Add README section on performance counters

[nazavode]

This PR adds a full description of all performance counters reported by the post-processing script. All descriptions have been discussed with hw designers to avoid discrepancies with the micro-architecture.

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt	min_llvm_mlir
conv2d_d1_s1_3x3 1x1x8x8xf64	1559	1489		1503	1486	1489
ddot 128xf64	929	937	961	570	550	929
dense 8x8xf64	3239	3238		2730	2709	3238
dsum 8x16xf32	1202	794	3543	209	187	794
fill 16x16xf64	370	347	2665	294	299	347
matmul 8x8xf64	2562	2560	13602	2333	2304	2560
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	1435	1201	13548	1115	1094	1201
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	1948	1948	13548	1997	1976	1948
relu 16x16xf64	1081	1074	4974	322	334	1074
softmax 16xf64	32613					32613

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt	min_llvm_mlir
conv2d_d1_s1_3x3 1x1x8x8xf64	1559	1489		1503	1486	1489
ddot 128xf64	929	937	961	570	550	929
dense 8x8xf64	3239	3238		2730	2709	3238
dsum 8x16xf32	1202	794	3543	209	187	794
fill 16x16xf64	370	347	2665	294	299	347
matmul 8x8xf64	2562	2560	13602	2333	2304	2560
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	1435	1201	13548	1115	1094	1201
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	1948	1948	13548	1997	1976	1948
relu 16x16xf64	1081	1074	4974	322	334	1074
softmax 16xf64	32613					32613

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt	min_llvm_mlir
conv2d_d1_s1_3x3 1x1x8x8xf64	1559	1489		1503	1486	1489
ddot 128xf64	929	937	961	570	550	929
dense 8x8xf64	3239	3238		2730	2709	3238
dsum 8x16xf32	1202	794	2019	209	187	794
fill 16x16xf64	370	347	2665	294	299	347
matmul 8x8xf64	2515	2823		2336	2323	2515
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	1435	1201		1115	1094	1201
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	1948	1948		1997	1976	1948
relu 16x16xf64	1081	1074	3460	322	334	1074

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt
conv2d_d1_s1_3x3 1x1x8x8xf64	0.23	0.21		0.18	0.18
ddot 128xf64	0.14	0.14	0.13	0.00	0.00
dense 8x8xf64	0.20	0.20		0.22	0.23
dsum 8x16xf32	0.11	0.16	0.06	0.00	0.69
fill 16x16xf64	0.00	0.00	0.00	0.02	0.86
matmul 8x8xf64	0.20	0.18		0.25	0.25
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	0.31	0.37		0.35	0.36
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	0.22	0.22		0.19	0.20
relu 16x16xf64	0.24	0.24	0.07	0.01	0.01

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt	min_llvm_mlir
conv2d_d1_s1_3x3 1x1x8x8xf64	1559	1489		1503	1486	1489
ddot 128xf64	929	937	961	570	550	929
dense 8x8xf64	3239	3238		2730	2709	3238
dsum 8x16xf32	1202	794	2019	209	187	794
fill 16x16xf64	370	347	2665	294	299	347
matmul 8x8xf64	2515	2823		2336	2323	2515
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	1435	1201		1115	1094	1201
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	1948	1948		1997	1976	1948
relu 16x16xf64	1081	1074	3460	322	334	1074

[github-actions]

kernels	baseline	linalg	scf_xdsl	snitch_stream	snrt
conv2d_d1_s1_3x3 1x1x8x8xf64	0.23	0.21		0.18	0.18
ddot 128xf64	0.14	0.14	0.13	0.00	0.00
dense 8x8xf64	0.20	0.20		0.22	0.23
dsum 8x16xf32	0.11	0.16	0.06	0.00	0.69
fill 16x16xf64	0.00	0.00	0.00	0.02	0.86
matmul 8x8xf64	0.20	0.18		0.25	0.25
pooling_nchw_max_d1_s2_3x3 1x1x16x16xf64	0.31	0.37		0.35	0.36
pooling_nchw_sum_d1_s2_3x3 1x1x16x16xf64	0.22	0.22		0.19	0.20
relu 16x16xf64	0.24	0.24	0.07	0.01	0.01