoolean Möbius transform into algebraic normal form for eight-input logic functions
6.9× faster
hardware-prime-implicantsFour-variable Boolean truth-table prime implicant enumeration
8.4× faster
Image editing
image-blur3×3 Gaussian blur with clamped image borders
7.9× faster
image-compositeAlpha compositing and clamped color adjustment
6.2× faster
image-error-diffusionFloyd–Steinberg monochrome error-diffusion dithering
9.1× faster
image-median3×3 median filtering of interior grayscale pixels
12.3× faster
image-resizeBilinear image downsampling at pixel centers
13× faster
image-ycocgReversible integer RGB-to-YCoCg lifting color transform
12.5× faster
JSON & serialization
json-as (simd)AssemblyScript json-as SIMD: serialize/deserialize canonical payload
1.3× slower
serialization-msgpackMessagePack shortest-form unsigned integer encoding
13× faster
serialization-protobufPacked Protobuf unsigned-varint stream decoding, with fixture encoding included
9.4× faster
yyjsonJSON & serialization corpus
3.8× faster
Language runtimes
language-brainfuckBrainfuck tape interpreter with bracket scanning and loop control
13.2× faster
language-forthTiny Forth-style text interpreter: decimal literals, arithmetic and output words
11.5× faster
language-register-vmEight-register bytecode interpreter with arithmetic, table loads and conditional skips
10.3× faster
Numerical arithmetic
libtommathNumerical arithmetic corpus
2.3× faster
nbodyBenchmarks-Game N-body: leapfrog solar-system integration (f64 mul/add/div/sqrt)
26.9× faster
numeric-cordicQ16 CORDIC sine/cosine rotation over angles in the convergence interval
4.3× faster
numeric-euclidean-gcdEuclidean greatest-common-divisor reduction of integer pairs
4.3× faster
numeric-integer-sqrtInteger Newton square-root iteration with floor-result correction
4.2× faster
numeric-simpsonComposite Simpson quadrature of a quadratic polynomial over integer sample points
4.8× faster
Machine learning
ml-convolutionValid 3×3 int8 convolution followed by ReLU activation
9.4× faster
ml-decision-treeDepth-eight decision-tree classifier over eight input features
10.3× faster
ml-inferenceInt8 dense inference: 64 input channels, 32 outputs, ReLU activation
14.5× faster
ml-kmeansFive Lloyd iterations of eight-cluster two-dimensional k-means
7.5× faster
ml-knnFive-nearest-neighbor classification against 32 eight-dimensional training vectors
5.3× faster
ml-max-pooling2×2 maximum-pooling reduction of activation planes
6.3× faster
Stencils
polybench-adiPolyBench/C adi (small dataset)
5.5× faster
polybench-fdtd-2dPolyBench/C fdtd-2d (small dataset)
4.4× faster
polybench-jacobi-2dPolyBench/C jacobi-2d (small dataset)
4.8× faster
polybench-seidel-2dPolyBench/C seidel-2d (small dataset)
4.3× faster
stencil-lattice-boltzmannD2Q9 lattice-Boltzmann fluid collision/streaming with BGK relaxation
6.6× faster
stencil-wave-equationSecond-order two-dimensional wave-equation time stepping on a periodic grid
29.7× faster
Linear algebra
polybench-choleskyPolyBench/C cholesky (small dataset)
5.3× faster
polybench-durbinPolyBench/C durbin (small dataset)
4.5× faster
polybench-gemmPolyBench/C gemm (small dataset)
5.2× faster
polybench-gramschmidtPolyBench/C gramschmidt (small dataset)
7.8× faster
polybench-luPolyBench/C lu (small dataset)
5.7× faster
polybench-trisolvPolyBench/C trisolv (small dataset)
4.9× faster
Statistics
polybench-correlationPolyBench/C correlation (small dataset)
4.6× faster
stats-bootstrapDeterministic bootstrap resampling of the mean, 64 resample replicates
10.4× faster
stats-giniQ16 empirical categorical Gini impurity from class-frequency counts
8.7× faster
stats-linear-regressionQ8 ordinary least-squares slope and intercept from sufficient statistics
7.7× faster
stats-quickselectLomuto quickselect median without fully sorting observations
