Mojo module

tensor_core

Tensor Core Module for High-Performance Matrix Operations

Provides abstractions for using GPU Tensor Cores to perform optimized matrix operations. It supports both NVIDIA and AMD GPU architectures with hardware-specific optimizations.

Key Components:

TensorCore: Core struct that encapsulates tensor core operations with support for various data types and matrix shapes. It handles loading matrix fragments, performing matrix multiply-accumulate operations, and storing results.
Matrix Fragment Management: Functions for loading and storing matrix fragments to/from shared memory with hardware-specific optimizations.
Matrix Multiply-Accumulate (MMA): Optimized implementations of matrix multiplication operations using tensor cores.

Supported Operations:

Matrix loading with various layouts and swizzling patterns
Matrix multiply-accumulate (D = A * B + C)
Matrix storing with hardware-specific optimizations

Supported Data Types:

NVIDIA: float32, bfloat16, float16, float8_e4m3fn, float8_e5m2
AMD: float32, bfloat16, float16

Supported Matrix Shapes:

NVIDIA: 16×8×8, 16×8×4, 16×8×16, 8×8×4, 16×8×32
AMD: 16×16×4, 16×16×16, 32×32×8

Aliases

shape_16x16x16 = IndexList(16, 16, 16):
shape_16x16x4 = IndexList(16, 16, 4):
shape_16x8x16 = IndexList(16, 8, 16):
shape_16x8x32 = IndexList(16, 8, 32):
shape_16x8x4 = IndexList(16, 8, 4):
shape_16x8x8 = IndexList(16, 8, 8):
shape_32x32x8 = IndexList(32, 32, 8):
shape_8x8x4 = IndexList(8, 8, 4):
shape_null = IndexList(0, 0, 0):

Structs

TensorCore: TensorCore provides an abstraction for GPU tensor core hardware to perform optimized matrix operations.

Functions

get_fragment_size: Calculates the fragment size per thread for a given MMA shape.
get_mma_shape: Returns the appropriate matrix multiply-accumulate (MMA) shape for tensor core operations.
num_matrix_reg: Calculates the number of matrix registers required per thread.

Key Components:​

Supported Operations:​

Supported Data Types:​

Supported Matrix Shapes:​

Aliases​

Structs​

Functions​