As was discussed in the SciPy integration thread, Python libraries use double-precision floating-point numbers by default.
So, in this release, I've extended the spatial distance functions in the underlying SimSIMD - Cos, L2sq, IP with support for double arguments with specialized implementations on AVX-512-capable x86 CPUs and SVE-capable Arm CPUs.
Benchmarking SimSIMD vs. SciPy on Intel Sapphire Rapids CPU
- Vector dimensions: 1536
- Vectors count: 1000
- Hardware capabilities: serial, x86_avx2, x86_avx512, x86_avx2fp16, x86_avx512fp16, x86_avx512vpopcntdq, x86_avx512vnni
- NumPy BLAS dependency: openblas64
- NumPy LAPACK dependency: dep140640983012528
Between 2 Vectors, Batch Size: 1
| Datatype |
Method |
Ops/s |
SimSIMD Ops/s |
SimSIMD Improvement |
f64 |
scipy.cosine |
63,612 |
572,605 |
9.00 x |
f64 |
scipy.sqeuclidean |
238,547 |
915,596 |
3.84 x |
f64 |
numpy.inner |
449,499 |
986,522 |
2.19 x |
Between 2 Vectors, Batch Size: 1,000
| Datatype |
Method |
Ops/s |
SimSIMD Ops/s |
SimSIMD Improvement |
f64 |
scipy.cosine |
68,962 |
1,457,172 |
21.13 x |
f64 |
scipy.sqeuclidean |
247,727 |
1,535,547 |
6.20 x |
f64 |
numpy.inner |
463,509 |
1,512,004 |
3.26 x |
Benchmarking SimSIMD vs. SciPy on AWS Graviton 3
- Vector dimensions: 1536
- Vectors count: 1000
- Hardware capabilities: serial, arm_neon, arm_sve
- NumPy BLAS dependency: openblas64
- NumPy LAPACK dependency: openblas64
Between 2 Vectors, Batch Size: 1
| Datatype |
Method |
Ops/s |
SimSIMD Ops/s |
SimSIMD Improvement |
f64 |
scipy.cosine |
40,729 |
725,382 |
17.81 x |
f64 |
scipy.sqeuclidean |
160,812 |
728,114 |
4.53 x |
f64 |
numpy.inner |
473,443 |
767,374 |
1.62 x |
f64 |
scipy.jensenshannon |
15,684 |
38,528 |
2.46 x |
f64 |
scipy.kl_div |
49,983 |
61,811 |
1.24 x |
Between 2 Vectors, Batch Size: 1,000
| Datatype |
Method |
Ops/s |
SimSIMD Ops/s |
SimSIMD Improvement |
f64 |
scipy.cosine |
41,130 |
1,460,850 |
35.52 x |
f64 |
scipy.sqeuclidean |
162,147 |
1,486,255 |
9.17 x |
f64 |
numpy.inner |
473,856 |
1,580,136 |
3.33 x |
Hashes
- docs.tar.gz :
def474428a4d67076e68dfd16b660a53bf51fad12af7e4c6ee77e1555b220b8f
- usearch-v2.8.14.tar.gz :
6ae186618120b6c710ff3ed1bf31e9a58610e7b837bccdeae79000247c2b24a3
- usearch-v2.8.14.zip :
a1da6b34bc23111926b16be9d36a7403988405a00c833417b27b8ccd9c70227f
- usearch_linux_amd_2.8.14.deb :
ae4995f9504a9ab90921e3091a5aa6af432de647f1a6c835ee5cb2622dd2f8a3
- usearch_linux_arm_2.8.14.deb :
dafcee294630b7c17adaed9aebb668d4cbfc5fe269f35f66b10bf458d66d899d
- usearch_macOS_arm64_2.8.14.zip :
65e0a8d0259400e35de692c9afb3406d6bd4db5dd2f46632676890c41ba1537c
- usearch_macOS_x86_64_2.8.14.zip :
1fa9a6e6983f5b6fe5dd2b82ce566262dcbd1d1e8671ec39b9d01ceddc3b80dd
- usearch_wasm_linux_arm64_2.8.14.tar.gz :
edb5846ab0b38b1095b12f2b8ca771748394911bf6efd74df6d5ba66f36328d9
- usearch_wasm_linux_x86_64_2.8.14.tar.gz :
84e4162db1dc83157f7a4032278c9cfc01a68ac59c4b1af1b5fc35e43905c515
- usearch_wasm_macos_arm64_2.8.14.zip :
8c966ef2f5e425cf82f472bd4b74f0e7cd7fa39b860dd9f25f488370afc035fe
- usearch_wasm_macos_x86_64_2.8.14.zip :
7e48cc5d0e34b586f36b86671c8410b2b2deca31962beb0ccf6230546ac42621
- usearch_wasm_windows_x64_2.8.14.tar.gz :
02b1382c68fe8ef52d55da639a38ad15c7449d579230ef5ae78465776fd689fa
- usearch_wasm_windows_x86_2.8.14.tar.gz :
cf922953ca61c8f9101fac33a2aa06d70fd09ef8b8c73d64a6623f73dcfe09cc
- usearch_windows_x64_2.8.14.tar :
8a74a122fafae229f65df8de5a6e264eb16a7e5eba691629d6439b0b54ea8b74
- usearch_windows_x86_2.8.14.tar :
96c6625fc2ca723cc44e864a75f5c72dc3ab9aeb5fe5f0f80ba32783121f6766