5.7× faster
stats-welfordQ8 online Welford mean and variance accumulator
6× faster
Search & indexing
search-aho-corasickAho–Corasick multi-pattern automaton with suffix-failure links
19.3× faster
search-bk-treeBK-tree fuzzy lookup using Hamming distance over 16-bit symbol vectors
7.7× faster
search-horspoolBoyer–Moore–Horspool substring search with bad-character shifts
12.1× faster
search-kmpKnuth–Morris–Pratt substring matching with prefix failure links
3.8× faster
search-rabin-karpRabin–Karp rolling polynomial hash with exact match verification
4.2× faster
Video processing
video-dctSeparable Q10 8×8 cosine transform (unscaled DCT-II coefficients)
16.1× faster
video-deblockingThresholded vertical block-boundary deblocking with clamped pair updates
11.3× faster
video-motion8×8 block motion estimation over a 5×5 displacement search
11.1× faster
video-optical-flowQ6 Lucas–Kanade optical flow from 3×3 patch normal equations
11.6× faster
video-temporal-denoiseTen-frame exponential temporal denoising with integer 3:1 history weighting
9.3× faster
video-yuv420Integer BT.601 RGB-to-YUV420 conversion with 2×2 chroma averaging
9.1× faster
Computer vision
vision-componentsFour-connected binary component labeling by union-find with canonical minimum labels
7.3× faster
vision-dilationBinary 3×3 morphological dilation with cropped image borders
16.1× faster
vision-distance-transformTwo-pass city-block distance transform of a binary image
10.5× faster
vision-fast-cornersFAST-9 corner detection on the 16-sample radius-three ring
6.3× faster
vision-otsuOtsu histogram-based global threshold selection for grayscale segmentation
6.8× faster
vision-sobelSobel horizontal/vertical edges and gradient magnitude
7.4× faster
Audio
audio-adpcmCustom adaptive 4-bit differential PCM: saturating predictor and variable quantizer step
1.2× less
audio-autocorrelationPitch-analysis autocorrelation over 32 positive sample lags
1.6× less
audio-biquadQ8 second-order low-pass IIR biquad with reset filter history
1.3× less
audio-fir32-tap fixed-point FIR audio filtering
1.6× less
drwavAudio corpus
wago
85.6 KB
wazero
137 KB
1.6× less
kissfftAudio corpus
wago
39.1 KB
wazero
58.9 KB
1.5× less
3D & rendering
bezier-tessellationCubic 3D Bezier tessellation using Bernstein evaluation at 17 samples
1.1× more
frustum-cullingOrthographic sphere/frustum culling against six clipping planes
1.2× more
mesh-skinningFixed-point skeletal skinning with four bone influences and 3×4 matrices
1.6× less
ray-boxFixed-point slab ray/AABB intersection with positive ray directions
1.2× less
raytracerecursive Whitted ray tracer: n x n frame, 4 spheres + checker plane, depth-4 mirrors (heavy f64 vector math)
wago
11.9 KB
wazero
14 KB
1.2× less
triangle-rasterCPU triangle coverage rasterization into a 128×128 depth buffer
1.6× less
Bioinformatics
bio-global-alignmentNeedleman-Wunsch global DNA alignment score with linear gap penalty
wago
1.3 KB
wazero
2.1 KB
1.7× less
bio-kmer-countComplete histogram of overlapping DNA 5-mers
1.7× less
bio-lcsLongest common subsequence length of two deterministic DNA strings
wago
1.2 KB
wazero
2.1 KB
1.8× less
bio-local-alignmentSmith-Waterman best local DNA alignment score with linear gap penalty
wago
1.2 KB
wazero
2.1 KB
1.7× less
bio-reverse-complementReverse complement of a deterministic DNA sequence
1.2× more
polybench-nussinovPolyBench/C nussinov (small dataset)
1.5× less
Hashing & cryptography
blake3Hashing & cryptography corpus
wago
29.4 KB
wazero
37.8 KB
1.3× less
checksum-crc32IEEE CRC-32 byte-stream checksum using reflected bitwise polynomial reduction
1.1× more
crypto-aes128AES-128 ECB encryption with key expansion and algebraic S-box setup included
1.5× less
monocypherHashing & cryptography corpus
wago
