cuSPARSELt: A High-Performance CUDA Library for Sparse Matrix-Matrix Multiplication

NVIDIA cuSPARSELt is a high-performance CUDA library dedicated to general matrix-matrix operations in which at least one operand is a structured sparse matrix with 50% sparsity ratio:

\begin{equation*} D = Activation(\alpha op(A) \cdot op(B) + \beta op(C) + bias) \end{equation*}

where \(op(A)/op(B)\) refers to in-place operations such as transpose/non-transpose, and \(alpha, beta\) are scalars or vectors.

The cuSPARSELt APIs allow flexibility in the algorithm/operation selection, epilogue, and matrix characteristics, including memory layout, alignment, and data types.

Download: developer.nvidia.com/cusparselt/downloads

Provide Feedback: Math-Libs-Feedback@nvidia.com

Examples: cuSPARSELt Example 1, cuSPARSELt Example 2

Blog post:

Key Features

NVIDIA Sparse MMA tensor core support

Mixed-precision computation support:

Input A/B

Input C

Output D

Compute

Block scaled

Support SM arch

FP32

FP32

FP32

FP32

No

8.0, 8.6, 8.7 9.0, 10.0, 10.3 11.0, 12.0, 12.1

BF16

BF16

BF16

FP32

FP16

FP16

FP16

FP32

FP16

FP16

FP16

FP16

No

9.0

INT8

INT8

INT8

INT32

No

8.0, 8.6, 8.7 9.0, 10.0, 11.0 12.0, 12.1

INT32

INT32

FP16

FP16

BF16

BF16

INT8

INT8

INT8

INT32

No

8.0, 8.6, 8.7 9.0, 10.0, 11.0 12.0, 12.1

INT32

INT32

FP16

FP16

BF16

BF16

E4M3

FP16

E4M3

FP32

No

9.0, 10.0, 10.3 11.0, 12.0, 12.1

BF16

E4M3

FP16

FP16

BF16

BF16

FP32

FP32

E5M2

FP16

E5M2

FP32

No

9.0, 10.0, 10.3 11.0, 12.0, 12.1

BF16

E5M2

FP16

FP16

BF16

BF16

FP32

FP32

E4M3

FP16

E4M3

FP32

A/B/D_OUT_SCALE = VEC64_UE8M0

D_SCALE = 32F

10.0, 10.3, 11.0 12.0, 12.1

BF16

E4M3

FP16

FP16

A/B_SCALE = VEC64_UE8M0

BF16

BF16

FP32

FP32

E2M1

FP16

E2M1

FP32

A/B/D_SCALE = VEC32_UE4M3

D_SCALE = 32F

10.0, 10.3, 11.0 12.0, 12.1

BF16

E2M1

FP16

FP16

A/B_SCALE = VEC32_UE4M3

BF16

BF16

FP32

FP32

Matrix pruning and compression functionalities
Activation functions, bias vector, and output scaling
Batched computation (multiple matrices in a single run)
GEMM Split-K mode
Auto-tuning functionality (see cusparseLtMatmulSearch() <cusparseLtMatmulSearch-label>)
NVTX ranging and Logging functionalities

Input A/B	Input C	Output D	Compute	Block scaled	Support SM arch
FP32	FP32	FP32	FP32	No	8.0, 8.6, 8.7 9.0, 10.0, 10.3 11.0, 12.0, 12.1
BF16	BF16	BF16	FP32
FP16	FP16	FP16	FP32
FP16	FP16	FP16	FP16	No	9.0
INT8	INT8	INT8	INT32	No	8.0, 8.6, 8.7 9.0, 10.0, 11.0 12.0, 12.1
INT32	INT32
FP16	FP16
BF16	BF16
INT8	INT8	INT8	INT32	No	8.0, 8.6, 8.7 9.0, 10.0, 11.0 12.0, 12.1
INT32	INT32
FP16	FP16
BF16	BF16
E4M3	FP16	E4M3	FP32	No	9.0, 10.0, 10.3 11.0, 12.0, 12.1
BF16	E4M3
FP16	FP16
BF16	BF16
FP32	FP32
E5M2	FP16	E5M2	FP32	No	9.0, 10.0, 10.3 11.0, 12.0, 12.1
BF16	E5M2
FP16	FP16
BF16	BF16
FP32	FP32
E4M3	FP16	E4M3	FP32	A/B/D_OUT_SCALE = VEC64_UE8M0 D_SCALE = 32F	10.0, 10.3, 11.0 12.0, 12.1
BF16	E4M3
FP16	FP16	A/B_SCALE = VEC64_UE8M0
BF16	BF16
FP32	FP32
E2M1	FP16	E2M1	FP32	A/B/D_SCALE = VEC32_UE4M3 D_SCALE = 32F	10.0, 10.3, 11.0 12.0, 12.1
BF16	E2M1
FP16	FP16	A/B_SCALE = VEC32_UE4M3
BF16	BF16
FP32	FP32

