blocks|key|1321474|text|好的，CUDA是对C的一个修改，为了编写CUDA内核，您必须用C编写代码，然后用nvidia的CUDA编译器编译成可执行的表单。生成的本机代码可以使用JNI与Java链接。所以，从技术上讲，您不能用Java编写内核代码。有JCUDA+http://www.jcuda.de/jcuda/JCuda.html，它为您提供了用于通用内存/设备管理的CUDA+apis和一些用CUDA和JNI封装实现的Java方法(FFT，一些线性代数方法)。等等)|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|1321475|另一方面，OpenCL只是一个API。OpenCL内核是传递给API的普通字符串，因此使用来自OpenCL的OpenCL，您应该能够指定自己的内核。java的OpenCL绑定可以在这里找到，http://www.jocl.org/。|1321476|entityMap|0|LINK|mutability|MUTABLE|url|http://www.jcuda.de/jcuda/JCuda.html|1|http://www.jocl.org/^0|39|10|0|0|2N|K|1|0^^$0|@$1|2|3|4|5|6|7|P|8|@]|9|@$A|Q|B|R|1|S]]|C|$]]|$1|D|3|E|5|6|7|T|8|@]|9|@$A|U|B|V|1|W]]|C|$]]|$1|F|3|-4|5|6|7|X|8|@]|9|@]|C|$]]]|G|$H|$5|I|J|K|C|$L|M]]|N|$5|I|J|K|C|$L|O]]]]

Well CUDA is a modification of C, to write CUDA kernel you have to code in C, and then compile to executable form with nvidia's CUDA compiler. Produced native code could then be linked with Java using JNI. So technically you can't write kernel code from Java. There is JCUDA <a href="http://www.jcuda.de/jcuda/JCuda.html" rel="noreferrer">http://www.jcuda.de/jcuda/JCuda.html</a>, it provides you with cuda's apis for general memory/device menagement and some Java methods that are implemented in CUDA and JNI wrapped (FFT, some linear algebra methods.. etc etc..). 

On the other hand OpenCL is just an API. OpenCL kernels are plain strings passed to the API so using OpenCL from Java you should be able to specify your own kernels. OpenCL binding for java can be found here <a href="http://www.jocl.org/" rel="noreferrer">http://www.jocl.org/</a>.

blocks|key|34901|text|我一直在使用JOCL，我对此非常满意。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|34902|OpenCL相对于CUDA的主要缺点(至少对我来说)是缺乏可用的库(+main、CUDPP等)。然而，CUDA可以很容易地移植到OpenCL上，通过查看这些库(算法、策略等)的工作原理，实际上非常好，因为您可以从中学到很多东西。|34903|entityMap^0|0|0^^$0|@$1|2|3|4|5|6|7|F|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|G|8|@]|9|@]|A|$]]|$1|D|3|-4|5|6|7|H|8|@]|9|@]|A|$]]]|E|$]]

I've been using JOCL and I'm very happy with it.

The main disadvantage of OpenCL over CUDA (at least for me) is the lack of available libraries (Thrust, CUDPP, etc). However CUDA can be easily ported to OpenCL, and by looking at how those libraries work (algorithms, strategies, etc) is actually very nice as you learn a lot with it.

blocks|key|1584945|text|AFAIK，JavaCL+/+OpenCL4Java是目前所有平台上唯一可用的OpenCL绑定(包括MacOS+X、FreeBSD、Linux、Solaris，都是因使用南国防军而在Intel+32、64位和ppc变体中使用)。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|1584946|它的演示程序实际上在Java上运行得很好，至少在Mac和Windows上是这样的(为了避免在Linux上随机崩溃，请看这个wiki页面，比如这个粒子演示+)。|1584947|它还附带了一些实用程序(GPGPU随机数生成、基本并行约简、线性代数)和一个Scala+DSL。|1584948|最后，它是可用的最古老的绑定(自2009年6月以来)和它有一个活跃的用户社区。。|1584949|(免责声明:我是JavaCL的作者：-)|1584950|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/nativelibs4java/JavaCL|1|https://github.com/twall/jna/|2|http://code.google.com/p/javacl/wiki/TroubleShootingJavaCLOnLinux|3|http://nativelibs4java.sourceforge.net/webstart/OpenCL/ParticlesDemo.jnlp|4|http://ochafik.free.fr/blog/?p=207|5|http://groups.google.com/group/nativelibs4java|6|http://code.google.co/^0|6|K|0|2D|4|1|0|1N|8|2|20|4|3|0|12|9|4|0|R|C|5|0|8|6|6|0^^$0|@$1|2|3|4|5|6|7|15|8|@]|9|@$A|16|B|17|1|18]|$A|19|B|1A|1|1B]]|C|$]]|$1|D|3|E|5|6|7|1C|8|@]|9|@$A|1D|B|1E|1|1F]|$A|1G|B|1H|1|1I]]|C|$]]|$1|F|3|G|5|6|7|1J|8|@]|9|@$A|1K|B|1L|1|1M]]|C|$]]|$1|H|3|I|5|6|7|1N|8|@]|9|@$A|1O|B|1P|1|1Q]]|C|$]]|$1|J|3|K|5|6|7|1R|8|@]|9|@$A|1S|B|1T|1|1U]]|C|$]]|$1|L|3|-4|5|6|7|1V|8|@]|9|@]|C|$]]]|M|$N|$5|O|P|Q|C|$R|S]]|T|$5|O|P|Q|C|$R|U]]|V|$5|O|P|Q|C|$R|W]]|X|$5|O|P|Q|C|$R|Y]]|Z|$5|O|P|Q|C|$R|10]]|11|$5|O|P|Q|C|$R|12]]|13|$5|O|P|Q|C|$R|14]]]]

