-
Notifications
You must be signed in to change notification settings - Fork 0
/
mlsys.html
147 lines (133 loc) · 8.73 KB
/
mlsys.html
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
<!DOCTYPE HTML>
<!--
Phantom by HTML5 UP
html5up.net | @ajlkn
Free for personal and commercial use under the CCA 3.0 license (html5up.net/license)
-->
<html>
<head>
<title>Horizon Lab @ URCS</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
<link rel="stylesheet" href="assets/css/main.css" />
<link rel="icon" href="favicon.ico" />
<noscript><link rel="stylesheet" href="assets/css/noscript.css" /></noscript>
</head>
<body class="is-preload">
<!-- Wrapper -->
<div id="wrapper">
<!-- Header -->
<header id="header">
<div class="inner">
<!-- Logo -->
<a href="index.html" class="logo">
<span class="symbol"><img src="images/urseal.svg" alt="" /></span><span class="head">Horizon Lab @ URCS</span>
</a>
<!-- Nav -->
<nav>
<ul>
<li><a href="#menu">Menu</a></li>
</ul>
</nav>
<!-- Menu -->
<nav id="menu">
<h2>Menu</h2>
<ul id="list">
<li><a href="index.html">Home</a></li>
<li><a href="people.html">People</a></li>
<li><a href="index.html#proj">Projects</a></li>
<li><a href="pubs.html">Publications</a></li>
<li><a href="artifacts.html">Artifacts</a></li>
<li><a href="teach.html">Courses & Outreach</a></li>
</ul>
</nav>
</div>
</header>
<!-- Main -->
<div id="main">
<div class="inner">
<header>
<h1 class="title">Systems and Architecture for ML</h1>
<div class="midbmargin">
</div>
<h2 class="courses">Recent Publications</h2>
<h3><span class="em">Systems and Hardware Support</span></h3>
<div class="smallbmargin">
[<span class="confname">ASPLOS 2024</span>] <a class="paperlink" href="pubs/asplos24-amanda.pdf">Amanda: Unified Instrumentation Framework for Deep Neural Networks</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">ISCA 2023</span>] <a class="paperlink" href="pubs/isca23-olive.pdf">OliVe: Accelerating Large Language Models via Hardware-friendly Outlier-Victim Pair Quantization</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">HPCA 2022</span>] <a class="paperlink" href="https://arxiv.org/pdf/2107.07983.pdf">S2TA: Exploiting Structured Sparsity for Energy-Efficient Mobile CNN Acceleration</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">MICRO 2022</span>] <a class="paperlink" href="https://arxiv.org/pdf/2208.14286.pdf">ANT: Exploiting Adaptive Numerical Data Type for Low-bit Deep Neural Network Quantization</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">ICLR 2022</span>] <a class="paperlink" href="https://arxiv.org/pdf/2202.07471.pdf">SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">IISWC 2021</span>] <a class="paperlink" href="https://arxiv.org/pdf/2110.03901.pdf">Characterizing and Demystifying the Implicit Convolution Algorithm on Commercial Matrix-Multiplication Accelerators</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">ISPASS 2020</span>] <a class="paperlink" href="pubs/ispass20.pdf">A Systematic Methodology for Characterizing Scalability of DNN Accelerators using SCALE-Sim</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">DAC 2020</span>] <a class="paperlink" href="https://arxiv.org/pdf/2002.08326.pdf">Balancing Efficiency and Flexibility for DNN Acceleration via Temporal GPU-Systolic Array Integration</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">SC 2020</span>] <a class="paperlink" href="pubs/sc20.odf">Accelerating Sparse DNN Models Without Hardware-Support via Tile-wise Sparsity</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">IPDPS 2018</span>] <a class="paperlink" href="pubs/ipdps18.pdf">BitFlow: Exploiting Vector Parallelism for Binary Neural Networks on CPU</a><br>
</div>
<h3><span class="em">Adversarial Robustness</span></h3>
<div class="smallbmargin">
[<span class="confname">MICRO 2020</span>] <a class="paperlink" href="pubs/micro20-ptolemy.pdf">Ptolemy: Architecture Support for Robust Deep Learning</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">CVPR 2019</span>] <a class="paperlink" href="https://openaccess.thecvf.com/content_CVPR_2019/papers/Qiu_Adversarial_Defense_Through_Network_Profiling_Based_Path_Extraction_CVPR_2019_paper.pdf">Adversarial Defense Through Network Profiling Based Path Extraction</a><br>
</div>
<h3><span class="em">Resource-Guaranteeing Deep Learning</span></h3>
<div class="smallbmargin">
[<span class="confname">CVPR 2020</span>] <a class="paperlink" href="https://openaccess.thecvf.com/content_CVPR_2020/papers/Yang_Automatic_Neural_Network_Compression_by_Sparsity-Quantization_Joint_Learning_A_Constrained_CVPR_2020_paper.pdf">Automatic Neural Network Compression by Sparsity-Quantization Joint Learning: A Constrained Optimization-based Approach</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">CVPR 2019</span>] <a class="paperlink" href="http://openaccess.thecvf.com/content_CVPR_2019/papers/Yang_ECC_Platform-Independent_Energy-Constrained_Deep_Neural_Network_Compression_via_a_Bilinear_CVPR_2019_paper.pdf">ECC: Energy-Constrained Deep Neural Network Compression via a Bilinear Regression Model</a><br>
</div>
<div class="smallbmargin">
[<span class="confname">ICLR 2019</span>] <a class="paperlink" href="https://openreview.net/forum?id=BylBr3C9K7">Energy-Constrained Compression for Deep Neural Networks via Weighted Sparse Projection and Layer Input Masking</a><br>
</div>
</div>
</div>
<!-- Footer -->
<footer id="footer">
<div class="inner">
<!--<section>
<h2>Follow</h2>
<ul class="icons">
<li><a href="#" class="icon style2 fa-twitter"><span class="label">Twitter</span></a></li>
<li><a href="#" class="icon style2 fa-facebook"><span class="label">Facebook</span></a></li>
<li><a href="#" class="icon style2 fa-instagram"><span class="label">Instagram</span></a></li>
<li><a href="#" class="icon style2 fa-dribbble"><span class="label">Dribbble</span></a></li>
<li><a href="#" class="icon style2 fa-github"><span class="label">GitHub</span></a></li>
<li><a href="#" class="icon style2 fa-500px"><span class="label">500px</span></a></li>
<li><a href="#" class="icon style2 fa-phone"><span class="label">Phone</span></a></li>
<li><a href="#" class="icon style2 fa-envelope-o"><span class="label">Email</span></a></li>
</ul>
</section>-->
<ul class="copyright">
<li>© Horizon Lab. All rights reserved</li><li>Design: <a href="http://html5up.net">HTML5 UP</a></li>
</ul>
</div>
</footer>
</div>
<!-- Scripts -->
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/browser.min.js"></script>
<script src="assets/js/breakpoints.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>