API

Public kernel functions declared in arm_nnfunctions*.h: 426 across 2 headers. Start with an operator family, then follow a function through to its parameters, return values and source.

Machine-readable model · llms.txt · llms-full.txt

compute

Convolution→

126 functions

Conv2D, depthwise, transpose convolution, wrappers, and buffer helpers.

compute

Fully connected→

33 functions

Dense layers, batch matmul paths, and scratch sizing helpers.

graph glue

Elementwise→

65 functions

Add, sub, mul, square difference, min/max, batch norm, select, and arithmetic glue.

selection

Reduction and comparison→

40 functions

Argmin/argmax, min/max reductions, comparisons, means, vector sums, and where.

activation

Activation→

27 functions

ReLU, LeakyReLU, PReLU, Hard-Swish, Logistic, Tanh, and clamp.

layout

Data movement→

77 functions

Pad, reshape, transpose, concatenate, gather, resize, strided slice, tile, broadcast, scatter, mirror pad, and sequence/slice-update utilities.

tail

Pooling, softmax, quantization→

27 functions

Classifier tail APIs plus dtype conversion and requantization utilities.

sequence

Sequence→

31 functions

LSTM, GRU, and SVDF functions for temporal workloads.

Convolution Functions

Conv2D, depthwise, transpose convolution, wrappers, and buffer helpers. 126 functions.

