forked from neurosim/MLP_NeuroSim_V3.0
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmain.cpp
171 lines (153 loc) · 8.86 KB
/
main.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
/*******************************************************************************
* Copyright (c) 2015-2017
* School of Electrical, Computer and Energy Engineering, Arizona State University
* PI: Prof. Shimeng Yu
* All rights reserved.
*
* This source code is part of NeuroSim - a device-circuit-algorithm framework to benchmark
* neuro-inspired architectures with synaptic devices(e.g., SRAM and emerging non-volatile memory).
* Copyright of the model is maintained by the developers, and the model is distributed under
* the terms of the Creative Commons Attribution-NonCommercial 4.0 International Public License
* http://creativecommons.org/licenses/by-nc/4.0/legalcode.
* The source code is free and you can redistribute and/or modify it
* by providing that the following conditions are met:
*
* 1) Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
*
* 2) Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* Developer list:
* Pai-Yu Chen Email: pchen72 at asu dot edu
*
* Xiaochen Peng Email: xpeng15 at asu dot edu
********************************************************************************/
#include <cstdio>
#include <iostream>
#include <fstream>
#include <sstream>
#include <string>
#include <stdlib.h>
#include <random>
#include <vector>
#include "Cell.h"
#include "Array.h"
#include "formula.h"
#include "NeuroSim.h"
#include "Param.h"
#include "IO.h"
#include "Train.h"
#include "Test.h"
#include "Mapping.h"
#include "Definition.h"
#include "omp.h"
using namespace std;
int main() {
gen.seed(0);
/* Load in MNIST data */
ReadTrainingDataFromFile("patch60000_train.txt", "label60000_train.txt");
ReadTestingDataFromFile("patch10000_test.txt", "label10000_test.txt");
/* Initialization of synaptic array from input to hidden layer */
//arrayIH->Initialization<IdealDevice>();
arrayIH->Initialization<RealDevice>();
//arrayIH->Initialization<MeasuredDevice>();
//arrayIH->Initialization<SRAM>(param->numWeightBit);
//arrayIH->Initialization<DigitalNVM>(param->numWeightBit,true);
//arrayIH->Initialization<HybridCell>(); // the 3T1C+2PCM cell
//arrayIH->Initialization<_2T1F>();
/* Initialization of synaptic array from hidden to output layer */
//arrayHO->Initialization<IdealDevice>();
arrayHO->Initialization<RealDevice>();
//arrayHO->Initialization<MeasuredDevice>();
//arrayHO->Initialization<SRAM>(param->numWeightBit);
//arrayHO->Initialization<DigitalNVM>(param->numWeightBit,true);
//arrayHO->Initialization<HybridCell>(); // the 3T1C+2PCM cell
//arrayHO->Initialization<_2T1F>();
omp_set_num_threads(16);
/* Initialization of NeuroSim synaptic cores */
param->relaxArrayCellWidth = 0;
NeuroSimSubArrayInitialize(subArrayIH, arrayIH, inputParameterIH, techIH, cellIH);
param->relaxArrayCellWidth = 1;
NeuroSimSubArrayInitialize(subArrayHO, arrayHO, inputParameterHO, techHO, cellHO);
/* Calculate synaptic core area */
NeuroSimSubArrayArea(subArrayIH);
NeuroSimSubArrayArea(subArrayHO);
/* Calculate synaptic core standby leakage power */
NeuroSimSubArrayLeakagePower(subArrayIH);
NeuroSimSubArrayLeakagePower(subArrayHO);
/* Initialize the neuron peripheries */
NeuroSimNeuronInitialize(subArrayIH, inputParameterIH, techIH, cellIH, adderIH, muxIH, muxDecoderIH, dffIH, subtractorIH);
NeuroSimNeuronInitialize(subArrayHO, inputParameterHO, techHO, cellHO, adderHO, muxHO, muxDecoderHO, dffHO, subtractorHO);
/* Calculate the area and standby leakage power of neuron peripheries below subArrayIH */
double heightNeuronIH, widthNeuronIH;
NeuroSimNeuronArea(subArrayIH, adderIH, muxIH, muxDecoderIH, dffIH, subtractorIH, &heightNeuronIH, &widthNeuronIH);
double leakageNeuronIH = NeuroSimNeuronLeakagePower(subArrayIH, adderIH, muxIH, muxDecoderIH, dffIH, subtractorIH);