34.7 KB
wazero
57.8 KB
1.7× less
sha256standards-correct SHA-256 of an n-KiB buffer (i32 rotate/shift/xor/add, 64-round)
1.6× less
xxhashHashing & cryptography corpus
wago
3.2 KB
wazero
4.9 KB
1.5× less
Databases & analytics
columnar-queryColumnar range filter with grouped SUM and COUNT over 256 keys
1.7× less
db-bitmap-indexBitmap filter intersection and set-bit counting
1.7× less
db-btreeBulk-built eight-way B-tree index and point lookups
wago
2.3 KB
wazero
4.2 KB
1.8× less
db-hash-joinChained hash equijoin emitting matching left/right row pairs
1.8× less
db-sort-merge-joinSort-merge equijoin with grouped result multiplicities
wago
1.2 KB
wazero
1.9 KB
1.7× less
Compilers & development tools
compiler-constant-foldConstant propagation and folding over a mixed literal/variable expression DAG
wago
1.1 KB
wazero
1.7 KB
1.6× less
compiler-dead-codeBackward reachability elimination of dead nodes in an acyclic IR dependency graph
wago
1.2 KB
wazero
1.9 KB
1.6× less
compiler-register-allocationLinear-scan register allocation with eight registers and furthest-end spilling
wago
2.5 KB
wazero
5.8 KB
2.3× less
Compression
compression-lzwLempel–Ziv–Welch dictionary compression with 12-bit code-space limit
1.4× less
compression-rleByte run-length encoding of mixed repeated and noisy input spans
1.2× more
lz4Compression corpus
1.8× less
zlibCompression corpus
wago
38.4 KB
wazero
72.4 KB
1.9× less
zstdCompression corpus
wago
165 KB
wazero
280 KB
1.7× less
Integer & graph algorithms
coremarkInteger & graph algorithms corpus
wago
21.8 KB
wazero
32 KB
1.5× less
fannkuchBenchmarks-Game fannkuch-redux: permute + pancake-flip (branch/array churn, no f64)
wago
5.1 KB
wazero
7.3 KB
1.4× less
graph-dijkstraDense-graph Dijkstra single-source shortest paths
wago
1.6 KB
wazero
3.1 KB
1.9× less
graph-kruskalKruskal minimum-spanning-tree weight using sorted edges and disjoint sets
2× less
polybench-floyd-warshallPolyBench/C floyd-warshall (small dataset)
wago
4.7 KB
wazero
5.1 KB
1.1× less
Calls & memory
dispatchcall_indirect through a function table
1.1× less
fib_recrecursion; call/return heavy
2.1× more
linked_listsingly-linked list build + dependent-load pointer chase
1.6× less
many_funcs300 small functions; per-function compile cost
wago
8.2 KB
wazero
18.9 KB
2.3× less
memorylinear-memory load/store loop
1.6× more
memory_treerecursive call tree with load/store churn at every node
1.1× more
tinyminimal module; fixed pipeline overhead
1.3× less
Documents & layout
document-bidi-reorderBidirectional run reordering from supplied embedding levels (L2 stage only)
1.7× less
document-edit-distanceTwo-row Levenshtein document-token edit distance
wago
1.3 KB
wazero
2.4 KB
1.9× less
document-kerningGlyph positioning with pair-kerning adjustments and proportional advances
1.3× less
document-layoutProportional glyph advances and greedy word line breaking into 640-unit lines
1.2× less
document-optimal-wrapDynamic-programming minimum-raggedness line breaking with quadratic slack cost
1.7× less
document-page-packingFirst-fit decreasing page packing of document block heights
wago
1.4 KB
wazero
2.4 KB
1.8× less
Text & parsing
fastfloatText & parsing corpus
wago
48.6 KB
wazero
70.9 KB
1.5× less
pcre2Text & parsing corpus
wago
483 KB
wazero
1.1 MB
2.3× less
text-substitutionLiteral non-overlapping error and warning substitutions over a deterministic record stream
1.9× less
utf-as (simd)AssemblyScript utf-as SIMD: UTF-8 length and validation of mixed text
wago
21.7 KB
wazero
33.1 KB
1.5× less
utf8procText & parsing corpus
wago
30.6 KB
wazero
51.1 KB
1.7× less
Files & directories
files-content-chunkingGear-hash content-defined chunk boundaries with 32-byte minimum and 1024-byte maximum
parity
files-glob-matchWildcard filename matching with star backtracking and single-character wildcards