AFAIK, <a href="https://github.com/nativelibs4java/JavaCL" rel="noreferrer" title="JavaCL / OpenCL4Java">JavaCL / OpenCL4Java</a> is the only OpenCL binding that is available on all platforms right now (including MacOS X, FreeBSD, Linux, Windows, Solaris, all in Intel 32, 64 bits and ppc variants, thanks to its use of <a href="https://github.com/twall/jna/" rel="noreferrer" title="JNA">JNA</a>).

It has demos that actually run fine from Java Web Start at least on Mac and Windows (to avoid random crashes on Linux, please see <a href="http://code.google.com/p/javacl/wiki/TroubleShootingJavaCLOnLinux" rel="noreferrer" title="this wiki page">this wiki page</a>, such as this <a href="http://nativelibs4java.sourceforge.net/webstart/OpenCL/ParticlesDemo.jnlp" rel="noreferrer" title="Particles Demo">Particles Demo</a>.

It also comes with a few utilities (GPGPU random number generation, basic parallel reduction, linear algebra) and a <a href="http://ochafik.free.fr/blog/?p=207" rel="noreferrer" title="Scala DSL">Scala DSL</a>.

Finally, it's the oldest bindings available (since june 2009) and <a href="http://groups.google.com/group/nativelibs4java" rel="noreferrer" title="NativeLibs4Java&#39;s user group">it has an active user community</a>.

(Disclaimer: I'm <a href="http://code.google.co" rel="noreferrer">JavaCL</a>'s author :-))

blocks|key|35003|text|您还可以考虑阿帕拉皮。它允许您用Java编写代码，并将尝试在运行时将字节码转换为OpenCL。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|35004|全部披露。我是Aparapi开发者。|35005|entityMap|0|LINK|mutability|MUTABLE|url|http://github.com/aparapi/aparapi^0|6|4|0|0|0^^$0|@$1|2|3|4|5|6|7|N|8|@]|9|@$A|O|B|P|1|Q]]|C|$]]|$1|D|3|E|5|6|7|R|8|@]|9|@]|C|$]]|$1|F|3|-4|5|6|7|S|8|@]|9|@]|C|$]]]|G|$H|$5|I|J|K|C|$L|M]]]]

You may also consider <a href="http://github.com/aparapi/aparapi" rel="nofollow noreferrer">Aparapi</a>. It allows you to write your code in Java and will attempt to convert bytecode to OpenCL at runtime. 

Full disclosure. I am the Aparapi developer.

blocks|key|1321516|text|我知道很晚了，但是看看这个：https://github.com/pcpratts/rootbeer1|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|1321517|我没有使用它，但似乎比其他解决方案更容易使用。|1321518|从项目页面：|1321519|OpenCL比CUDA或绑定更先进。使用绑定，开发人员必须将复杂的对象图序列化为基本类型数组。对于Rootbeer，这是自动完成的。此外，通过语言绑定，开发人员必须用CUDA或OpenCL编写GPU内核。使用Rootbeer，对Java字节码进行静态分析(使用碳烟)，并自动生成CUDA代码。|1321520|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/pcpratts/rootbeer1^0|E|11|0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|R|8|@]|9|@$A|S|B|T|1|U]]|C|$]]|$1|D|3|E|5|6|7|V|8|@]|9|@]|C|$]]|$1|F|3|G|5|6|7|W|8|@]|9|@]|C|$]]|$1|H|3|I|5|6|7|X|8|@]|9|@]|C|$]]|$1|J|3|-4|5|6|7|Y|8|@]|9|@]|C|$]]]|K|$L|$5|M|N|O|C|$P|Q]]]]

I know it's late but take a look at this: <a href="https://github.com/pcpratts/rootbeer1">https://github.com/pcpratts/rootbeer1</a>

I have not worked with it but seems much easier to use than other solutions.