Function	Summary	Module
arm_convolve_1_x_n_f16	1xN convolution, dispatch by layout.	Convolution Functions
arm_convolve_1_x_n_f16_acc16	1xN convolution, dispatch by layout.	Convolution Functions
arm_convolve_1_x_n_f16_get_buffer_size	Get the buffer size required by 1xN convolution.	Convolution Functions
arm_convolve_1_x_n_f32	1xN convolution, dispatch by layout.	Convolution Functions
arm_convolve_1_x_n_f32_get_buffer_size	Get the buffer size required by 1xN convolution.	Convolution Functions
arm_convolve_1_x_n_nhwc_f16	1xN convolution, NHWC layout.	Convolution Functions
arm_convolve_1_x_n_nhwc_f16_acc16	1xN convolution, NHWC layout.	Convolution Functions
arm_convolve_1_x_n_nhwc_f32	1xN convolution, NHWC layout.	Convolution Functions
arm_convolve_1_x_n_s4	1xn convolution for s4 weights	arm_nnfunctions.h
arm_convolve_1_x_n_s4_get_buffer_size	Get the required additional buffer size for 1xn convolution.	arm_nnfunctions.h
arm_convolve_1_x_n_s8	1xn convolution	arm_nnfunctions.h
arm_convolve_1_x_n_s8_get_buffer_size	Get the required additional buffer size for 1xn convolution.	arm_nnfunctions.h
arm_convolve_1x1_f16	1x1 convolution, dispatch by layout.	Convolution Functions
arm_convolve_1x1_f16_acc16	1x1 convolution, dispatch by layout.	Convolution Functions
arm_convolve_1x1_f16_get_buffer_size	Get the buffer size required by 1x1 convolution.	Convolution Functions
arm_convolve_1x1_f32	1x1 convolution, dispatch by layout.	Convolution Functions
arm_convolve_1x1_f32_get_buffer_size	Get the buffer size required by 1x1 convolution.	Convolution Functions
arm_convolve_1x1_nhwc_f16	1x1 convolution, NHWC layout.	Convolution Functions
arm_convolve_1x1_nhwc_f16_acc16	1x1 convolution, NHWC layout.	Convolution Functions
arm_convolve_1x1_nhwc_f32	1x1 convolution, NHWC layout.	Convolution Functions
arm_convolve_1x1_out_s8	Optimised convolution for 1x1 output images (shape of BX1x1xCOUT) for 8x8 computations.	arm_nnfunctions.h
arm_convolve_1x1_out_s8_get_buffer_size	Get the required scratch buffer size for armconvolve1x1outs8().	arm_nnfunctions.h
arm_convolve_1x1_s16_ns_np_nd	Pointwise s16 convolution function: no stride, no padding, no dilation.	arm_nnfunctions.h
arm_convolve_1x1_s4	s4 version for 1x1 convolution with support for non-unity stride values	arm_nnfunctions.h
arm_convolve_1x1_s4_fast	Fast s4 version for 1x1 convolution (non-square shape).	arm_nnfunctions.h
arm_convolve_1x1_s4_fast_get_buffer_size	Get the required buffer size for armconvolve1x1s4fast.	arm_nnfunctions.h
arm_convolve_1x1_s8	s8 version for 1x1 convolution with support for non-unity stride values	arm_nnfunctions.h
arm_convolve_1x1_s8_fast	Fast s8 version for 1x1 convolution (non-square shape).	arm_nnfunctions.h
arm_convolve_1x1_s8_fast_get_buffer_size	Get the required buffer size for armconvolve1x1s8fast.	arm_nnfunctions.h
arm_convolve_even_s4	Basic s4 convolution function with a requirement of even number of kernels.	arm_nnfunctions.h
arm_convolve_even_s4_get_buffer_size	Get the required buffer size for armconvolveevens4.	arm_nnfunctions.h
arm_convolve_f16	Convolution, dispatch by layout.	Convolution Functions
arm_convolve_f16_acc16	Convolution, dispatch by layout.	Convolution Functions
arm_convolve_f16_get_buffer_size	Get the temporary buffer size required by convolution.	Convolution Functions
arm_convolve_f32	Convolution, dispatch by layout.	Convolution Functions
arm_convolve_f32_get_buffer_size	Get the temporary buffer size required by convolution.	Convolution Functions
arm_convolve_nhwc_f16	Convolution, NHWC layout.	Convolution Functions
arm_convolve_nhwc_f16_acc16	Convolution, NHWC layout.	Convolution Functions
arm_convolve_nhwc_f32	Convolution, NHWC layout.	Convolution Functions
arm_convolve_s16	Basic s16 convolution function.	arm_nnfunctions.h
arm_convolve_s16_fast_small_kernel	armconvolves16fastsmallkernel function.	arm_nnfunctions.h
arm_convolve_s16_get_buffer_size	Get the required buffer size for s16 convolution function.	arm_nnfunctions.h
arm_convolve_s16_group_ch_mult_1	s16 grouped convolution optimized for the case where filterdims->c == 1 and inputch == outputch (channel multiplier = 1).	arm_nnfunctions.h
arm_convolve_s4	Basic s4 convolution function.	arm_nnfunctions.h
arm_convolve_s4_get_buffer_size	Get the required buffer size for s4 convolution function.	arm_nnfunctions.h
arm_convolve_s8	Basic s8 convolution function.	arm_nnfunctions.h
arm_convolve_s8_3x3_c16_s1	s8 3x3 convolution over 16 input channels with unit stride.	arm_nnfunctions.h
arm_convolve_s8_get_buffer_size	Get the required buffer size for s8 convolution function.	arm_nnfunctions.h
arm_convolve_s8_get_buffer_size_mve	Get the required buffer size for armconvolves8 for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_convolve_s8_get_weights_sum_size	Get the required buffer size for s8 convolution and depthwise convolution weight sum.	arm_nnfunctions.h
arm_convolve_s8_small_cin	s8 convolution for input depths of 1 to 3, such as the first layer of an image or audio model.	arm_nnfunctions.h
arm_convolve_weight_sum	Pre-computes per-output-channel weight sums for a standard convolution.	arm_nnfunctions.h
arm_convolve_wrapper_f16	Convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_convolve_wrapper_f16_acc16	Convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_convolve_wrapper_f16_get_buffer_size	Get the buffer size required by the convolution wrapper.	Convolution Functions
arm_convolve_wrapper_f32	Convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_convolve_wrapper_f32_get_buffer_size	Get the buffer size required by the convolution wrapper.	Convolution Functions
arm_convolve_wrapper_s16	s16 convolution layer wrapper function with the main purpose to call the optimal kernel available in cmsis-nn to perform the convolution.	arm_nnfunctions.h
arm_convolve_wrapper_s16_get_buffer_size	Get the required buffer size for armconvolvewrappers16.	arm_nnfunctions.h
arm_convolve_wrapper_s16_get_buffer_size_dsp	Get the required buffer size for armconvolvewrappers16 for for processors with DSP extension.	arm_nnfunctions.h
arm_convolve_wrapper_s16_get_buffer_size_mve	Get the required buffer size for armconvolvewrappers16 for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_convolve_wrapper_s4	s4 convolution layer wrapper function with the main purpose to call the optimal kernel available in cmsis-nn to perform the convolution.	arm_nnfunctions.h
arm_convolve_wrapper_s4_get_buffer_size	Get the required buffer size for armconvolvewrappers4.	arm_nnfunctions.h
arm_convolve_wrapper_s4_get_buffer_size_dsp	Get the required buffer size for armconvolvewrappers4 for processors with DSP extension.	arm_nnfunctions.h
arm_convolve_wrapper_s4_get_buffer_size_mve	Get the required buffer size for armconvolvewrappers4 for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_convolve_wrapper_s8	s8 convolution layer wrapper function with the main purpose to call the optimal kernel available in cmsis-nn to perform the convolution.	arm_nnfunctions.h
arm_convolve_wrapper_s8_get_buffer_size	Get the required buffer size for armconvolvewrappers8.	arm_nnfunctions.h
arm_convolve_wrapper_s8_get_buffer_size_dsp	Get the required buffer size for armconvolvewrappers8 for processors with DSP extension.	arm_nnfunctions.h
arm_convolve_wrapper_s8_get_buffer_size_mve	Get the required buffer size for armconvolvewrappers8 for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_depthwise_conv_3x3_s8	Optimized s8 depthwise convolution function for 3x3 kernel size with some constraints on the input arguments(documented below).	arm_nnfunctions.h
arm_depthwise_conv_f16	Depthwise convolution, dispatch by layout.	Convolution Functions
arm_depthwise_conv_f16_acc16	Depthwise convolution, dispatch by layout.	Convolution Functions
arm_depthwise_conv_f16_get_buffer_size	Get the temporary buffer size required by depthwise convolution.	Convolution Functions
arm_depthwise_conv_f32	Depthwise convolution, dispatch by layout.	Convolution Functions
arm_depthwise_conv_f32_get_buffer_size	Get the temporary buffer size required by depthwise convolution.	Convolution Functions
arm_depthwise_conv_fast_s16	Optimized s16 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_fast_s16_get_buffer_size	Get the required buffer size for optimized s16 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_s16	Basic s16 depthwise convolution function that doesn’t have any constraints on the input dimensions.	arm_nnfunctions.h
arm_depthwise_conv_s4	Basic s4 depthwise convolution function that doesn’t have any constraints on the input dimensions.	arm_nnfunctions.h
arm_depthwise_conv_s4_opt	Optimized s4 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_s4_opt_get_buffer_size	Get the required buffer size for optimized s4 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_s8	Basic s8 depthwise convolution function that doesn’t have any constraints on the input dimensions.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt	Optimized s8 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_3x3	s8 3x3 depthwise convolution that reads the input in place, without the im2col copy of armdepthwiseconvs8opt(), for layers in its gate.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_3x3_c64_s1	armdepthwiseconvs8opt3x3() specialized for 64 channels and a vertical stride of 1, with fixed channel offsets and edge columns that skip their zero taps.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_3x3_get_buffer_size	Get the scratch size in bytes of armdepthwiseconvs8opt3x3() and armdepthwiseconvs8opt3x3c64s1().	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_channelwise	The channel-vectorized path of armdepthwiseconvs8opt() on its own, without the planar attempt.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_get_buffer_size	Get the required buffer size for optimized s8 depthwise convolution function with constraint that inchannel equals outchannel.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_planar	The planar path of armdepthwiseconvs8opt() on its own: s8 depthwise convolution vectorized across the output pixels of each channel plane.	arm_nnfunctions.h
arm_depthwise_conv_s8_opt_planar_supported	Whether armdepthwiseconvs8opt() runs a layer on its planar path, vectorized across the output pixels of each channel plane rather than across channels.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_f16	Depthwise convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_depthwise_conv_wrapper_f16_acc16	Depthwise convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_depthwise_conv_wrapper_f16_get_buffer_size	Get the buffer size required by the depthwise convolution wrapper.	Convolution Functions
arm_depthwise_conv_wrapper_f32	Depthwise convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_depthwise_conv_wrapper_f32_get_buffer_size	Get the buffer size required by the depthwise convolution wrapper.	Convolution Functions
arm_depthwise_conv_wrapper_s16	Wrapper function to pick the right optimized s16 depthwise convolution function.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s16_get_buffer_size	Get size of additional buffer required by armdepthwiseconvwrappers16().	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s16_get_buffer_size_dsp	Get size of additional buffer required by armdepthwiseconvwrappers16() for processors with DSP extension.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s16_get_buffer_size_mve	Get size of additional buffer required by armdepthwiseconvwrappers16() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s4	Wrapper function to pick the right optimized s4 depthwise convolution function.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s4_get_buffer_size	Get size of additional buffer required by armdepthwiseconvwrappers4().	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s4_get_buffer_size_dsp	Get size of additional buffer required by armdepthwiseconvwrappers4() for processors with DSP extension.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s4_get_buffer_size_mve	Get size of additional buffer required by armdepthwiseconvwrappers4() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s8	Wrapper function to pick the right optimized s8 depthwise convolution function.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s8_get_buffer_size	Get size of additional buffer required by armdepthwiseconvwrappers8().	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s8_get_buffer_size_dsp	Get size of additional buffer required by armdepthwiseconvwrappers8() for processors with DSP extension.	arm_nnfunctions.h
arm_depthwise_conv_wrapper_s8_get_buffer_size_mve	Get size of additional buffer required by armdepthwiseconvwrappers8() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_depthwise_convolve_weight_sum	Pre-computes per-channel weight sums for a depthwise convolution.	arm_nnfunctions.h
arm_depthwise_nhwc_conv_f16	Depthwise convolution, NHWC layout.	Convolution Functions
arm_depthwise_nhwc_conv_f16_acc16	Depthwise convolution, NHWC layout.	Convolution Functions
arm_depthwise_nhwc_conv_f32	Depthwise convolution, NHWC layout.	Convolution Functions
arm_transpose_conv_f16	Transpose convolution, dispatch by layout.	Convolution Functions
arm_transpose_conv_f16_get_buffer_size	Get the temporary buffer size required by transpose convolution.	Convolution Functions
arm_transpose_conv_f16_get_reverse_conv_buffer_size	Get the reverse-convolution workspace size used by transpose convolution helpers.	Convolution Functions
arm_transpose_conv_f32	Transpose convolution, dispatch by layout.	Convolution Functions
arm_transpose_conv_f32_get_buffer_size	Get the temporary buffer size required by transpose convolution.	Convolution Functions
arm_transpose_conv_f32_get_reverse_conv_buffer_size	Get the reverse-convolution workspace size used by transpose convolution helpers.	Convolution Functions
arm_transpose_conv_nhwc_f16	Transpose convolution, NHWC layout.	Convolution Functions
arm_transpose_conv_nhwc_f32	Transpose convolution, NHWC layout.	Convolution Functions
arm_transpose_conv_s8	Basic s8 transpose convolution function.	arm_nnfunctions.h
arm_transpose_conv_s8_get_buffer_size	Get the required buffer size for ctx in s8 transpose conv function.	arm_nnfunctions.h
arm_transpose_conv_s8_get_buffer_size_mve	Get size of additional buffer required by armtransposeconvs8() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_transpose_conv_s8_get_reverse_conv_buffer_size	Get the required buffer size for outputctx in s8 transpose conv function.	arm_nnfunctions.h
arm_transpose_conv_wrapper_f16	Transpose convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_transpose_conv_wrapper_f32	Transpose convolution wrapper using the CMSIS-NN baseline path.	Convolution Functions
arm_transpose_conv_wrapper_s8	Wrapper to select optimal transposed convolution algorithm depending on parameters.	arm_nnfunctions.h