1.3× less
files-merkle-treeBinary Merkle content tree with a non-cryptographic FNV fixture hash
wago
1.5 KB
wazero
2.2 KB
1.5× less
files-path-normalizeLexical path normalization by dot-segment stack reduction
1.7× less
files-path-trieDirectory-component trie construction with deduplicated prefixes
wago
1.6 KB
wazero
2.6 KB
1.6× less
Games & physics
game-cellular-automataEight generations of Conway Life on a toroidal game grid
wago
2.1 KB
wazero
3.9 KB
1.9× less
game-collision-impulseOne-dimensional elastic collision impulse resolution for contacting bodies
1.1× less
game-flockingLocal-neighbor boid cohesion and separation steering
1.5× less
game-spatial-hashUniform-grid broadphase pair enumeration for collision candidates
1.7× less
grid-pathfindingFour-neighbor breadth-first shortest paths over a seeded obstacle grid
1.8× less
particle-physics64 simulation steps of particle integration, gravity and reflecting wall collisions
1.1× more
Geospatial
geo-convex-hullAndrew monotone-chain convex hull of two-dimensional map points
wago
1.8 KB
wazero
3.6 KB
2× less
geo-geohashBinary geohash encoding by interleaving quantized longitude/latitude bits
1.2× less
geo-point-in-polygonEven-odd point-in-polygon classification against an eight-vertex map polygon
1.1× less
geo-polygon-clippingInteger Sutherland–Hodgman clipping of map polygons to a tile rectangle
1.6× less
geo-polyline-simplifyRamer–Douglas–Peucker polyline simplification with squared perpendicular-distance tolerance
wago
1.6 KB
wazero
2.6 KB
1.6× less
map-point-segmentMap geometry classification by exact squared distance to a finite line segment
1.1× more
Graphics & images
graphics-bresenhamBresenham integer line rasterization into a 128×128 coverage image
wago
1.1 KB
wazero
1.6 KB
1.4× less
graphics-png-paethPNG Paeth scanline residual filtering with four-byte pixels
1.1× more
graphics-reed-solomonGF(256) polynomial multiplication used in Reed–Solomon image-code parity construction
wago
4.7 KB
wazero
4.3 KB
1.1× more
nanosvgGraphics & images corpus
wago
161 KB
wazero
227 KB
1.4× less
polybench-derichePolyBench/C deriche (small dataset)
wago
6.5 KB
wazero
9.6 KB
1.5× less
qoiGraphics & images corpus
wago
6.1 KB
wazero
9.9 KB
1.6× less
Hardware design
hardware-boolean-anfBoolean Möbius transform into algebraic normal form for eight-input logic functions
wago
2.1 KB
wazero
4.3 KB
2.1× less
hardware-prime-implicantsFour-variable Boolean truth-table prime implicant enumeration
wago
1.4 KB
wazero
1.9 KB
1.4× less
Image editing
image-blur3×3 Gaussian blur with clamped image borders
wago
1.6 KB
wazero
2.6 KB
1.7× less
image-compositeAlpha compositing and clamped color adjustment
1.4× less
image-error-diffusionFloyd–Steinberg monochrome error-diffusion dithering
wago
1.4 KB
wazero
3.1 KB
2.3× less
image-median3×3 median filtering of interior grayscale pixels
wago
3.4 KB
wazero
6.2 KB
1.8× less
image-resizeBilinear image downsampling at pixel centers
wago
1.1 KB
wazero
1.8 KB
1.7× less
image-ycocgReversible integer RGB-to-YCoCg lifting color transform
1.3× more
JSON & serialization
json-as (simd)AssemblyScript json-as SIMD: serialize/deserialize canonical payload
wago
88.5 KB
wazero
102 KB
1.2× less
serialization-msgpackMessagePack shortest-form unsigned integer encoding
1.3× more
serialization-protobufPacked Protobuf unsigned-varint stream decoding, with fixture encoding included
1.3× less
yyjsonJSON & serialization corpus
wago
288 KB
wazero
519 KB
1.8× less
Language runtimes
language-brainfuckBrainfuck tape interpreter with bracket scanning and loop control
1.5× less
language-forthTiny Forth-style text interpreter: decimal literals, arithmetic and output words
wago
1.3 KB
wazero
2.4 KB
1.8× less
