gemm_bias

Computes a matrix-matrix product using general integer matrices with bias.

Description

The gemm_bias routines compute a scalar-matrix-matrix product and add the result to a scalar-matrix product, using general integer matrices with biases/offsets. The operation is defined as:

\[\scriptstyle C \leftarrow alpha*(op(A) - A\_offset)*(op(B) - B\_offset) + beta*C + C\_offset\]

where:

op(X) is one of op(X) = X, or op(X) = X^T, or op(X) = X^H,

alpha and beta are scalars,

A_offset is an m-by-k matrix with every element equal to the value ao,

B_offset is a k-by-n matrix with every element equal to the value bo,

C_offset is an m-by-n matrix defined by the co buffer as described below,

A, B, and C are matrices,

op(A) is m x k, op(B) is k x n, and C is m x n.

gemm_bias supports the following precisions.

Ta	Tb
std::uint8_t	std::uint8_t
std::int8_t	std::uint8_t
std::uint8_t	std::int8_t
std::int8_t	std::int8_t

gemm_bias (Buffer Version)

Syntax

namespace oneapi::math::blas::column_major {
    void gemm_bias(sycl::queue &queue,
                   oneapi::math::transpose transa,
                   oneapi::math::transpose transb,
                   oneapi::math::offset offset_type,
                   std::int64_t m,
                   std::int64_t n,
                   std::int64_t k,
                   float alpha,
                   sycl::buffer<Ta,1> &a,
                   std::int64_t lda,
                   Ta ao,
                   sycl::buffer<Tb,1> &b,
                   std::int64_t ldb,
                   Tb bo,
                   float beta,
                   sycl::buffer<std::int32_t,1> &c,
                   std::int64_t ldc,
                   sycl::buffer<std::int32_t,1> &co)
}

namespace oneapi::math::blas::row_major {
    void gemm_bias(sycl::queue &queue,
                   oneapi::math::transpose transa,
                   oneapi::math::transpose transb,
                   oneapi::math::offset offset_type,
                   std::int64_t m,
                   std::int64_t n,
                   std::int64_t k,
                   float alpha,
                   sycl::buffer<Ta,1> &a,
                   std::int64_t lda,
                   Ta ao,
                   sycl::buffer<Tb,1> &b,
                   std::int64_t ldb,
                   Tb bo,
                   float beta,
                   sycl::buffer<std::int32_t,1> &c,
                   std::int64_t ldc,
                   sycl::buffer<std::int32_t,1> &co)
}

Input Parameters

queue

The queue where the routine should be executed.

transa

Specifies op(A), the transposition operation applied to A. See oneMath defined datatypes for more details.

transb

Specifies op(B), the transposition operation applied to B. See oneMath defined datatypes for more details.

offset_type

Specifies the form of C_offset used in the matrix multiplication. See oneMath defined datatypes for more details.

m

Number of rows of op(A) and C. Must be at least zero.

n

Number of columns of op(B) and C. Must be at least zero.

k

Number of columns of op(A) and rows of op(B). Must be at least zero.

alpha

Scaling factor for the matrix-matrix product.

a

The buffer holding the input matrix A.

	A not transposed	A transposed
Column major	A is an m-by-k matrix so the array a must have size at least lda*k.	A is an k-by-m matrix so the array a must have size at least lda*m
Row major	A is an m-by-k matrix so the array a must have size at least lda*m.	A is an k-by-m matrix so the array a must have size at least lda*k

See Matrix Storage for more details.

lda

The leading dimension of A. It must be positive.

	A not transposed	A transposed
Column major	lda must be at least m.	lda must be at least k.
Row major	lda must be at least k.	lda must be at least m.

ao

Specifies the scalar offset value for matrix A.

b

Buffer holding the input matrix B.

	B not transposed	B transposed
Column major	B is an k-by-n matrix so the array b must have size at least ldb*n.	B is an n-by-k matrix so the array b must have size at least ldb*k
Row major	B is an k-by-n matrix so the array b must have size at least ldb*k.	B is an n-by-k matrix so the array b must have size at least ldb*n

See Matrix Storage for more details.

ldb

The leading dimension of B. It must be positive.

	B not transposed	B transposed
Column major	ldb must be at least k.	ldb must be at least n.
Row major	ldb must be at least n.	ldb must be at least k.

bo

Specifies the scalar offset value for matrix B.

beta

Scaling factor for matrix C.

c

Buffer holding the input/output matrix C. It must have a size of at least ldc*n if column major layout is used to store matrices or at least ldc*m if row major layout is used to store matrices . See Matrix Storage for more details.

ldc

The leading dimension of C. It must be positive and at least m if column major layout is used to store matrices or at least n if row major layout is used to store matrices.

co

Buffer holding the offset values for matrix C.

If offset_type = offset::fix, the co array must have size at least 1.

If offset_type = offset::col, the co array must have size at least max(1,m).

If offset_type = offset::row, the co array must have size at least max(1,n).

Output Parameters

c

Output buffer, overwritten by alpha * (op(A) - A_offset)*(op(B) - B_offset) + beta * C + C_offset.

Notes

If beta = 0, matrix C does not need to be initialized before calling gemm_bias.