Fully-Connected Layer Functions

Dense layers, batch matmul paths, and scratch sizing helpers. 33 functions.

Function	Summary	Module
arm_batch_matmul_f16	Batched matrix multiplication.	Fully-connected Layer Functions
arm_batch_matmul_f16_get_buffer_size	Get the temporary buffer size required by batched matrix multiplication.	Fully-connected Layer Functions
arm_batch_matmul_f32	Batched matrix multiplication.	Fully-connected Layer Functions
arm_batch_matmul_f32_get_buffer_size	Get the temporary buffer size required by batched matrix multiplication.	Fully-connected Layer Functions
arm_batch_matmul_s16	Batch matmul function with 16 bit input and output.	arm_nnfunctions.h
arm_batch_matmul_s8	Batch matmul function with 8 bit input and output.	arm_nnfunctions.h
arm_batch_matmul_s8_get_buffer_size	Get size of the scratch buffer required by armbatchmatmuls8().	arm_nnfunctions.h
arm_batch_matmul_s8_get_buffer_size_dsp	Get size of the scratch buffer required by armbatchmatmuls8() for processors with DSP extension.	arm_nnfunctions.h
arm_batch_matmul_s8_get_buffer_size_mve	Get size of the scratch buffer required by armbatchmatmuls8() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_fully_connected_f16	Fully connected layer, dispatch by layout.	Fully-connected Layer Functions
arm_fully_connected_f16_acc16	Fully connected layer, dispatch by layout.	Fully-connected Layer Functions
arm_fully_connected_f16_get_buffer_size	Get the temporary buffer size required by the fully connected layer.	Fully-connected Layer Functions
arm_fully_connected_f32	Fully connected layer, dispatch by layout.	Fully-connected Layer Functions
arm_fully_connected_f32_get_buffer_size	Get the temporary buffer size required by the fully connected layer.	Fully-connected Layer Functions
arm_fully_connected_nhwc_f16	Fully connected layer, NHWC layout.	Fully-connected Layer Functions
arm_fully_connected_nhwc_f16_acc16	Fully connected layer, NHWC layout.	Fully-connected Layer Functions
arm_fully_connected_nhwc_f32	Fully connected layer, NHWC layout.	Fully-connected Layer Functions
arm_fully_connected_per_channel_s16	Basic s16 Fully Connected function using per channel quantization.	arm_nnfunctions.h
arm_fully_connected_per_channel_s16_get_buffer_size	Get size of additional buffer required by armfullyconnectedperchannels16().	arm_nnfunctions.h
arm_fully_connected_per_channel_s16_get_buffer_size_dsp	Get size of additional buffer required by armfullyconnectedperchannels16() for processors with DSP extension.	arm_nnfunctions.h
arm_fully_connected_per_channel_s16_get_buffer_size_mve	Get size of additional buffer required by armfullyconnectedperchannels16() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_fully_connected_per_channel_s8	Basic s8 Fully Connected function using per channel quantization.	arm_nnfunctions.h
arm_fully_connected_s16	Basic s16 Fully Connected function.	arm_nnfunctions.h
arm_fully_connected_s16_get_buffer_size	Get size of additional buffer required by armfullyconnecteds16().	arm_nnfunctions.h
arm_fully_connected_s16_get_buffer_size_dsp	Get size of additional buffer required by armfullyconnecteds16() for processors with DSP extension.	arm_nnfunctions.h
arm_fully_connected_s16_get_buffer_size_mve	Get size of additional buffer required by armfullyconnecteds16() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_fully_connected_s4	Basic s4 Fully Connected function.	arm_nnfunctions.h
arm_fully_connected_s8	Basic s8 Fully Connected function.	arm_nnfunctions.h
arm_fully_connected_s8_get_buffer_size	Get size of additional buffer required by armfullyconnecteds8().	arm_nnfunctions.h
arm_fully_connected_s8_get_buffer_size_dsp	Get size of additional buffer required by armfullyconnecteds8() for processors with DSP extension.	arm_nnfunctions.h
arm_fully_connected_s8_get_buffer_size_mve	Get size of additional buffer required by armfullyconnecteds8() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_fully_connected_wrapper_s16	s16 Fully Connected layer wrapper function	arm_nnfunctions.h
arm_fully_connected_wrapper_s8	s8 Fully Connected layer wrapper function	arm_nnfunctions.h

