forked from m4rs-mt/ILGPU
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
428 lines (418 loc) · 24.9 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
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
---
layout: home
extras-js: index/extras-js.html
---
{%- assign social = site.social_links -%}
{% assign its = site.docs | where: "path", site.data.sidebar.main.path %}
{% for it in its %}
{% assign docs_url = it.url | relative_url %}
{% endfor %}
<!-- ======= Hero Section ======= -->
<section id="hero" class="d-flex align-items-center">
<div class="container">
<div class="row">
<div class="col-lg-6 d-flex flex-column justify-content-center pt-4 pt-lg-0 order-2 order-lg-1"
data-aos="fade-up" data-aos-delay="200">
<h1>A Modern GPU Compiler for .NET Programs</h1>
<h2>Developed by Marcel K<span id="easter-egg" onclick="easteregg()">ö</span>ster</h2>
<div class="d-flex justify-content-center justify-content-lg-start">
<a href="#learn-more" class="btn-get-started btn-hero-1 scrollto">Learn More</a>
<a href="#cta" class="btn-get-started scrollto">Get Started</a>
</div>
</div>
<div class="col-lg-6 order-1 order-lg-2 hero-img" data-aos="zoom-in" data-aos-delay="200">
<img src="{{ "assets/img/hero-section.png" | relative_url }}" class="img-fluid animated"
alt="graphics cpu">
</div>
</div>
</div>
</section><!-- End Hero -->
<main id="main">
<!-- ======= Learn More Section ======= -->
<section id="learn-more" class="why-us section-bg">
<div class="container-fluid" data-aos="fade-up">
<div class="row" style="min-height: 554px !important;">
<div class="col-lg-6 d-flex flex-column align-items-center order-2 order-lg-1 img" data-aos="zoom-in"
data-aos-delay="150">
<div class="image-holder">
<img src="{{ "assets/img/learn-more.png" | relative_url }}" alt="graphics cpu">
</div>
</div>
<div class="col-lg-6 d-flex flex-column justify-content-center align-items-stretch order-1 order-lg-2">
<div class="content">
<h3>Learn more about ILGPU</h3>
<p>
A modern, lightweight & fast GPU compiler for high-performance .Net programs
</p>
</div>
<div class="accordion-list">
<ul>
<li>
<a data-bs-toggle="collapse" class="collapse"
data-bs-target="#accordion-list-1"><span>•</span>What is ILGPU?<i
class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="accordion-list-1" class="collapse show" data-bs-parent=".accordion-list">
<p>
ILGPU is a new JIT (just-in-time) compiler for high-performance GPU programs
(also known as kernels) written in .Net-based languages. ILGPU is completely
written in C# without any native dependencies which allows you to write GPU
programs that are truly portable.
</p>
</div>
</li>
<li>
<a data-bs-toggle="collapse" data-bs-target="#accordion-list-2" class="collapsed"><span>•</span>What
does ILGPU do?<i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="accordion-list-2" class="collapse" data-bs-parent=".accordion-list">
<p>
It combines the convenience of C++ AMP with the high performance of CUDA.
Functions in the scope of kernels do not have to be annotated (e.g. default C#
functions) and are allowed to work on value types. All kernels (including all
hardware features like shared memory, atomics and warp shuffles) can be executed
and debugged on the CPU using the integrated multi-threaded CPU accelerator.
</p>
</div>
</li>
<li>
<a data-bs-toggle="collapse" data-bs-target="#accordion-list-3" class="collapsed"><span>•</span>Do
I need to pay for ILGPU?<i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="accordion-list-3" class="collapse" data-bs-parent=".accordion-list">
<p>
Nope! ILGPU is released under the University of Illinois/NCSA Open Source
License. It is a free project and open-source project supported mainly by
<a href="https://www.gresearch.co.uk/"
style="display: inline; padding-right: unset; font-weight: 400; font-size: 1rem;"
target="_blank">G-Research</a>. Support the project with contributions or a
small donation in order
to speed up the development process and to keep the project alive.