language-register-vmEight-register bytecode interpreter with arithmetic, table loads and conditional skips
1.7× less
Numerical arithmetic
libtommathNumerical arithmetic corpus
wago
113 KB
wazero
147 KB
1.3× less
nbodyBenchmarks-Game N-body: leapfrog solar-system integration (f64 mul/add/div/sqrt)
wago
3.4 KB
wazero
5.7 KB
1.7× less
numeric-cordicQ16 CORDIC sine/cosine rotation over angles in the convergence interval
1.2× more
numeric-euclidean-gcdEuclidean greatest-common-divisor reduction of integer pairs
parity
numeric-integer-sqrtInteger Newton square-root iteration with floor-result correction
parity
numeric-simpsonComposite Simpson quadrature of a quadratic polynomial over integer sample points
1.2× less
Machine learning
ml-convolutionValid 3×3 int8 convolution followed by ReLU activation
1.8× less
ml-decision-treeDepth-eight decision-tree classifier over eight input features
1.1× more
ml-inferenceInt8 dense inference: 64 input channels, 32 outputs, ReLU activation
1.7× less
ml-kmeansFive Lloyd iterations of eight-cluster two-dimensional k-means
1.9× less
ml-knnFive-nearest-neighbor classification against 32 eight-dimensional training vectors
wago
2.2 KB
wazero
2.8 KB
1.3× less
ml-max-pooling2×2 maximum-pooling reduction of activation planes
1.5× less
Stencils
polybench-adiPolyBench/C adi (small dataset)
wago
9.3 KB
wazero
11.6 KB
1.2× less
polybench-fdtd-2dPolyBench/C fdtd-2d (small dataset)
wago
9.5 KB
wazero
10.7 KB
1.1× less
polybench-jacobi-2dPolyBench/C jacobi-2d (small dataset)
wago
6.4 KB
wazero
6.1 KB
1.1× more
polybench-seidel-2dPolyBench/C seidel-2d (small dataset)
1.5× less
stencil-lattice-boltzmannD2Q9 lattice-Boltzmann fluid collision/streaming with BGK relaxation
wago
5.6 KB
wazero
7.3 KB
1.3× less
stencil-wave-equationSecond-order two-dimensional wave-equation time stepping on a periodic grid
wago
2.2 KB
wazero
4.2 KB
1.9× less
Linear algebra
polybench-choleskyPolyBench/C cholesky (small dataset)
wago
6.4 KB
wazero
9.9 KB
1.5× less
polybench-durbinPolyBench/C durbin (small dataset)
wago
4.6 KB
wazero
6.8 KB
1.5× less
polybench-gemmPolyBench/C gemm (small dataset)
wago
5.5 KB
wazero
7.4 KB
1.3× less
polybench-gramschmidtPolyBench/C gramschmidt (small dataset)
wago
5.6 KB
wazero
8.4 KB
1.5× less
polybench-luPolyBench/C lu (small dataset)
wago
6.3 KB
wazero
9.9 KB
1.6× less
polybench-trisolvPolyBench/C trisolv (small dataset)
1.5× less
Statistics
polybench-correlationPolyBench/C correlation (small dataset)
1.3× less
stats-bootstrapDeterministic bootstrap resampling of the mean, 64 resample replicates
1.5× less
stats-giniQ16 empirical categorical Gini impurity from class-frequency counts
wago
2.3 KB
wazero
2.9 KB
1.2× less
stats-linear-regressionQ8 ordinary least-squares slope and intercept from sufficient statistics
1.9× more
stats-quickselectLomuto quickselect median without fully sorting observations
1.6× less
stats-welfordQ8 online Welford mean and variance accumulator
parity
Search & indexing
search-aho-corasickAho–Corasick multi-pattern automaton with suffix-failure links
wago
6.2 KB
wazero
9.7 KB
1.6× less
search-bk-treeBK-tree fuzzy lookup using Hamming distance over 16-bit symbol vectors
wago
1.4 KB
wazero
2.3 KB
1.6× less
search-horspoolBoyer–Moore–Horspool substring search with bad-character shifts
wago
2.7 KB
wazero
4.5 KB
1.7× less
search-kmpKnuth–Morris–Pratt substring matching with prefix failure links
wago
1.8 KB
wazero
3.3 KB
1.8× less
search-rabin-karpRabin–Karp rolling polynomial hash with exact match verification
wago
3.9 KB
wazero
6.6 KB
1.7× less
Video processing
video-dctSeparable Q10 8×8 cosine transform (unscaled DCT-II coefficients)
1.3× less
video-deblockingThresholded vertical block-boundary deblocking with clamped pair updates
wago
1.2 KB
wazero
2.2 KB
1.9× less
video-motion8×8 block motion estimation over a 5×5 displacement search