Elementwise Functions

Add, sub, mul, square difference, min/max, batch norm, select, and arithmetic glue. 65 functions.

Function	Summary	Module
arm_abs_s16	s16 elementwise absolute value	arm_nnfunctions.h
arm_abs_s8	s8 elementwise absolute value	arm_nnfunctions.h
arm_add_s16	s16 elementwise add of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_add_s8	s8 elementwise add of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_add_scalar_s16	s16 elementwise add of scalar and vector	arm_nnfunctions.h
arm_add_scalar_s8	s8 elementwise add of scalar and vector	arm_nnfunctions.h
arm_batch_norm_f16	Apply batch normalization.	NNSupport
arm_batch_norm_f32	Apply batch normalization.	NNSupport
arm_elementwise_add_broadcast_f16	Elementwise add with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_add_broadcast_f32	Elementwise add with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_add_f16	Elementwise add with optional output clamp.	Elementwise Functions
arm_elementwise_add_f32	Elementwise add with optional output clamp.	Elementwise Functions
arm_elementwise_add_fp16	Legacy float16 elementwise add with fused clamp, kept only for source compatibility with callers that predate armelementwiseaddf16().	Elementwise Functions
arm_elementwise_add_s16	s16 elementwise add of two vectors	arm_nnfunctions.h
arm_elementwise_add_s8	s8 elementwise add of two vectors	arm_nnfunctions.h
arm_elementwise_mul_broadcast_f16	Elementwise multiply with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_mul_broadcast_f32	Elementwise multiply with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_mul_f16	Elementwise multiply with optional output clamp.	Elementwise Functions
arm_elementwise_mul_f32	Elementwise multiply with optional output clamp.	Elementwise Functions
arm_elementwise_mul_s16	s16 elementwise multiplication	arm_nnfunctions.h
arm_elementwise_mul_s8	s8 elementwise multiplication	arm_nnfunctions.h
arm_elementwise_prelu_s16	Elementwise S16 PReLU activation function.	arm_nnfunctions.h
arm_elementwise_prelu_s8	Elementwise S8 PReLU activation function.	arm_nnfunctions.h
arm_elementwise_squared_difference_f16	Elementwise squared difference of two float16 vectors.	Elementwise Functions
arm_elementwise_squared_difference_s16	s16 elementwise squared difference of two vectors.	arm_nnfunctions.h
arm_elementwise_squared_difference_s8	s8 elementwise squared difference of two vectors.	arm_nnfunctions.h
arm_elementwise_sub_broadcast_f16	Elementwise subtract with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_sub_broadcast_f32	Elementwise subtract with TensorFlow Lite NHWC broadcasting and an output clamp.	Elementwise Functions
arm_elementwise_sub_f16	Elementwise subtract with optional output clamp.	Elementwise Functions
arm_elementwise_sub_f32	Elementwise subtract with optional output clamp.	Elementwise Functions
arm_elementwise_sub_s16	s16 elementwise subtract of two vectors	arm_nnfunctions.h
arm_elementwise_sub_s8	s8 elementwise subtract of two vectors	arm_nnfunctions.h
arm_maximum_f16	Elementwise maximum with TensorFlow Lite NHWC broadcasting: each dimension of the two inputs must be equal or 1, and outputdims must be their broadcast shape.	Elementwise Functions
arm_maximum_f32	Elementwise maximum with TensorFlow Lite NHWC broadcasting: each dimension of the two inputs must be equal or 1, and outputdims must be their broadcast shape.	Elementwise Functions
arm_maximum_s16	s16 elementwise maximum w/ support for broadcasting and scalar inputs.	arm_nnfunctions.h
arm_maximum_s8	s8 elementwise maximum w/ support for broadcasting and scalar inputs.	arm_nnfunctions.h
arm_minimum_f16	Elementwise minimum with TensorFlow Lite NHWC broadcasting: each dimension of the two inputs must be equal or 1, and outputdims must be their broadcast shape.	Elementwise Functions
arm_minimum_f32	Elementwise minimum with TensorFlow Lite NHWC broadcasting: each dimension of the two inputs must be equal or 1, and outputdims must be their broadcast shape.	Elementwise Functions
arm_minimum_s16	s16 elementwise minimum w/ support for broadcasting and scalar inputs.	arm_nnfunctions.h
arm_minimum_s8	s8 elementwise minimum w/ support for broadcasting and scalar inputs.	arm_nnfunctions.h
arm_mul_s16	s16 elementwise multiplication of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_mul_s8	s8 elementwise multiplication of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_mul_scalar_s16	s16 elementwise multiplication of scalar and vector	arm_nnfunctions.h
arm_mul_scalar_s8	s8 elementwise multiplication of scalar and vector	arm_nnfunctions.h
arm_nn_abs_f16	Elementwise absolute value.	Elementwise Functions
arm_nn_abs_f32	Elementwise absolute value.	Elementwise Functions
arm_nn_sqrt_f16	Elementwise square root of a float16 tensor.	Elementwise Functions
arm_nn_sqrt_f32	Elementwise square root.	Elementwise Functions
arm_rsqrt_f16	Elementwise reciprocal square root of a float16 tensor, 1 / sqrt(x).	Elementwise Functions
arm_rsqrt_f32	Elementwise reciprocal square root, 1 / sqrt(x).	Elementwise Functions
arm_rsqrt_s16_per_op	INT16 reciprocal square root using a per-operator LUT.	arm_nnfunctions.h
arm_rsqrt_s16_universal	INT16 reciprocal square root using a shared universal LUT.	arm_nnfunctions.h
arm_select_v2_s16	SELECTV2 with broadcast for int16 tensors.	arm_nnfunctions.h
arm_select_v2_s8	SELECTV2 with broadcast for int8 tensors.	arm_nnfunctions.h
arm_sqrt_s16	s16 elementwise square root using piecewise LUT with linear interpolation	arm_nnfunctions.h
arm_sqrt_s16_tablefree	s16 elementwise square root without a lookup table	arm_nnfunctions.h
arm_sqrt_s8	s8 elementwise square root	arm_nnfunctions.h
arm_squared_difference_s16	s16 elementwise squared difference of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_squared_difference_s8	s8 elementwise squared difference of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_squared_difference_scalar_s16	s16 elementwise squared difference of scalar and vector.	arm_nnfunctions.h
arm_squared_difference_scalar_s8	s8 elementwise squared difference of scalar and vector.	arm_nnfunctions.h
arm_sub_s16	s16 elementwise subtraction of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_sub_s8	s8 elementwise subtraction of two tensors with support for broadcasting.	arm_nnfunctions.h
arm_sub_scalar_s16	s16 elementwise subtract of scalar and vector (scalar - vector)	arm_nnfunctions.h
arm_sub_scalar_s8	s8 elementwise subtract of scalar and vector (scalar - vector)	arm_nnfunctions.h

