-
Notifications
You must be signed in to change notification settings - Fork 1
/
index.html
246 lines (225 loc) · 16.3 KB
/
index.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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
<!DOCTYPE html>
<!-- Modeled off of Bootstrap's fine starter template. Thanks! -->
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>NOPE 2022</title>
<!-- Bootstrap -->
<!-- Latest compiled and minified CSS -->
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.min.css">
<!-- HTML5 shim and Respond.js for IE8 support of HTML5 elements and media queries -->
<!-- WARNING: Respond.js doesn't work if you view the page via file:// -->
<!--[if lt IE 9]>
<script src="https://oss.maxcdn.com/html5shiv/3.7.2/html5shiv.min.js"></script> <script src="https://oss.maxcdn.com/respond/1.4.2/respond.min.js"></script>
<![endif]-->
<link href="https://fonts.googleapis.com/css?family=Questrial|PT+Sans" rel="stylesheet" type="text/css">
<link rel="stylesheet" href="assets/nope.css">
</head>
<body data-spy="scroll" data-target=".navbar" data-offset="50">
<!-- Scrollspy? Scrollspy! http://www.w3schools.com/bootstrap/bootstrap_scrollspy.asp -->
<!-- Fixed Navbar -->
<nav class="navbar navbar-default navbar-fixed-top">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle collapsed" data-toggle="collapse" data-target="#mynav">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="#">NOPE 2022</a>
</div>
<div class="collapse navbar-collapse" id="mynav">
<ul class="nav navbar-nav">
<!-- note: scrollspy doesn't play nicely with other options in <a>
tags. So the that's why we have an extra div with the toggle
features (which are so that the drop-down menu auto-hides
itself on mobile). -->
<!--<li><a href="#program"><div data-toggle="collapse" data-target="#mynav">Program</div></a></li>--!>
<li><a href="#about"><div data-toggle="collapse" data-target="#mynav">About</div></a></li>
<li><a href="#cfp"><div data-toggle="collapse" data-target="#mynav">Call for Papers</div></a></li>
<li><a href="#committee"><div data-toggle="collapse" data-target="#mynav">Committee</div></a></li>
<li><a href="#submission"><div data-toggle="collapse" data-target="#mynav">Submit a Paper</div></a></li>
</ul>
</div>
</div>
</nav>
<div id="page" class="container">
<div class="container page-header">
<h1>2022 Workshop on<br/>
<span class="spacing"> </span>Negative results,<br/>
<span class="spacing">  </span>Opportunities,<br/>
<span class="spacing">   </span>Perspectives, and<br/>
<span class="spacing">    </span>Experiences</h1>
<h3>In conjunction with <a href="https://asplos-conference.org/">ASPLOS-27</a>
<h3>February 28, 2022</h3>
<h3>Virtual Event</h3>
<h2>Please register for ASPLOS for Zoom access to this event!</h2>
</div>
<div id="program" class="container section">
<h3>Program</h3>
Schedule :
<table class="table table-striped">
<thead>
<tr>
<th style="width: 20%"></th>
<th style="width: 50%"></th>
<th style="width: 30%"></th>
</tr>
</thead>
<tbody>
<tr>
<td>9:00 - 9:15am ET</td>
<td>Opening remarks</td>
<td><a href="http://vlsiarch.eecs.harvard.edu/">Organizers<br/>Harvard</a></td>
</tr>
<tr>
<td>9:15 - 9:45am ET</td>
<td>Failure as a learning opportunity for growth</td>
<td><a href="https://sites.google.com/site/gillespokam/home">Gilles Pokam<br/>Intel</a></td>
</tr>
<td>9:45 - 10:15am ET</td>
<td>Can Chips Prove Their Own Correctness?</td>
<td><a href="https://engineering.nyu.edu/faculty/siddharth-garg">Siddharth Garg<br/>New York University</a></td>
</tr>
<td>10:15am - 10:30am ET</td>
<td>Break
</td>
<td></td>
</tr>
<tr>
<td>10:30am - 10:50am ET</td>
<td>CRAP: Collecting Resources Across different Processing levels<br/>
<a role="button" data-toggle="collapse" href="#abstract-thomas" aria-expanded="false" aria-controls="abstract-thomas">[Abstract]</a>
</td>
<td><a href="https://samueltphd.github.io/">Samuel Thomas<br/>Brown University</a></td>
</tr>
<tr class="hidden"><td></td></tr><!--this counter-balances the collapsed table row to keep the overall table striping correct-->
<tr class="collapse" id="abstract-thomas"><td colspan="3">
<div class="abstract">
<p>Near-Memory Processing (NMP) is an emerging technology that can help accelerate pointer-chasing applications. However, it is necessary to invoke further evaluation of NMP on workloads that are more complex than data structures. We identify garbage collection as a more complex application for NMP. In this paper, we demonstrate that we can get 2x performance improvement in short-lived processes and a 2.3x reduction in L2 traffic by integrating NMP into the built-in Java garbage collector. However, we will discuss architectural reasons why these results do not extend to long-lived processes. We will discuss why these results should still be encouraging motivation for NMP-awareness in the JDK. <a href="papers/2022/thomas.pdf">View Paper</a></p>