1.8× less
video-optical-flowQ6 Lucas–Kanade optical flow from 3×3 patch normal equations
wago
3.4 KB
wazero
4.9 KB
1.4× less
video-temporal-denoiseTen-frame exponential temporal denoising with integer 3:1 history weighting
wago
4.1 KB
wazero
5.9 KB
1.4× less
video-yuv420Integer BT.601 RGB-to-YUV420 conversion with 2×2 chroma averaging
1.7× less
Computer vision
vision-componentsFour-connected binary component labeling by union-find with canonical minimum labels
wago
1.9 KB
wazero
3.5 KB
1.8× less
vision-dilationBinary 3×3 morphological dilation with cropped image borders
wago
1.3 KB
wazero
2.2 KB
1.7× less
vision-distance-transformTwo-pass city-block distance transform of a binary image
wago
2.4 KB
wazero
4.1 KB
1.7× less
vision-fast-cornersFAST-9 corner detection on the 16-sample radius-three ring
wago
1.9 KB
wazero
3.2 KB
1.7× less
vision-otsuOtsu histogram-based global threshold selection for grayscale segmentation
wago
1.7 KB
wazero
1.9 KB
1.1× less
vision-sobelSobel horizontal/vertical edges and gradient magnitude
1.7× less
Audio
audio-adpcmrun · Audio corpus
wago
47.1µs
wazero
55.7µs
1.2× faster
audio-autocorrelationrun · Audio corpus
1.3× faster
audio-biquadrun · Audio corpus
wago
17.4µs
wazero
28.2µs
1.6× faster
audio-firrun · Audio corpus
wago
48.1µs
wazero
88.4µs
1.8× faster
drwavpcm-decode-seek · Audio corpus
1.3× faster
kissfftcomplex-roundtrip · Audio corpus
2× faster
3D & rendering
bezier-tessellationrun · 3D & rendering corpus
wago
24.5µs
wazero
24.6µs
parity
frustum-cullingrun · 3D & rendering corpus
wago
16.6µs
wazero
16.9µs
parity
mesh-skinningrun · 3D & rendering corpus
wago
43.3µs
wazero
53.4µs
1.2× faster
ray-boxrun · 3D & rendering corpus
wago
38.6µs
wazero
37.4µs
parity
raytracerender · 3D & rendering corpus
parity
triangle-rasterrun · 3D & rendering corpus
2.3× faster
Bioinformatics
bio-global-alignmentrun · Bioinformatics corpus
1.6× faster
bio-kmer-countrun · Bioinformatics corpus
wago
39.2µs
wazero
49.3µs
1.3× faster
bio-lcsrun · Bioinformatics corpus
1.6× faster
bio-local-alignmentrun · Bioinformatics corpus
1.9× faster
bio-reverse-complementrun · Bioinformatics corpus
wago
34.5µs
wazero
34.8µs
parity
polybench-nussinovpolybench_run · Bioinformatics corpus
1.8× faster
Hashing & cryptography
blake3hash · Hashing & cryptography corpus
parity
checksum-crc32run · Hashing & cryptography corpus
1.1× faster
crypto-aes128run · Hashing & cryptography corpus
wago
67.1µs
wazero
80.6µs
1.2× faster
monocypheraead-blake2b · Hashing & cryptography corpus
wago
33.9µs
wazero
42.1µs
1.2× faster
sha256hashN · Hashing & cryptography corpus
wago
56.7µs
wazero
56.8µs
parity
xxhashxxh64-boundaries · Hashing & cryptography corpus
2× faster
Databases & analytics
columnar-queryrun · Databases & analytics corpus
1.3× faster
db-bitmap-indexrun · Databases & analytics corpus
wago
14.8µs
wazero
20.5µs
1.4× faster
db-btreerun · Databases & analytics corpus
wago
12.9µs
wazero
25.1µs
2× faster
db-hash-joinrun · Databases & analytics corpus
1.6× faster
db-sort-merge-joinrun · Databases & analytics corpus
2.4× faster
Compilers & development tools
compiler-constant-foldrun · Compilers & development tools corpus
1.4× faster
compiler-dead-coderun · Compilers & development tools corpus
2.4× faster
compiler-register-allocationrun · Compilers & development tools corpus
wago
24.3µs
wazero
32.5µs
1.3× faster
Compression
compression-lzwrun · Compression corpus
3.7× faster
compression-rlerun · Compression corpus
wago
19.7µs
wazero
23.2µs
1.2× faster
lz4compress · Compression corpus
1.5× faster
zlibinflate · Compression corpus
wago
15.9µs
wazero
16.1µs
parity
zstddecompress · Compression corpus
2.1× slower
Integer & graph algorithms
coremark2k-performance · Integer & graph algorithms corpus
1.8× faster