Basic Math and Reduction

Argmin/argmax, min/max reductions, comparisons, means, vector sums, and where. 40 functions.

Function	Summary	Module
arm_argmax_f16	Returns the first maximum’s axis-relative INT32 index for a f16 tensor.	Reduction Functions
arm_argmax_f32	Returns the first maximum’s axis-relative INT32 index for a f32 tensor.	Reduction Functions
arm_argmax_s16	Compute ArgMax indices of an s16 tensor along a specific axis.	arm_nnfunctions.h
arm_argmax_s8	Compute ArgMax indices of an s8 tensor along a specific axis.	arm_nnfunctions.h
arm_argmin_f16	Returns the first minimum’s axis-relative INT32 index for a f16 tensor.	Reduction Functions
arm_argmin_f32	Returns the first minimum’s axis-relative INT32 index for a f32 tensor.	Reduction Functions
arm_argmin_s16	Compute ArgMin indices of an s16 tensor along a specific axis.	arm_nnfunctions.h
arm_argmin_s8	Compute ArgMin indices of an s8 tensor along a specific axis.	arm_nnfunctions.h
arm_comparison_s16	s16 elementwise comparison with support for broadcasting.	arm_nnfunctions.h
arm_comparison_s8	s8 elementwise comparison with support for broadcasting.	arm_nnfunctions.h
arm_equal_s16	s16 elementwise equality comparison with support for broadcasting.	arm_nnfunctions.h
arm_equal_s8	s8 elementwise equality comparison with support for broadcasting.	arm_nnfunctions.h
arm_greater_equal_s16	s16 elementwise greater-or-equal comparison with support for broadcasting.	arm_nnfunctions.h
arm_greater_equal_s8	s8 elementwise greater-or-equal comparison with support for broadcasting.	arm_nnfunctions.h
arm_greater_s16	s16 elementwise greater-than comparison with support for broadcasting.	arm_nnfunctions.h
arm_greater_s8	s8 elementwise greater-than comparison with support for broadcasting.	arm_nnfunctions.h
arm_less_equal_s16	s16 elementwise less-or-equal comparison with support for broadcasting.	arm_nnfunctions.h
arm_less_equal_s8	s8 elementwise less-or-equal comparison with support for broadcasting.	arm_nnfunctions.h
arm_less_s16	s16 elementwise less-than comparison with support for broadcasting.	arm_nnfunctions.h
arm_less_s8	s8 elementwise less-than comparison with support for broadcasting.	arm_nnfunctions.h
arm_mean_s16	Computes the mean of the input tensor along the specified axis.	arm_nnfunctions.h
arm_mean_s8	Computes the mean of the input tensor along the specified axis.	arm_nnfunctions.h
arm_nn_mean_f16	Computes the mean of a float16 tensor along the specified axes.	Reduction Functions
arm_nn_mean_f32	Computes the mean of a float32 tensor along the specified axes.	Reduction Functions
arm_not_equal_s16	s16 elementwise inequality comparison with support for broadcasting.	arm_nnfunctions.h
arm_not_equal_s8	s8 elementwise inequality comparison with support for broadcasting.	arm_nnfunctions.h
arm_reduce_max_f16	Reduces a f16 NHWC tensor to its maximum along a binary axis mask.	Reduction Functions
arm_reduce_max_f32	Reduces a f32 NHWC tensor to its maximum along a binary axis mask.	Reduction Functions
arm_reduce_max_s16	Computes the max of the input tensor along the specified axis.	arm_nnfunctions.h
arm_reduce_max_s8	Computes the max of the input tensor along the specified axis.	arm_nnfunctions.h
arm_reduce_min_f16	Reduces a f16 NHWC tensor to its minimum along a binary axis mask.	Reduction Functions
arm_reduce_min_f32	Reduces a f32 NHWC tensor to its minimum along a binary axis mask.	Reduction Functions
arm_reduce_min_s16	Computes the min of the input tensor along the specified axis.	arm_nnfunctions.h
arm_reduce_min_s8	Computes the min of the input tensor along the specified axis.	arm_nnfunctions.h
arm_reduce_sum_f16	Computes the sum of the input tensor along the specified axes.	Reduction Functions
arm_reduce_sum_f32	Computes the sum of the input tensor along the specified axes.	Reduction Functions
arm_vector_sum_s8	Calculate the sum of each row in vectordata, multiply by lhsoffset and optionally add s32 biasdata.	arm_nnfunctions.h
arm_vector_sum_s8_s64	Calculate the sum of each row in vectordata, multiply by lhsoffset and optionally add s64 biasdata.	arm_nnfunctions.h
arm_where_s16	WHERE operator: return coordinates of non-zero elements in condition (int16).	arm_nnfunctions.h
arm_where_s8	WHERE operator: return coordinates of non-zero elements in condition.	arm_nnfunctions.h

Activation Functions

ReLU, LeakyReLU, PReLU, Hard-Swish, Logistic, Tanh, and clamp. 27 functions.