</div>
</td></tr>
<tr>
<td>10:50am - 11:10am ET</td>
<td>Too Noisy To Extract : Pitfalls of Model Extraction Attacks<br/>
<a role="button" data-toggle="collapse" href="#abstract-alrafi" aria-expanded="false" aria-controls="abstract-alrafi">[Abstract]</a>
</td>
<td>Mujahid Al Rafi<br/>UC Merced</td>
</tr>
<tr class="hidden"><td></td></tr><!--this counter-balances the collapsed table row to keep the overall table striping correct-->
<tr class="collapse" id="abstract-alrafi"><td colspan="3">
<div class="abstract">
<p>Deep learning solutions become one of the most important intellectual properties of the solution providers. A few studies raised the alarm about the leakage of important model parameters through model extraction attacks. This paper discusses the practicality of such attacks by exploring the baseline assumptions of the existing attack scenarios. <a href="papers/2022/alrafi.pdf">View Paper</a></p>
</div>
</td></tr>
<tr>
<td>11:10am - 11:30am ET</td>
<td>Homomorphically Encrypted Computation using Stochastic Encodings<br/>
<a role="button" data-toggle="collapse" href="#abstract-hsiao" aria-expanded="false" aria-controls="abstract-hsiao">[Abstract]</a>
</td>
<td>Julie Hsiao<br/>Facebook Reality Labs - Research</td>
</tr>
<tr class="hidden"><td></td></tr><!--this counter-balances the collapsed table row to keep the overall table striping correct-->
<tr class="collapse" id="abstract-hsiao"><td colspan="3">
<div class="abstract">
<p>Homomorphic encryption (HE) is a privacy-preserving technique that enables computation directly over ciphertext. Unfortunately, a key challenge for HE is that implementations can be impractically slow and have limits on computation that can be efficiently implemented. For instance, in Boolean constructions of HE like TFHE, arithmetic operations need to be decomposed into constituent elementary logic gates to implement so performance depends on logical circuit depth. For even heavily quantized fixed-point arithmetic operations, these HE circuit implementations can be slow. This paper explores the merit of using stochastic computing (SC) encodings to reduce the logical depth required for HE computation to enable more efficient implementations. Contrary to computation in the plaintext space where many efficient hardware implementations are available, HE provides support for only a limited number of primitive operators and their performance may not directly correlate to their plaintext performance. Our results show that by layering SC encodings on top of TFHE, we observe similar challenges and limitations that SC faces in the plaintext space. Additional breakthroughs would require more support from the HE libraries to make SC with HE a viable solution. <a href="papers/2022/hsiao.pdf">View Paper</a></p>
</div>
</td></tr>
<tr>
<td>11:30am - Noon ET</td>
<td>Industry and its Academic Research -- How does it work?</td>
<td><a href="https://www.intel.com/content/www/us/en/research/researchers/mandy-pant.html">Mondira Pant<br/>Intel</a></td>
</tr>
<tr>
<td>Noon ET</td>
<td>The End. Discussion, Feedback. </td>
<td></td>
</tr>
</tbody>
</table>
</div>
<div id="about" class="container section">
<h3>Bringing historical perspectives to NOPE</h3>
<p>
At MICRO 2019, Lynn Conway gave a keynote describing her personal story and the "techno-social" factors which led to her contributions being undervalued and unseen. Inspired by her perspective and call to action, we aim to rebrand the aims and goals NOPE to be more inclusive of historical perspectives and descriptions of how technical concepts came to be commonplace.
</p>
<p>
NOPE will remain a place for open, honest port-mortems of research projects which ran into unexpected limitations and resulted in lessons learned. In addition, it will offer a venue to discuss contributions that have been underappreciated and misconstrued over time. In this way, the goals of NOPE are to reflect on negative outcomes, shed light on the origin of ideas, and offer a venue to revise our understanding and uncover opportunities to move forward by reflecting on mistakes that can be made throughout the research process.