Throws

This routine shall throw the following exceptions if the associated condition is detected. An implementation may throw additional implementation-specific exception(s) in case of error conditions not covered here.

oneapi::math::invalid_argument

oneapi::math::unsupported_device

oneapi::math::host_bad_alloc

oneapi::math::device_bad_alloc

oneapi::math::unimplemented

gemm_bias (USM Version)

Syntax

namespace oneapi::math::blas::column_major {
    sycl::event gemm_bias(sycl::queue &queue,
                          oneapi::math::transpose transa,
                          oneapi::math::transpose transb,
                          oneapi::math::offset offset_type,
                          std::int64_t m,
                          std::int64_t n,
                          std::int64_t k,
                          value_or_pointer<float> alpha,
                          const Ta *a,
                          std::int64_t lda,
                          Ta ao,
                          const Tb *b,
                          std::int64_t ldb,
                          Tb bo,
                          value_or_pointer<float> beta,
                          std::int32_t *c,
                          std::int64_t ldc,
                          const std::int32_t *co,
                          const std::vector<sycl::event> &dependencies = {})
}

namespace oneapi::math::blas::row_major {
    sycl::event gemm_bias(sycl::queue &queue,
                          oneapi::math::transpose transa,
                          oneapi::math::transpose transb,
                          oneapi::math::offset offset_type,
                          std::int64_t m,
                          std::int64_t n,
                          std::int64_t k,
                          value_or_pointer<float> alpha,
                          const Ta *a,
                          std::int64_t lda,
                          Ta ao,
                          const Tb *b,
                          std::int64_t ldb,
                          Tb bo,
                          value_or_pointer<float> beta,
                          std::int32_t *c,
                          std::int64_t ldc,
                          const std::int32_t *co,
                          const std::vector<sycl::event> &dependencies = {})
}

Input Parameters

queue

The queue where the routine should be executed.

transa

Specifies op(A), the transposition operation applied to A. See oneMath defined datatypes for more details.

transb

Specifies op(B), the transposition operation applied to B. See oneMath defined datatypes for more details.

offset_type

Specifies the form of C_offset used in the matrix multiplication. See oneMath defined datatypes for more details.

m

Number of rows of op(A) and C. Must be at least zero.

n

Number of columns of op(B) and C. Must be at least zero.

k

Number of columns of op(A) and rows of op(B). Must be at least zero.

alpha

Scaling factor for the matrix-matrix product. See Scalar Arguments in BLAS for more details.

a

Pointer to input matrix A.

	A not transposed	A transposed
Column major	A is an m-by-k matrix so the array a must have size at least lda*k.	A is an k-by-m matrix so the array a must have size at least lda*m
Row major	A is an m-by-k matrix so the array a must have size at least lda*m.	A is an k-by-m matrix so the array a must have size at least lda*k

See Matrix Storage for more details.

lda

The leading dimension of A. It must be positive.

	A not transposed	A transposed
Column major	lda must be at least m.	lda must be at least k.
Row major	lda must be at least k.	lda must be at least m.

ao

Specifies the scalar offset value for matrix A.

b

Pointer to input matrix B.

	B not transposed	B transposed
Column major	B is an k-by-n matrix so the array b must have size at least ldb*n.	B is an n-by-k matrix so the array b must have size at least ldb*k
Row major	B is an k-by-n matrix so the array b must have size at least ldb*k.	B is an n-by-k matrix so the array b must have size at least ldb*n

See Matrix Storage for more details.

ldb

The leading dimension of B. It must be positive.

	B not transposed	B transposed
Column major	ldb must be at least k.	ldb must be at least n.
Row major	ldb must be at least n.	ldb must be at least k.

bo

Specifies the scalar offset value for matrix B.

beta

Scaling factor for matrix C. See Scalar Arguments in BLAS for more details.

c

Pointer to input/output matrix C. It must have a size of at least ldc*n if column major layout is used to store matrices or at least ldc*m if row major layout is used to store matrices . See Matrix Storage for more details.

ldc

The leading dimension of C. It must be positive and at least m if column major layout is used to store matrices or at least n if row major layout is used to store matrices.

co

Pointer to offset values for matrix C.

If offset_type = offset::fix, the co array must have size at least 1.

If offset_type = offset::col, the co array must have size at least max(1,m).

If offset_type = offset::row, the co array must have size at least max(1,n).

dependencies

List of events to wait for before starting computation, if any. If omitted, defaults to no dependencies.

Output Parameters

c

Pointer to the output matrix, overwritten by alpha * (op(A) - A_offset)*(op(B) - B_offset) + beta * C + C_offset.

Notes

If beta = 0, matrix C does not need to be initialized before calling gemm_bias.

Return Values

Output event to wait on to ensure computation is complete.

Throws

This routine shall throw the following exceptions if the associated condition is detected. An implementation may throw additional implementation-specific exception(s) in case of error conditions not covered here.

oneapi::math::invalid_argument

oneapi::math::unsupported_device

oneapi::math::host_bad_alloc

oneapi::math::device_bad_alloc

oneapi::math::unimplemented

Parent topic: BLAS-like Extensions