Function	Summary	Module
arm_clamp_s16	S16 clamp function.	arm_nnfunctions.h
arm_clamp_s8	S8 clamp function.	arm_nnfunctions.h
arm_hard_swish_compat_s8	S8 Hard-Swish activation function (compatibility version).	arm_nnfunctions.h
arm_hard_swish_f16	Hard swish activation for float16 data.	Activation Functions
arm_hard_swish_f32	Hard swish activation for float32 data.	Activation Functions
arm_hard_swish_precise_s16	S16 Hard-Swish activation function (precise version).	arm_nnfunctions.h
arm_hard_swish_precise_s8	S8 Hard-Swish activation function (precise version).	arm_nnfunctions.h
arm_leaky_relu_s16	S16 Leaky ReLU activation function.	arm_nnfunctions.h
arm_leaky_relu_s8	S8 Leaky ReLU activation function.	arm_nnfunctions.h
arm_logistic_s16	Logistic activation function for s16.	arm_nnfunctions.h
arm_nn_activation_f16	Elementwise activation.	Activation Functions
arm_nn_activation_f32	Elementwise activation.	Activation Functions
arm_nn_activation_s16	s16 neural network activation function using direct table look-up	arm_nnfunctions.h
arm_prelu_f16	Parametric ReLU for float32 data.	Activation Functions
arm_prelu_f32	Parametric ReLU for float32 data.	Activation Functions
arm_prelu_s16	S16 PReLU activation function.	arm_nnfunctions.h
arm_prelu_s8	S8 PReLU activation function.	arm_nnfunctions.h
arm_prelu_scalar_s16	Scalar S16 PReLU activation function.	arm_nnfunctions.h
arm_prelu_scalar_s8	Scalar S8 PReLU activation function.	arm_nnfunctions.h
arm_relu_generic_s16	S16 ReLU activation function (generic version) This generic version allows to set custom lower and upper bounds to implement ReLU6, etc.	arm_nnfunctions.h
arm_relu_generic_s8	S8 ReLU activation function (generic version) This generic version allows to set custom lower and upper bounds to implement ReLU6, etc.	arm_nnfunctions.h
arm_relu_q15	Q15 RELU function.	arm_nnfunctions.h
arm_relu_q7	Q7 RELU function.	arm_nnfunctions.h
arm_relu_s16	S16 ReLU activation function lower and upper bounds are quantized representations of 0 and 32767.	arm_nnfunctions.h
arm_relu_s8	S8 ReLU activation function lower and upper bounds are quantized representations of 0 and 127.	arm_nnfunctions.h
arm_relu6_q7	Q7 RELU6 function.	arm_nnfunctions.h
arm_tanh_s16	Tanh activation function for s16.	arm_nnfunctions.h

Data Movement

Pad, reshape, transpose, concatenate, gather, resize, strided slice, tile, broadcast, scatter, mirror pad, and sequence/slice-update utilities. 77 functions.

