Navigation :

GPU Kernel Information Aggregated by Layer

Search:

layer_index	layer_name	layer_type	layer_duration (us)	layer_gpu_duration (us)	layer_cpu_duration (us)	layer_flops	layer_dram_read_bytes	layer_dram_write_bytes	layer_achieved_occupancy (%)	layer_arithmetic_intensity (flops/byte)	layer_arithmetic_throughput (GFlops)	layer_memory_bound

layer_index	layer_name	layer_type	layer_duration (us)	layer_gpu_duration (us)	layer_cpu_duration (us)	layer_memory_bound
1	vgg4_relu0_fwd	Activation	196.67	186.00	10.67	true
2	vgg4_conv1_fwd	Convolution	286584.33	879.67	285704.67	true
3	vgg4_relu1_fwd	Activation	206.33	186.00	20.33	true
4	vgg4_pool0_fwd	Pooling	7511.33	131.33	7380.00	true
5	vgg4_conv2_fwd	Convolution	122470.33	441.67	122028.67	true
6	vgg4_relu2_fwd	Activation	106.00	94.67	11.33	true
7	vgg4_conv3_fwd	Convolution	245914.67	669.67	245245.00	true
8	vgg4_relu3_fwd	Activation	114.33	94.00	20.33	true
9	vgg4_pool1_fwd	Pooling	4259.67	74.67	4185.00	true
10	vgg4_conv4_fwd	Convolution	113114.67	411.67	112703.00	true
11	vgg4_relu4_fwd	Activation	64.00	47.00	17.00	true
12	vgg4_conv5_fwd	Convolution	228811.33	711.67	228099.67	true
13	vgg4_relu5_fwd	Activation	65.67	46.00	19.67	true
14	vgg4_conv6_fwd	Convolution	228606.33	711.00	227895.33	true
15	vgg4_relu6_fwd	Activation	66.67	46.33	20.33	true
16	vgg4_pool2_fwd	Pooling	2574.67	43.00	2531.67	true
17	vgg4_conv7_fwd	Convolution	112785.67	518.00	112267.67	true
18	vgg4_relu7_fwd	Activation	42.33	9.00	33.33	true
19	vgg4_conv8_fwd	Convolution	223792.00	950.67	222841.33	true
20	vgg4_relu8_fwd	Activation	42.33	9.00	33.33	true
21	vgg4_conv9_fwd	Convolution	226563.67	956.00	225607.67	true
22	vgg4_relu9_fwd	Activation	42.00	9.00	33.00	true
23	vgg4_pool3_fwd	Pooling	1429.33	16.00	1413.33	true
24	vgg4_conv10_fwd	Convolution	56645.33	428.00	56217.33	true
25	vgg4_relu10_fwd	Activation	23.33	5.00	18.33	true
26	vgg4_conv11_fwd	Convolution	57040.00	422.33	56617.67	true
27	vgg4_relu11_fwd	Activation	22.67	5.00	17.67	true
28	vgg4_conv12_fwd	Convolution	57081.67	421.33	56660.33	true
29	vgg4_relu12_fwd	Activation	22.33	5.00	17.33	true
30	vgg4_pool4_fwd	Pooling	419.67	8.67	411.00	true
31	vgg4_dense0_fwd	FullyConnected	225574.00	3036.67	222537.33	true
32	vgg4_dense0_relu_fwd	Activation	26.33	3.00	23.33	true
33	vgg4_dropout0_fwd	Dropout	17.33	3.00	14.33	true
34	vgg4_dense1_fwd	FullyConnected	36664.67	495.00	36169.67	true
35	vgg4_dense1_relu_fwd	Activation	25.00	3.00	22.00	true
36	vgg4_dropout1_fwd	Dropout	16.67	2.00	14.67	true
37	vgg4_dense2_fwd	FullyConnected	9180.00	125.00	9055.00	true

Showing 1 to 37 of 37 entries

Download as CSV