NineToothed
NineToothed is a Triton-based domain-specific language (DSL). By introducing tensor-oriented meta-programming (TOM), it makes writing high-performance GPU kernels easier.
Installation
We can use pip to install ninetoothed.
pip install ninetoothed
After successfully running the above command, ninetoothed will be installed. However, to fully utilize its capabilities, you also need to install a deep learning framework supported by ninetoothed. For trial purposes, we recommend installing torch.
Usage
Thanks to tensor-oriented meta-programming, NineToothed can be written using the arrange-and-apply paradigm, which involves separately defining arrangement, application, and tensors, and then integrating them using ninetoothed.make to generate the kernel.
Matrix Multiplication
Here is the code we need for matrix multiplication:
import ninetoothed
import ninetoothed.language as ntl
from ninetoothed import Tensor, block_size
BLOCK_SIZE_M = block_size()
BLOCK_SIZE_N = block_size()
BLOCK_SIZE_K = block_size()
def arrangement(input, other, output):
output_arranged = output.tile((BLOCK_SIZE_M, BLOCK_SIZE_N))
input_arranged = input.tile((BLOCK_SIZE_M, BLOCK_SIZE_K))
input_arranged = input_arranged.tile((1, -1))
input_arranged = input_arranged.expand((-1, output_arranged.shape[1]))
input_arranged.dtype = input_arranged.dtype.squeeze(0)
other_arranged = other.tile((BLOCK_SIZE_K, BLOCK_SIZE_N))
other_arranged = other_arranged.tile((-1, 1))
other_arranged = other_arranged.expand((output_arranged.shape[0], -1))
other_arranged.dtype = other_arranged.dtype.squeeze(1)
return input_arranged, other_arranged, output_arranged
def application(input, other, output):
accumulator = ntl.zeros(output.shape, dtype=ntl.float32)
for k in range(input.shape[0]):
accumulator += ntl.dot(input[k], other[k])
output = accumulator
tensors = (Tensor(2), Tensor(2), Tensor(2))
kernel = ninetoothed.make(arrangement, application, tensors)
Useful Links
License
This project is distributed under the Apache-2.0 license. See the included LICENSE file for details.
1 maintainer
Metadata
1 maintainer
Release files for ninetoothed 0.26.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| ninetoothed-0.26.0.tar.gz | 70.5 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| ninetoothed-0.26.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 117.8 kB
Release files / ninetoothed-0.26.0.tar.gz
| Download URL | ninetoothed-0.26.0.tar.gz |
|---|---|
| Size | 70.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
bb01c9ed2fc8f541ecde187f9976a62bd0370d3cbad6bc91ba6390927aa28321
|
|
BLAKE2b-256 checksum How to use checksums |
32175b0fa96254e1859be76e4a720b1449f922e2c5fab06bfafb1282124aa55b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / ninetoothed-0.26.0-py3-none-any.whl
| Download URL | ninetoothed-0.26.0-py3-none-any.whl |
|---|---|
| Size | 47.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
7f15a35b308f00659c63508ff14f964f431234397e5dc97f73e2e4008ab51f08
|
|
BLAKE2b-256 checksum How to use checksums |
22813f083e7dc3138ac56d1fa289d5ff3e5233792693796a691c8f2c53c1a398
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency log