Function	Summary	Module
arm_batch_to_space_nd_s16	Batch to Space ND function for s16 data type.	arm_nnfunctions.h
arm_batch_to_space_nd_s8	Batch to Space ND function for s8 data type.	arm_nnfunctions.h
arm_broadcast_to_s16	Broadcast an int16 tensor to a target shape.	arm_nnfunctions.h
arm_broadcast_to_s8	Broadcast an int8 tensor to a target shape.	arm_nnfunctions.h
arm_concatenation_f16	Concatenate float32 tensors of any rank along one axis.	NNSupport
arm_concatenation_f16_w	Concatenate tensors along the W axis.	NNSupport
arm_concatenation_f16_x	Concatenate tensors along the X axis.	NNSupport
arm_concatenation_f16_y	Concatenate tensors along the Y axis.	NNSupport
arm_concatenation_f16_z	Concatenate tensors along the Z axis.	NNSupport
arm_concatenation_f32	Concatenate float32 tensors of any rank along one axis.	NNSupport
arm_concatenation_f32_w	Concatenate tensors along the W axis.	NNSupport
arm_concatenation_f32_x	Concatenate tensors along the X axis.	NNSupport
arm_concatenation_f32_y	Concatenate tensors along the Y axis.	NNSupport
arm_concatenation_f32_z	Concatenate tensors along the Z axis.	NNSupport
arm_concatenation_s16	int16/uint16 concatenation function to be used for concatenating N-tensors along the target axis	arm_nnfunctions.h
arm_concatenation_s32	int32/uint32 concatenation function to be used for concatenating N-tensors along the target axis	arm_nnfunctions.h
arm_concatenation_s8	int8/uint8 concatenation function to be used for concatenating N-tensors along the target axis	arm_nnfunctions.h
arm_concatenation_s8_w	int8/uint8 concatenation function to be used for concatenating N-tensors along the W axis (Batch size) This function should be called for each input tensor to…	arm_nnfunctions.h
arm_concatenation_s8_x	int8/uint8 concatenation function to be used for concatenating N-tensors along the X axis This function should be called for each input tensor to concatenate.	arm_nnfunctions.h
arm_concatenation_s8_y	int8/uint8 concatenation function to be used for concatenating N-tensors along the Y axis This function should be called for each input tensor to concatenate.	arm_nnfunctions.h
arm_concatenation_s8_z	int8/uint8 concatenation function to be used for concatenating N-tensors along the Z axis This function should be called for each input tensor to concatenate.	arm_nnfunctions.h
arm_depth_to_space_s16	Depth to Space function for s16 data type.	arm_nnfunctions.h
arm_depth_to_space_s8	Depth to Space function for s8 data type.	arm_nnfunctions.h
arm_dynamic_update_slice_s16	Update a slice of an int16 operand tensor at runtime-determined indices.	arm_nnfunctions.h
arm_dynamic_update_slice_s8	Update a slice of an int8 operand tensor at runtime-determined indices.	arm_nnfunctions.h
arm_gather_f16	Gather contiguous slices along an axis.	Gather Functions:
arm_gather_f32	Gather contiguous slices along an axis.	Gather Functions:
arm_gather_nd_f16	Gather contiguous slices using coordinate tuples.	Gather Functions:
arm_gather_nd_f32	Gather contiguous slices using coordinate tuples.	Gather Functions:
arm_gather_nd_s16	Gathernd slices for int16 tensors.	arm_nnfunctions.h
arm_gather_nd_s8	Gathernd slices for int8 tensors.	arm_nnfunctions.h
arm_gather_s16	Gather elements along an axis for int16 tensors.	arm_nnfunctions.h
arm_gather_s8	Gather elements along an axis for int8 tensors.	arm_nnfunctions.h
arm_mirror_pad_s16	Mirror-pad an int16 tensor (REFLECT or SYMMETRIC mode).	arm_nnfunctions.h
arm_mirror_pad_s8	Mirror-pad an int8 tensor (REFLECT or SYMMETRIC mode).	arm_nnfunctions.h
arm_nn_fill_f16	Fill a float16 vector with one value; bit copy of value, NaN payload included.	Elementwise Functions
arm_nn_fill_f32	Fill a float32 vector with one value.	Elementwise Functions
arm_pack_f16	Stack float32 tensors of equal shape along a new axis (TFLite PACK).	NNSupport
arm_pack_f32	Stack float32 tensors of equal shape along a new axis (TFLite PACK).	NNSupport
arm_pad_f16	Pad a tensor with a constant value.	Pad Layer Functions:
arm_pad_f32	Pad a tensor with a constant value.	Pad Layer Functions:
arm_pad_s16	Expands the size of the input by adding constant values before and after the data, in all dimensions.	arm_nnfunctions.h
arm_pad_s8	Expands the size of the input by adding constant values before and after the data, in all dimensions.	arm_nnfunctions.h
arm_reshape_f16	Reshape by copying data without changing element order.	NNSupport
arm_reshape_f32	Reshape by copying data without changing element order.	NNSupport
arm_reshape_s8	Reshape a s8 vector into another with different shape.	arm_nnfunctions.h
arm_resize_nearest_neighbor_f16	Nearest-neighbor resize of a float32 NHWC tensor.	Reshape Functions
arm_resize_nearest_neighbor_f16_get_buffer_size	Scratch size in bytes for armresizenearestneighborf32() / armresizenearestneighborf16().	Reshape Functions
arm_resize_nearest_neighbor_f32	Nearest-neighbor resize of a float32 NHWC tensor.	Reshape Functions
arm_resize_nearest_neighbor_f32_get_buffer_size	Scratch size in bytes for armresizenearestneighborf32() / armresizenearestneighborf16().	Reshape Functions
arm_resize_nearest_neighbor_s16	Nearest neighbor resize function for s16 data.	arm_nnfunctions.h
arm_resize_nearest_neighbor_s8	Nearest neighbor resize function for s8 data.	arm_nnfunctions.h
arm_reverse_sequence_s16	Reverse variable-length sequences along a dimension for int16.	arm_nnfunctions.h
arm_reverse_sequence_s8	Reverse variable-length sequences along a dimension for int8.	arm_nnfunctions.h
arm_scatter_nd_s16	Scatter updates into a zero-initialized output tensor for int16.	arm_nnfunctions.h
arm_scatter_nd_s8	Scatter updates into a zero-initialized output tensor for int8.	arm_nnfunctions.h
arm_space_to_batch_nd_s16	Space to Batch ND function for s16 data type.	arm_nnfunctions.h
arm_space_to_batch_nd_s8	Space to Batch ND function for s8 data type.	arm_nnfunctions.h
arm_space_to_depth_s16	Space to Depth function for s16 data type.	arm_nnfunctions.h
arm_space_to_depth_s8	Space to Depth function for s8 data type.	arm_nnfunctions.h
arm_split_f16	Split a float32 tensor of any rank into several tensors along one axis.	Elementwise Functions
arm_split_f32	Split a float32 tensor of any rank into several tensors along one axis.	NNSupport
arm_split_s16	int16/uint16 split function to be used for splitting a tensor into multiple tensors along the target axis	arm_nnfunctions.h
arm_split_s8	int8/uint8 split function to be used for splitting a tensor into multiple tensors along the target axis	arm_nnfunctions.h
arm_strided_slice_f16	Strided slice for float32 data (pure copy, TensorFlow Lite compatible).	Elementwise Functions
arm_strided_slice_f32	Strided slice for float32 data (pure copy, TensorFlow Lite compatible).	Slicing Functions:
arm_strided_slice_s16	Strided slice function for int16 data.	arm_nnfunctions.h
arm_strided_slice_s32	Strided slice function for int32 data.	arm_nnfunctions.h
arm_strided_slice_s8	Strided slice function for int8 data.	arm_nnfunctions.h
arm_tile_s16	Tile an int16 tensor along each dimension.	arm_nnfunctions.h
arm_tile_s8	Tile an int8 tensor along each dimension.	arm_nnfunctions.h
arm_transpose_f16	Transpose a floating-point tensor.	NNSupport
arm_transpose_f32	Transpose a floating-point tensor.	NNSupport
arm_transpose_s16	Basic s16 transpose function.	arm_nnfunctions.h
arm_transpose_s8	Basic transpose function.	arm_nnfunctions.h
arm_unpack_f16	Unstack a float32 tensor along one axis into inputshape[axis] tensors (TFLite UNPACK).	NNSupport
arm_unpack_f32	Unstack a float32 tensor along one axis into inputshape[axis] tensors (TFLite UNPACK).	NNSupport

Classifier Tail

Classifier tail APIs plus dtype conversion and requantization utilities. 27 functions.