</p>
<h3>Why do we need to talk about negative results?</h3>
<p>
Not all research projects end up with positive results. Sometimes ideas that sound enticing at first run into unexpected complexity, high overheads, or turn out simply infeasible. Such projects often end up in a proverbial researcher's drawer, and the community as a whole is not aware of dead-end or hard-to-advance research directions. <strong>NOPE is a venue that encourages publishing such results in all their "badness".</strong>
</p>
<h3>What is a "good failure"?</h3>
<p>
The best negative results help us learn from our mistakes. They can illuminate hidden obstacles or demonstrate why we need a change of course. An ideal submission to NOPE has a novel idea which sounds plausible from first principles or design intuition, but yields little to no improvement (in performance, power, area, …) in practice. The paper drills down into the reasons for the lack of improvement and proposes a plausible explanation – different technology trends, unexpected implementation complexity.
</p>
<h3>Prior NOPE Workshops</h3>
<a href="http://nope.pub/2021/">NOPE 2021</a> Virtal Event</br>
<a href="http://nope.pub/2020/">NOPE 2020</a> Lausanne, Switzerland (postponed)</br>
<a href="http://nope.pub/2019/">NOPE 2019</a> Providence, RI</br>
<a href="http://nope.pub/2017/">NOPE 2017</a> Cambridge, MA</br>
<a href="http://nope.pub/2016/">NOPE 2016</a> Taipei, Taiwan</br>
<a href="http://nope.pub/2015/">NOPE 2015</a> Waikiki, HI</br>
(with <a href="http://www.microarch.org/">MICRO</a> 2015-2017)
</div>
<div id="cfp" class="container section">
<h3>Call for Papers</h3>
<p>
Our goal is to find papers which the community can learn from and might otherwise have trouble finding a suitable venue, so we take a broad view of what constitutes a "negative" result.
</p>
<p>
We invite submissions from all sub-areas of systems and computer architecture. Submissions should focus on discussing historical perspectives, uncovering a misunderstood or misappropriated technical concept, analyzing the reasons for failure, especially in light of underlying assumptions. Submissions based on opinion, speculation, and non-fundamental circumstances ("there was a bug in the simulator") are not encouraged, as they do not provide concrete evidence as to whether an idea is bad. Topics of interest include:
</p>
<ul>
<li>Thorough evaluations of failed projects which uncover and characterize the root cause.</li>
<li>Reflections and historical perspectives on specific ideas, general trends, and "techno-social" factors in the systems and computer architecture community.</li>
<li>Papers which describe both positive and negative results, with an emphasis on the underlying reasons behind why some succeeded and other failed.</li>
<li>Cradle-to-grave examination of completed projects, specifically to dissect dead-ends and unworkable solutions encountered along the way.</li>
<li>Design space explorations or comprehensive experiments which suggest a particular technique is unlikely to work under a realistic set of assumptions.</li>
<li>Research which uncovers fundamental limitations in scalability, performance, accuracy, or other quantifiable metrics.</li>
<li>Any research which serves to share the lessons of failure to the broader community, such that we can avoid repeating them in the future.</li>
</ul>
<h3>Important dates</h3>
Paper submission: February 14, 2022<br>
Author notification: February 16, 2022<br>
Camera-ready version: February 23, 2022<br>
</div>
<div id="submission" class="container section">
<h3>Submission Guidelines</h3>
<p>We believe in substance over style, and we encourage authors to prioritize delivering their message over conforming to a particular template. That being said, we anticipate papers will be 2 pages, and we encourage authors to use a two-column format. Papers need not be anonymized.</p>
<p>Additionally, we ask that you also include a short, 1-paragraph abstract in your submission email. This should be suitable for inclusion on the NOPE website and program handouts.</p>
<p>Accepted submissions are expected to prepare a poster or short presentation to foster discussion.</p>
<h3>Submit</h3>
<p>Please submit your papers via email to: <a href="mailto:[email protected]?subject=Submission">[email protected]</a> by 5:00pm ET on the deadline.</p>
</div>
<div id="committee" class="container section">
<h3>Organizers</h3>
Lillian Pentecost, Harvard<br>
Udit Gupta, Harvard<br>
Thierry Tambe, Harvard<br>
Jeff (Jun) Zhang, Harvard<br>
David Brooks, Harvard<br>
Brandon Reagen, NYU / Facebook<br>
<h3>Contact</h3>
<p>Questions? <a href="mailto:[email protected]?subject=Question about the workshop">Send us an email.</a></p>
</div>
</div>
<!-- jQuery (necessary for Bootstrap's JavaScript plugins) -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.2/jquery.min.js"></script>
<!-- Include all compiled plugins (below), or include individual files as needed -->
<!-- Latest compiled and minified JavaScript -->
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.2/js/bootstrap.min.js"></script>
<!-- Analytics -->
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
})(window,document,'script','//www.google-analytics.com/analytics.js','ga');
ga('create', 'UA-66549897-1', 'auto');
ga('send', 'pageview');
</script>
</body>
</html>