/* Calculate the area and standby leakage power of neuron peripheries below subArrayHO */
double heightNeuronHO, widthNeuronHO;
NeuroSimNeuronArea(subArrayHO, adderHO, muxHO, muxDecoderHO, dffHO, subtractorHO, &heightNeuronHO, &widthNeuronHO);
double leakageNeuronHO = NeuroSimNeuronLeakagePower(subArrayHO, adderHO, muxHO, muxDecoderHO, dffHO, subtractorHO);
/* Print the area of synaptic core and neuron peripheries */
double totalSubArrayArea = subArrayIH->usedArea + subArrayHO->usedArea;
double totalNeuronAreaIH = adderIH.area + muxIH.area + muxDecoderIH.area + dffIH.area + subtractorIH.area;
double totalNeuronAreaHO = adderHO.area + muxHO.area + muxDecoderHO.area + dffHO.area + subtractorHO.area;
printf("Total SubArray (synaptic core) area=%.4e m^2\n", totalSubArrayArea);
printf("Total Neuron (neuron peripheries) area=%.4e m^2\n", totalNeuronAreaIH + totalNeuronAreaHO);
printf("Total area=%.4e m^2\n", totalSubArrayArea + totalNeuronAreaIH + totalNeuronAreaHO);
/* Print the standby leakage power of synaptic core and neuron peripheries */
printf("Leakage power of subArrayIH is : %.4e W\n", subArrayIH->leakage);
printf("Leakage power of subArrayHO is : %.4e W\n", subArrayHO->leakage);
printf("Leakage power of NeuronIH is : %.4e W\n", leakageNeuronIH);
printf("Leakage power of NeuronHO is : %.4e W\n", leakageNeuronHO);
printf("Total leakage power of subArray is : %.4e W\n", subArrayIH->leakage + subArrayHO->leakage);
printf("Total leakage power of Neuron is : %.4e W\n", leakageNeuronIH + leakageNeuronHO);
/* Initialize weights and map weights to conductances for hardware implementation */
WeightInitialize();
if (param->useHardwareInTraining)
WeightToConductance();
srand(0); // Pseudorandom number seed
ofstream mywriteoutfile;
mywriteoutfile.open("output.csv");
for (int i=1; i<=param->totalNumEpochs/param->interNumEpochs; i++){
Train(param->numTrainImagesPerEpoch, param->interNumEpochs,param->optimization_type);
if (!param->useHardwareInTraining && param->useHardwareInTestingFF) { WeightToConductance(); }
Validate();
if (HybridCell *temp = dynamic_cast<HybridCell*>(arrayIH->cell[0][0]))
WeightTransfer();
else if(_2T1F *temp = dynamic_cast<_2T1F*>(arrayIH->cell[0][0]))
WeightTransfer_2T1F();
mywriteoutfile << i*param->interNumEpochs << ", " << (double)correct/param->numMnistTestImages*100 << endl;
printf("Accuracy at %d epochs is : %.2f%\n", i*param->interNumEpochs, (double)correct/param->numMnistTestImages*100);
/* Here the performance metrics of subArray also includes that of neuron peripheries (see Train.cpp and Test.cpp) */
printf("\tRead latency=%.4e s\n", subArrayIH->readLatency + subArrayHO->readLatency);
printf("\tWrite latency=%.4e s\n", subArrayIH->writeLatency + subArrayHO->writeLatency);
printf("\tRead energy=%.4e J\n", arrayIH->readEnergy + subArrayIH->readDynamicEnergy + arrayHO->readEnergy + subArrayHO->readDynamicEnergy);
printf("\tWrite energy=%.4e J\n", arrayIH->writeEnergy + subArrayIH->writeDynamicEnergy + arrayHO->writeEnergy + subArrayHO->writeDynamicEnergy);
if(HybridCell* temp = dynamic_cast<HybridCell*>(arrayIH->cell[0][0])){
printf("\tTransfer latency=%.4e s\n", subArrayIH->transferLatency + subArrayHO->transferLatency);
printf("\tTransfer latency=%.4e s\n", subArrayIH->transferLatency);
printf("\tTransfer energy=%.4e J\n", arrayIH->transferEnergy + subArrayIH->transferDynamicEnergy + arrayHO->transferEnergy + subArrayHO->transferDynamicEnergy);
}
else if(_2T1F* temp = dynamic_cast<_2T1F*>(arrayIH->cell[0][0])){
printf("\tTransfer latency=%.4e s\n", subArrayIH->transferLatency);
printf("\tTransfer energy=%.4e J\n", arrayIH->transferEnergy + subArrayIH->transferDynamicEnergy + arrayHO->transferEnergy + subArrayHO->transferDynamicEnergy);
}
// printf("\tThe total weight update = %.4e\n", totalWeightUpdate);
// printf("\tThe total pulse number = %.4e\n", totalNumPulse);
}
// print the summary:
printf("\n");
return 0;
}