Function	Summary	Module
arm_avg_pool_f16	Average pooling.	Pooling Functions
arm_avg_pool_f32	Average pooling.	Pooling Functions
arm_avgpool_s16	s16 average pooling function.	arm_nnfunctions.h
arm_avgpool_s16_get_buffer_size	Get the required buffer size for S16 average pooling function.	arm_nnfunctions.h
arm_avgpool_s16_get_buffer_size_dsp	Get the required buffer size for S16 average pooling function for processors with DSP extension.	arm_nnfunctions.h
arm_avgpool_s16_get_buffer_size_mve	Get the required buffer size for S16 average pooling function for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_avgpool_s8	s8 average pooling function.	arm_nnfunctions.h
arm_avgpool_s8_get_buffer_size	Get the required buffer size for S8 average pooling function.	arm_nnfunctions.h
arm_avgpool_s8_get_buffer_size_dsp	Get the required buffer size for S8 average pooling function for processors with DSP extension.	arm_nnfunctions.h
arm_avgpool_s8_get_buffer_size_mve	Get the required buffer size for S8 average pooling function for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_dequantize_f16_f32	Widen a float16 vector to float32.	Quantization Functions:
arm_dequantize_s16_f32	Dequantize an int16t array back to floating-point format.	arm_nnfunctions.h
arm_dequantize_s8_f32	Dequantize an int8t array back to floating-point format.	arm_nnfunctions.h
arm_max_pool_f16	Max pooling.	Pooling Functions
arm_max_pool_f32	Max pooling.	Pooling Functions
arm_max_pool_s16	s16 max pooling function.	arm_nnfunctions.h
arm_max_pool_s8	s8 max pooling function.	arm_nnfunctions.h
arm_quantize_f32_s16	Quantize a floating-point array into int16t format.	arm_nnfunctions.h
arm_quantize_f32_s8	Quantize a floating-point array into int8t format.	arm_nnfunctions.h
arm_requantize_s16_s16	Requantize an int16t array to another int16t range with a different scale.	arm_nnfunctions.h
arm_requantize_s8_s8	Requantize an int8t array to another int8t range with a different scale.	arm_nnfunctions.h
arm_softmax_f16	Softmax using the float-native API signature.	Softmax Functions
arm_softmax_f32	Softmax using the float-native API signature.	Softmax Functions
arm_softmax_s16	S16 softmax function.	arm_nnfunctions.h
arm_softmax_s8	S8 softmax function.	arm_nnfunctions.h
arm_softmax_s8_s16	S8 to s16 softmax function.	arm_nnfunctions.h
arm_softmax_u8	U8 softmax function.	arm_nnfunctions.h

Sequence Functions

LSTM, GRU, and SVDF functions for temporal workloads. 31 functions.

Function	Summary	Module
arm_gru_unidirectional_f16	Unidirectional GRU layer for float16 input, output and state.	LSTM Layer Functions
arm_gru_unidirectional_f16_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armgruunidirectionalf16().	LSTM Layer Functions
arm_gru_unidirectional_f32	Unidirectional GRU layer for float32 input, output and state.	LSTM Layer Functions
arm_gru_unidirectional_f32_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armgruunidirectionalf32().	LSTM Layer Functions
arm_lstm_unidirectional_f16	Unidirectional LSTM inference.	LSTM Layer Functions
arm_lstm_unidirectional_f16_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armlstmunidirectionalf16().	LSTM Layer Functions
arm_lstm_unidirectional_f16_temp2_get_buffer_size	Get size of the temp2 scratch buffer required by armlstmunidirectionalf16().	LSTM Layer Functions
arm_lstm_unidirectional_f32	Unidirectional LSTM inference.	LSTM Layer Functions
arm_lstm_unidirectional_f32_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armlstmunidirectionalf32().	LSTM Layer Functions
arm_lstm_unidirectional_f32_temp2_get_buffer_size	Get size of the temp2 scratch buffer required by armlstmunidirectionalf32().	LSTM Layer Functions
arm_lstm_unidirectional_s16	LSTM unidirectional function with 16 bit input and output and 16 bit gate output, 64 bit bias.	arm_nnfunctions.h
arm_lstm_unidirectional_s16_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armlstmunidirectionals16().	arm_nnfunctions.h
arm_lstm_unidirectional_s16_temp2_get_buffer_size	Get size of the temp2 scratch buffer required by armlstmunidirectionals16().	arm_nnfunctions.h
arm_lstm_unidirectional_s8	LSTM unidirectional function with 8 bit input and output and 16 bit gate output, 32 bit bias.	arm_nnfunctions.h
arm_lstm_unidirectional_s8_temp1_get_buffer_size	Get size of the temp1 scratch buffer required by armlstmunidirectionals8().	arm_nnfunctions.h
arm_lstm_unidirectional_s8_temp2_get_buffer_size	Get size of the temp2 scratch buffer required by armlstmunidirectionals8().	arm_nnfunctions.h
arm_svdf_f16	Stateful singular value decomposition filter, float16 variant.	SVDF Functions
arm_svdf_f16_input_ctx_get_buffer_size	Get size of the inputctx staging buffer required by armsvdff16().	SVDF Functions
arm_svdf_f16_output_ctx_get_buffer_size	Get size of the outputctx staging buffer required by armsvdff16().	SVDF Functions
arm_svdf_f32	Stateful singular value decomposition filter.	SVDF Functions
arm_svdf_f32_input_ctx_get_buffer_size	Get size of the inputctx staging buffer required by armsvdff32().	SVDF Functions
arm_svdf_f32_output_ctx_get_buffer_size	Get size of the outputctx staging buffer required by armsvdff32().	SVDF Functions
arm_svdf_s8	s8 SVDF function with 8 bit state tensor and 8 bit time weights	arm_nnfunctions.h
arm_svdf_s8_get_buffer_size	Get size of the kernel-sum buffer required by armsvdfs8().	arm_nnfunctions.h
arm_svdf_s8_get_buffer_size_dsp	Get size of the kernel-sum buffer required by armsvdfs8() for processors with DSP extension.	arm_nnfunctions.h
arm_svdf_s8_get_buffer_size_mve	Get size of the kernel-sum buffer required by armsvdfs8() for Arm(R) Helium Architecture case.	arm_nnfunctions.h
arm_svdf_s8_input_ctx_get_buffer_size	Get size of the inputctx staging buffer required by armsvdfs8().	arm_nnfunctions.h
arm_svdf_s8_output_ctx_get_buffer_size	Get size of the outputctx staging buffer required by armsvdfs8().	arm_nnfunctions.h
arm_svdf_state_s16_s8	s8 SVDF function with 16 bit state tensor and 16 bit time weights	arm_nnfunctions.h
arm_svdf_state_s16_s8_input_ctx_get_buffer_size	Get size of the inputctx staging buffer required by armsvdfstates16s8().	arm_nnfunctions.h
arm_svdf_state_s16_s8_output_ctx_get_buffer_size	Get size of the outputctx staging buffer required by armsvdfstates16s8().	arm_nnfunctions.h

Types and support

Structs, enums, macros and typedefs stay on the module pages rather than being repeated here. Open heliaCORE for the full tree.