From the project page: 

Rootbeer is more advanced than CUDA or OpenCL Java Language Bindings. With bindings the developer must serialize complex graphs of objects into arrays of primitive types. With Rootbeer this is done automatically. Also with language bindings, the developer must write the GPU kernel in CUDA or OpenCL. With Rootbeer a static analysis of the Java Bytecode is done (using Soot) and CUDA code is automatically generated.

blocks|key|35104|text|我还可以推荐jogamp.org的JOCL，在Linux、Mac和Windows上工作。例如，康拉德与JOCL结合使用了大量OpenCL。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|35105|entityMap|0|LINK|mutability|MUTABLE|url|http://jogamp.org/jocl/www/|1|http://conrad.stanford.edu/^0|6|F|0|1B|3|1|0^^$0|@$1|2|3|4|5|6|7|N|8|@]|9|@$A|O|B|P|1|Q]|$A|R|B|S|1|T]]|C|$]]|$1|D|3|-4|5|6|7|U|8|@]|9|@]|C|$]]]|E|$F|$5|G|H|I|C|$J|K]]|L|$5|G|H|I|C|$J|M]]]]

I can also recommend <a href="http://jogamp.org/jocl/www/" rel="nofollow">JOCL by jogamp.org</a>, works on Linux, Mac, and Windows. <a href="http://conrad.stanford.edu" rel="nofollow">CONRAD</a>, for example, uses heavily OpenCL in combination with JOCL.

blocks|key|1585083|text|您可以查看CUDA4J+API。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1585084|http://sett.com/gpgpu/the-cuda4j-api|offset|length|1585085|entityMap|0|LINK|mutability|MUTABLE|url^0|0|0|10|0|0^^$0|@$1|2|3|4|5|6|7|M|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|N|8|@]|9|@$D|O|E|P|1|Q]]|A|$]]|$1|F|3|-4|5|6|7|R|8|@]|9|@]|A|$]]]|G|$H|$5|I|J|K|A|$L|C]]]]

You can take a look at the CUDA4J API

<a href="http://sett.com/gpgpu/the-cuda4j-api" rel="nofollow">http://sett.com/gpgpu/the-cuda4j-api</a>

blocks|key|1585087|text|如果您想要做一些图像处理或几何操作，您可能需要一个线性代数库的gpu支持(例如，与CUDA+)。我会建议你ND4J女巫是线性白鹭与CUDA+GPU的支持，DeepLearning4J是建立在此基础上。这样，您就不必直接处理CUDA，也不必在c中使用低级代码。此外，如果您想使用DL4J处理更多的图像，您将可以访问特定的图像处理操作，例如卷积。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1585088|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|D|8|@]|9|@]|A|$]]|$1|B|3|-4|5|6|7|E|8|@]|9|@]|A|$]]]|C|$]]

If you want to do some image processing or geometric operations, you may want a linear algebra library with gpu support (with CUDA for instance). I would suggest you ND4J witch is the linear algrebra with CUDA GPU support on which DeepLearning4J is built. With that you don't have to deal with CUDA directly and have to low level code in c. Plus if you want to do more stuff with image with DL4J you will have access to specific image processing operations such as convolution.

General-purpose computing on graphics processing units (<a href="http://en.wikipedia.org/wiki/GPGPU" rel="noreferrer">GPGPU</a>) is a very attractive concept to harness the power of the GPU for any kind of computing. 

I'd love to use GPGPU for image processing, particles, and fast geometric operations.

Right now, it seems the two contenders in this space are CUDA and OpenCL. I'd like to know:

<ul>
<li>Is OpenCL usable yet from Java on Windows/Mac?</li>
<li>What are the libraries ways to interface to OpenCL/CUDA?</li>
<li>Is using JNA directly an option?</li>
<li>Am I forgetting something?</li>
</ul>

Any real-world experience/examples/war stories are appreciated.

Best approach for GPGPU/CUDA/OpenCL in Java?

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

教程

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云智能顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

EdgeOne AI 安全实战专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

聚焦“写作效率、视觉美观与运行性能”三方面进行全面升级，为您提供更高效、稳定的创作环境

社区富文本&Markdown编辑器全新改版上线，欢迎大家体验!

诚挚邀请您参与本次调研，分享您的真实使用感受与建议。您的反馈至关重要，感谢您的支持与参与！

社区新版编辑器体验调研

图形处理单元上的通用计算()是一个非常吸引人的概念，可以将GPU的能力用于任何类型的计算。我想使用GPGPU的图像处理，粒子和快速几何操作。目前，这一领域的两个竞争者似乎是CUDA和OpenCL。我想知道：OpenCL在Windows/Mac上还可用吗？图书馆与OpenCL/CUDA的接口方式是什么？直接使用JNA是一...

问Java中GPGPU/CUDA/OpenCL的最佳方法？
EN

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问Java中GPGPU/CUDA/OpenCL的最佳方法？EN