fannkuchrun · Integer & graph algorithms corpus
wago
1.13ms
wazero
1.49ms
1.3× faster
graph-dijkstrarun · Integer & graph algorithms corpus
1.9× faster
graph-kruskalrun · Integer & graph algorithms corpus
2× faster
polybench-floyd-warshallpolybench_run · Integer & graph algorithms corpus
wago
1.08ms
wazero
9.21ms
8.5× faster
Calls & memory
dispatchapply · Calls & memory corpus
1.6× faster
fib_recfib · Calls & memory corpus
3.2× faster
linked_listsum · Calls & memory corpus
1.3× faster
many_funcsrun · Calls & memory corpus
1.6× faster
memorysum · Calls & memory corpus
2.2× faster
memory_treerun · Calls & memory corpus
2.6× faster
tinyadd · Calls & memory corpus
1.6× faster
Documents & layout
document-bidi-reorderrun · Documents & layout corpus
wago
17.2µs
wazero
30.8µs
1.8× faster
document-edit-distancerun · Documents & layout corpus
1.5× faster
document-kerningrun · Documents & layout corpus
parity
document-layoutrun · Documents & layout corpus
1.6× faster
document-optimal-wraprun · Documents & layout corpus
1.3× faster
document-page-packingrun · Documents & layout corpus
2.4× faster
Text & parsing
fastfloatdecimal-parse · Text & parsing corpus
1.1× faster
pcre2compile-match · Text & parsing corpus
wago
10.6µs
wazero
20.4µs
1.9× faster
text-substitutionrun · Text & parsing corpus
wago
71.2µs
wazero
83.6µs
1.2× faster
utf-as (simd)validateN · Text & parsing corpus
1.7× faster
utf8procnormalize-casefold · Text & parsing corpus
1.5× slower
Files & directories
files-content-chunkingrun · Files & directories corpus
wago
42.5µs
wazero
65.7µs
1.5× faster
files-glob-matchrun · Files & directories corpus
1.7× faster
files-merkle-treerun · Files & directories corpus
1.1× faster
files-path-normalizerun · Files & directories corpus
wago
23.2µs
wazero
35.2µs
1.5× faster
files-path-trierun · Files & directories corpus
1.5× faster
Games & physics
game-cellular-automatarun · Games & physics corpus
1.5× faster
game-collision-impulserun · Games & physics corpus
wago
39.5µs
wazero
39.8µs
parity
game-flockingrun · Games & physics corpus
2.1× faster
game-spatial-hashrun · Games & physics corpus
wago
28.5µs
wazero
29.5µs
parity
grid-pathfindingrun · Games & physics corpus
wago
55.7µs
wazero
75.1µs
1.3× faster
particle-physicsrun · Games & physics corpus
1.1× faster
Geospatial
geo-convex-hullrun · Geospatial corpus
wago
18.4µs
wazero
39.2µs
2.1× faster
geo-geohashrun · Geospatial corpus
wago
67.5µs
wazero
63.6µs
1.1× slower
geo-point-in-polygonrun · Geospatial corpus
wago
23.1µs
wazero
28.9µs
1.3× faster
geo-polygon-clippingrun · Geospatial corpus
wago
51.8µs
wazero
66.6µs
1.3× faster
geo-polyline-simplifyrun · Geospatial corpus
1.2× faster
map-point-segmentrun · Geospatial corpus
wago
39.9µs
wazero
39.3µs
parity
Graphics & images
graphics-bresenhamrun · Graphics & images corpus
1.1× faster
graphics-png-paethrun · Graphics & images corpus
wago
70.9µs
wazero
79.1µs
1.1× faster
graphics-reed-solomonrun · Graphics & images corpus
parity
nanosvgparse-structure · Graphics & images corpus
1.6× faster
polybench-derichepolybench_run · Graphics & images corpus
1.2× faster
qoiencode · Graphics & images corpus
1.1× slower
Hardware design
hardware-boolean-anfrun · Hardware design corpus
2× faster
hardware-prime-implicantsrun · Hardware design corpus
1.9× faster
Image editing
image-blurrun · Image editing corpus
1.4× faster
image-compositerun · Image editing corpus
1.4× faster
image-error-diffusionrun · Image editing corpus
1.6× faster
image-medianrun · Image editing corpus
1.5× faster
image-resizerun · Image editing corpus
wago
34.6µs
wazero
52.7µs
1.5× faster
image-ycocgrun · Image editing corpus
wago
47.1µs
wazero
45.9µs
parity
JSON & serialization
serialization-msgpackrun · JSON & serialization corpus
wago
10.6µs
wazero
10.5µs
parity
serialization-protobufrun · JSON & serialization corpus