</p>
</div>
</li>
</ul>
</div>
</div>
</div>
</div>
</section><!-- End Learn More Section -->
<!-- ======= Highlights Section ======= -->
<section id="highlights" class="services">
<div class="container" data-aos="fade-up">
<div class="section-title">
<h2>Highlights</h2>
</div>
<div class="row" data-aos="fade-up">
<div class="col-xl-3 col-md-6 d-flex align-items-stretch">
<div class="icon-box">
<div class="icon"><i class="bx bxs-zap"></i></div>
<h4><a>High Performance</a></h4>
<p>High performance kernel compilation, dispatch and execution times. Furthermore, type-safe
kernel delegates avoid boxing.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-md-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-sort-alt-2"></i></div>
<h4><a>High Convenience</a></h4>
<p>Use the power of C# or F# to write high-level kernels and execute them on the GPU. No
need to program C++, Cuda or OpenCL.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-tachometer"></i></div>
<h4><a>CPU Accelerator</a></h4>
<p>Single- or multi-threaded execution of kernels on the CPU. This is also useful for debugging
or emulation of specific target platforms.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bxs-bug-alt"></i></div>
<h4><a>Advanced Debugging</a></h4>
<p>High-level kernel debugging using your favorite .Net debugger. Furthermore, the
single-threaded execution feature allows to focus on the algorithm instead of the
parallelism.</p>
</div>
</div>
</div>
<div class="row">
<div class="col-xl-3 col-md-6 d-flex align-items-stretch">
<div class="icon-box">
<div class="icon"><i class="bx bxs-no-entry"></i></div>
<h4><a>No Function Annotations</a></h4>
<p>Functions do not have to be annotated in order to use them in the scope of kernels.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-md-0">
<div class="icon-box">
<div class="icon"><i class="bx bxs-chip"></i></div>
<h4><a>Any-CPU Builds</a></h4>
<p>Compile your applications for any CPU. ILGPU will automatically adjust everything else for
X86 or X64 platforms.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-outline"></i></div>
<h4><a>Implicitly Grouped Kernels</a></h4>
<p>Implicitly grouped kernels let you implement high-level kernels without paying attention to
low-level index computations or tiling.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-layer"></i></div>
<h4><a>Multi-dimensional Indices</a></h4>
<p>Multi-dimensional index types simplify address computations and kernel writing.</p>
</div>
</div>
</div>
<div class="row">
<div class="col-xl-3 col-md-6 d-flex align-items-stretch">
<div class="icon-box">
<div class="icon"><i class="bx bxs-coin-stack"></i></div>
<h4><a>Array Views</a></h4>
<p>No pointer arithmetic and dramatically simplified index computations due to views to memory
regions.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-md-0">
<div class="icon-box">
<div class="icon"><i class="bx bxs-microchip"></i></div>
<h4><a>Shared Memory</a></h4>
<p>Support for shared (scratch-pad) memory in kernels via array views. Static or dynamic
allocation of shared memory is supported.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-atom"></i></div>
<h4><a>Atomics and Low-Level Intrinsics</a></h4>
<p>Easy access to atomic functions and low-level-intrinsics like warp shuffles. All functions
are supported during CPU debugging.</p>
</div>
</div>
<div class="col-xl-3 col-md-6 d-flex align-items-stretch mt-4 mt-xl-0">
<div class="icon-box">
<div class="icon"><i class="bx bx-list-plus"></i></div>
<h4><a>High-Performance Math Functions</a></h4>
<p>Default math functions and operations are mapped to high-performance math functions.
Furthermore, there is support for fast math and forced 32bit math to avoid doubles.</p>
</div>
</div>
</div>
</div>
</section><!-- End Highlights Section -->
<!-- ======= Features Comparison Section ======= -->
<section id="features" class="section-bg">
<div class="container" data-aos="fade-up">
<div class="section-title">
<h2>Features Comparison</h2>
</div>
<div id="chart-option">
<div class="container">
<div class="row">
<div class="col-md-12">
<div class="divider"></div>
</div>
</div>
<div class="row header"></div>
<div class="row">
<div class="col-md-12">
<table class="table table-borderless align-middle">
<thead class="features-header">
<th class="features-header-name w-50">Features</th>
<th class="features-header-col">ILGPU</th>
<th class="features-header-col">C++ AMP</th>
<th class="features-header-col">Cuda</th>
</thead>
<tbody class="features-body">
{% for feature in site.data.features %}
<tr>
<td class="features-body-name">
{{ feature.name }}
</td>
<td class="features-body-col">
<div class="spec">
{% if feature.ilgpu == "yes" %}
<i class='bx bx-check'></i>
{% elsif feature.ilgpu == "partial" %}
<i class='bx bx-check partial'></i>
{% elsif feature.ilgpu == "scheduled" %}
<i class='bx bx-check scheduled'></i>
{% endif %}
</div>
</td>
<td class="features-body-col">
<div class="spec">
{% if feature.cppamp == "yes" %}
<i class='bx bx-check'></i>
{% elsif feature.cppamp == "partial" %}
<i class='bx bx-check partial'></i>
{% endif %}
</div>
</td>
<td class="features-body-col">
<div class="spec">
{% if feature.cuda == "yes" %}
<i class='bx bx-check'></i>
{% elsif feature.cuda == "partial" %}
<i class='bx bx-check partial'></i>
{% endif %}
</div>
</td>
</tr>
{% endfor %}
</tbody>
</table>
</div>
</div>
<div class="row future-support">
<p>
<span class="partial">Yellow checkmarks</span> indicate partial or limited support.<br/>
Features marked with an <span class="scheduled">orange checkmark</span> will be available in
the future.