Support

Supported SM Architectures: SM 8.0, SM 8.6, SM 8.7, SM 8.9, SM 9.0, SM 10.0, SM 10.3, SM 11.0, SM 12.0, SM 12.1
Supported CPU architectures and operating systems:

OS	CPU archs
Windows	x86_64
Linux	x86_64, Arm64

Documentation

Please refer to https://docs.nvidia.com/cuda/cusparselt/index.html for the cuSPARSELt documentation.

Installation

The cuSPARSELt wheel can be installed as follows:

pip install nvidia-cusparselt-cuXX

where XX is the CUDA major version.

Metadata

Release files for nvidia-cusparselt-cu13 0.10.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Built distributions (wheels)

Table of built distributions (wheels) for nvidia-cusparselt-cu13 0.10.0
File	Interpreter	ABI	Platform	Reset
nvidia_cusparselt_cu13-0.10.0-py3-none-win_amd64.whl 155.4 MB Sep 30, 2026	Python 3	none	Windows x86-64	Details
nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_x86_64.whl 169.8 MB Sep 30, 2026	Python 3	none	Linux glibc 2.17+ x86-64	Details
nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_aarch64.whl 219.8 MB Sep 30, 2026	Python 3	none	Linux glibc 2.17+ ARM64	Details

Total release size: 545.0 MB

Release files / nvidia_cusparselt_cu13-0.10.0-py3-none-win_amd64.whl

Download URL	nvidia_cusparselt_cu13-0.10.0-py3-none-win_amd64.whl
Size	155.4 MB
Tags	Python 3 Windows x86-64
SHA-256 checksum How to use checksums	`f49f9b12a9c12b99aab1468fdd5fa031965dea88be585d647d60f5d50996bbd1`
BLAKE2b-256 checksum How to use checksums	`848e52267c84b0554a2b4f58079bc44bacd3f444927131e2e8c01a96d34d94a4`
Upload date	Sep 30, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/7.0.0 CPython/3.14.7`

Release files / nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_x86_64.whl

Download URL	nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_x86_64.whl
Size	169.8 MB
Tags	Linux glibc 2.17+ x86-64 Python 3
SHA-256 checksum How to use checksums	`b46f95b786b4ed96193d639527d95c1703f421d14a89afea1437a84ae3c36823`
BLAKE2b-256 checksum How to use checksums	`e7d9e337376cd3ec661feaea2075d50843a0f03dc94367f583674df9a18d2e6b`
Upload date	Sep 30, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/7.0.0 CPython/3.14.7`

Release files / nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_aarch64.whl

Download URL	nvidia_cusparselt_cu13-0.10.0-py3-none-manylinux2014_aarch64.whl
Size	219.8 MB
Tags	Linux glibc 2.17+ ARM64 Python 3
SHA-256 checksum How to use checksums	`6270a6d7b53111a50a2fc8e50b55729c37ab83c70ba83ee1ba954e1b4f5cc34c`
BLAKE2b-256 checksum How to use checksums	`83608e1b286cb38cf9d1a140faabac271205a161c3b77a89af62d8b11100a5fb`
Upload date	Sep 30, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/7.0.0 CPython/3.14.7`