wago
25.7µs
wazero
39.5µs
1.5× faster
yyjsonparse-edit-write · JSON & serialization corpus
2.5× faster
Language runtimes
language-brainfuckrun · Language runtimes corpus
1.6× faster
language-forthrun · Language runtimes corpus
wago
21.1µs
wazero
34.7µs
1.6× faster
language-register-vmrun · Language runtimes corpus
wago
25.3µs
wazero
35.2µs
1.4× faster
Numerical arithmetic
libtommathmodular-arithmetic · Numerical arithmetic corpus
wago
15.2µs
wazero
24.2µs
1.6× faster
nbodystep · Numerical arithmetic corpus
1.4× faster
numeric-cordicrun · Numerical arithmetic corpus
parity
numeric-euclidean-gcdrun · Numerical arithmetic corpus
1.8× faster
numeric-integer-sqrtrun · Numerical arithmetic corpus
1.3× faster
numeric-simpsonrun · Numerical arithmetic corpus
wago
731.5ns
wazero
822ns
1.1× faster
Machine learning
ml-convolutionrun · Machine learning corpus
wago
52.1µs
wazero
77.5µs
1.5× faster
ml-decision-treerun · Machine learning corpus
parity
ml-inferencerun · Machine learning corpus
wago
26.2µs
wazero
34.8µs
1.3× faster
ml-kmeansrun · Machine learning corpus
wago
46.5µs
wazero
54.9µs
1.2× faster
ml-knnrun · Machine learning corpus
wago
61.5µs
wazero
69.9µs
1.1× faster
ml-max-poolingrun · Machine learning corpus
wago
35.2µs
wazero
54.3µs
1.5× faster
Stencils
polybench-adipolybench_run · Stencils corpus
wago
1.81ms
wazero
1.90ms
parity
polybench-fdtd-2dpolybench_run · Stencils corpus
3.1× faster
polybench-jacobi-2dpolybench_run · Stencils corpus
7.1× faster
polybench-seidel-2dpolybench_run · Stencils corpus
wago
3.77ms
wazero
4.07ms
1.1× faster
stencil-lattice-boltzmannrun · Stencils corpus
1.6× faster
stencil-wave-equationrun · Stencils corpus
1.6× faster
Linear algebra
polybench-choleskypolybench_run · Linear algebra corpus
wago
1.16ms
wazero
2.49ms
2.1× faster
polybench-durbinpolybench_run · Linear algebra corpus
1.7× faster
polybench-gemmpolybench_run · Linear algebra corpus
4.2× faster
polybench-gramschmidtpolybench_run · Linear algebra corpus
1.8× faster
polybench-lupolybench_run · Linear algebra corpus
wago
1.37ms
wazero
2.99ms
2.2× faster
polybench-trisolvpolybench_run · Linear algebra corpus
wago
12.7µs
wazero
41.8µs
3.3× faster
Statistics
polybench-correlationpolybench_run · Statistics corpus
2× faster
stats-bootstraprun · Statistics corpus
wago
47.3µs
wazero
44.9µs
1.1× slower
stats-ginirun · Statistics corpus
1.3× faster
stats-linear-regressionrun · Statistics corpus
1.8× faster
stats-quickselectrun · Statistics corpus
wago
14.8µs
wazero
32.5µs
2.2× faster
stats-welfordrun · Statistics corpus
wago
19.1µs
wazero
19.4µs
parity
Search & indexing
search-aho-corasickrun · Search & indexing corpus
4.3× faster
search-bk-treerun · Search & indexing corpus
1.6× faster
search-horspoolrun · Search & indexing corpus
wago
56.5µs
wazero
70.3µs
1.2× faster
search-kmprun · Search & indexing corpus
1.3× faster
search-rabin-karprun · Search & indexing corpus
1.7× faster
Video processing
video-dctrun · Video processing corpus
1.2× faster
video-deblockingrun · Video processing corpus
wago
30.7µs
wazero
38.5µs
1.3× faster
video-motionrun · Video processing corpus
1.7× faster
video-optical-flowrun · Video processing corpus
wago
32.9µs
wazero
42.2µs
1.3× faster
video-temporal-denoiserun · Video processing corpus
1.1× faster
video-yuv420run · Video processing corpus
1.1× slower
Computer vision
vision-componentsrun · Computer vision corpus
wago
47.3µs
wazero
76.9µs
1.6× faster
vision-dilationrun · Computer vision corpus
1.9× faster
vision-distance-transformrun · Computer vision corpus
1.9× faster
vision-fast-cornersrun · Computer vision corpus
wago
1.65ms
wazero
2.05ms
1.2× faster
vision-otsurun · Computer vision corpus
wago
18.4µs
wazero
21.7µs
1.2× faster
vision-sobelrun · Computer vision corpus
1.4× faster