</p>
</div>
</div>
</div>
</div>
</section><!-- End Features Comparison Section -->
<!-- ======= Cta Section ======= -->
<section id="cta" class="cta">
<div class="container" data-aos="zoom-in">
<div class="row">
<div class="text-center text-lg-start">
<h3>Ready to get started?</h3>
<p>Take a look through the documentation, create your first ILGPU project, and join the community on
Discord!</p>
</div>
</div>
<div class="row">
<div class="col-lg-6 text-center">
<a class="cta-btn cta-btn-1" href="{{ docs_url }}">Documentation</a>
<a class="cta-btn" href="{{ social.github | default: '#' }}/" target="_blank"><i
class="bx bxl-github"></i>Github</a>
<a class="cta-btn" href="{{ social.discord | default: '#' }}" target="_blank"><i
class="bx bxl-discord"></i>Discord</a>
</div>
</div>
</div>
</section><!-- End Cta Section -->
<!-- ======= News Section ======= -->
<section id="news" class="pricing">
<div class="container" data-aos="fade-up">
<div class="section-title">
<h2>News</h2>
<p>Updates and news related to ILGPU and the community.</p>
</div>
<div class="row">
{% for post in site.posts limit: 3 %}
<div class="col-lg-4" data-aos="fade-up" data-aos-delay="100">
<div class="box">
<h5>{{ post.title }}</h5>
<p class="excerpt">{{ post.excerpt | strip_html }}</p>
{% comment %}<br/><br/>{% endcomment %}
<a href="{{ post.url | relative_url }}" class="buy-btn">Full Article</a>
</div>
</div>
{% endfor %}
</div>
</div>
</section><!-- End News Section -->
<!-- ======= Frequently Asked Questions Section ======= -->
<section id="faq" class="faq section-bg">
<div class="container" data-aos="fade-up">
<div class="section-title">
<h2>Frequently Asked Questions</h2>
</div>
<div class="faq-list">
<ul>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse" class="collapse"
data-bs-target="#faq-list-1">Are exceptions
supported? <i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-1" class="collapse show" data-bs-parent=".faq-list">
<p>
Exceptions require support for exception handlers and a limited support for reference
types. Changes of the "intended" control flow (which can be caused by exceptions) are
currently not supported. However, there might be a conversion phase in the future that
converts several exceptions into debug assertions.
</p>
</div>
</li>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse"
data-bs-target="#faq-list-2" class="collapsed">What
about debug assertions? <i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-2" class="collapse" data-bs-parent=".faq-list">
<p>
Debug assertions are supported on all accelerators and can be enabled via one of the
Context flags.
</p>
</div>
</li>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse"
data-bs-target="#faq-list-3" class="collapsed">Are
class types supported? And what about lambda functions? <i
class="bx bx-chevron-down icon-show"></i><i class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-3" class="collapse" data-bs-parent=".faq-list">
<p>
Reference types are currently not supported. However, a limited support for reference
types will be added in the future. This will also allow the implementation of delegates.
Lambda functions (or delegates in general) are currently not supported since they
require a limited support for reference types and custom code-transformation passes.
Support for lambda functions will be added in the future.
</p>
</div>
</li>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse"
data-bs-target="#faq-list-4" class="collapsed">Can
I debug a kernel on the GPU? <i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-4" class="collapse" data-bs-parent=".faq-list">
<p>
There is basic support for hardware-based kernel debugging and profiling. However,
CPU-based kernel debugging is recommended in all cases due to the advanced debugging and
testing capabilities.
</p>
</div>
</li>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse"
data-bs-target="#faq-list-5" class="collapsed">What
about .Net Standard support? <i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-5" class="collapse" data-bs-parent=".faq-list">
<p>
The new ILGPU version supports .Net 4.7 and .Net Standard 2.1 (e.g. .Net Core 3.1/.Net
5.0).
</p>
</div>
</li>
<li>
<i class="bx bx-help-circle icon-help"></i> <a data-bs-toggle="collapse"
data-bs-target="#faq-list-6" class="collapsed">What
about Linux and Mac support? <i class="bx bx-chevron-down icon-show"></i><i
class="bx bx-chevron-up icon-close"></i></a>
<div id="faq-list-6" class="collapse" data-bs-parent=".faq-list">
<p>
ILGPU supports .Net Core, which allows writing portable .Net applications. Since ILGPU
is written in C# and does not rely on native libraries in the current version, kernels
can be run on all .Net Core/.Net 5 compatible platforms. This allows you to compile your
application (including GPU code) only once.
</p>
</div>
</li>
</ul>
</div>
</div>
</section><!-- End Frequently Asked Questions Section -->
</main><!-- End #main -->