blocks|key|199732|text|只能通过内网接口获取。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|199733|F.e.https://github.com/linusyang/opencl-test-ios|offset|length|199734|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/linusyang/opencl-test-ios^0|0|4|18|0|0^^$0|@$1|2|3|4|5|6|7|N|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|O|8|@]|9|@$D|P|E|Q|1|R]]|A|$]]|$1|F|3|-4|5|6|7|S|8|@]|9|@]|A|$]]]|G|$H|$5|I|J|K|A|$L|M]]]]

Available only through private API.

F.e. <a href="https://github.com/linusyang/opencl-test-ios">https://github.com/linusyang/opencl-test-ios</a>

blocks|key|3679452|text|即使使用OpenCL作为私有框架，基于iOS的也不会给你带来GPU+(或者其他像DSP/FPGA，如果存在)的好处。它只是在arm处理器上提供了多个可用内核。我运行了以下代码来验证在iOS和OS中可以访问的OpenCL设备。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|BOLD|entityRanges|data|3679453|iOS上的输出|3679454|3679455|+ARM处理器设备|blockquote|3679456|3679457|3679458|OS+X上的输出|3679459|3679460|+Radeon+HD+4670|3679461|3679462|英特尔(R)酷睿(TM)+i3处理器540+@3.07+540|3679463|3679464|排除了错误检查的源。使用可用的OpenCL标头()和从(/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs/iPhoneOS8.1.sdk/System/Library/PrivateFrameworks)链接OpenCL|3679465|#include+"OpenCL/cl.h"
#include+<iostream>

cl_context_properties+prop[]+=+{+CL_CONTEXT_PLATFORM,+0,+0+};++++
//get+num+of+platforms
cl_uint+num;
cl_int+err+=+clGetPlatformIDs(0,+0,+&num);


//get+each+platform
cl_platform_id+*platforms+=+new+cl_platform_id[num];
err+=+clGetPlatformIDs(num,+platforms,+&num);

//create+context+for+platform
prop[1]+=+(cl_context_properties)+platforms[0];
cl_context+context+=+clCreateContextFromType(prop,+CL_DEVICE_TYPE_ALL,+NULL,+NULL,+&err);


//get+num+devices
size_t+numDevices=0;
clGetContextInfo(context,+CL_CONTEXT_DEVICES,+0,+NULL,+&numDevices);
cl_device_id+*devices+=+new+cl_device_id[+numDevices+];

//get+every+device
clGetContextInfo(context,+CL_CONTEXT_DEVICES,+numDevices,+devices,+0);

//get+info+of+every+device
for(+int+idx=0;+idx+<+numDevices;+%2B%2Bidx)+{

++++size_t+bufSize=0;
++++clGetDeviceInfo(devices[idx],+CL_DEVICE_NAME,+0,+NULL,+&bufSize);
++++if(+bufSize+>+0+)+{
++++++++char*+devName+=+new+char[+bufSize+];
++++++++clGetDeviceInfo(devices[idx],+CL_DEVICE_NAME,+bufSize,+devName,+0);
++++++++std::cout+<<+"Device+Name:+"+<<+devName+<<+'\n';
++++}
}|code-block|syntax|javascript|3679466|建议:到目前为止，我们需要使用OpenGL(2)或加速框架(3)。仍然不确定，出于什么原因/目的，在iPhone上复制OpenCL作为私有框架。|3679467|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/BradLarson/GPUImage|1|https://developer.apple.com/library/ios/documentation/Accelerate/Reference/AccelerateFWRef/_index.html^0|4|6|N|A|0|5|2|0|0|0|9|0|0|0|0|4|6|2|0|0|0|F|0|0|0|V|0|0|0|4R|0|0|0|20|M|1|0|U|1|1|0^^$0|@$1|2|3|4|5|6|7|1G|8|@$9|1H|A|1I|B|C]|$9|1J|A|1K|B|D]]|E|@]|F|$]]|$1|G|3|H|5|6|7|1L|8|@$9|1M|A|1N|B|D]]|E|@]|F|$]]|$1|I|3|-4|5|6|7|1O|8|@]|E|@]|F|$]]|$1|J|3|K|5|L|7|1P|8|@$9|1Q|A|1R|B|D]]|E|@]|F|$]]|$1|M|3|-4|5|6|7|1S|8|@]|E|@]|F|$]]|$1|N|3|-4|5|6|7|1T|8|@]|E|@]|F|$]]|$1|O|3|P|5|6|7|1U|8|@$9|1V|A|1W|B|D]|$9|1X|A|1Y|B|D]]|E|@]|F|$]]|$1|Q|3|-4|5|6|7|1Z|8|@]|E|@]|F|$]]|$1|R|3|S|5|L|7|20|8|@$9|21|A|22|B|D]]|E|@]|F|$]]|$1|T|3|-4|5|6|7|23|8|@]|E|@]|F|$]]|$1|U|3|V|5|6|7|24|8|@$9|25|A|26|B|D]]|E|@]|F|$]]|$1|W|3|-4|5|6|7|27|8|@]|E|@]|F|$]]|$1|X|3|Y|5|6|7|28|8|@$9|29|A|2A|B|D]]|E|@]|F|$]]|$1|Z|3|10|5|11|7|2B|8|@]|E|@]|F|$12|13]]|$1|14|3|15|5|6|7|2C|8|@$9|2D|A|2E|B|D]]|E|@$9|2F|A|2G|1|2H]|$9|2I|A|2J|1|2K]]|F|$]]|$1|16|3|-4|5|6|7|2L|8|@]|E|@]|F|$]]]|17|$18|$5|19|1A|1B|F|$1C|1D]]|1E|$5|19|1A|1B|F|$1C|1F]]]]

Even with using <code>OpenCL</code> as private framework, on iOS it won't give you the benefits of GPU ( or others like DSPs/FPGAs if existing ). It just gives you multiple cores available on arm processor. I ran the below code to verify the OpenCL devices accessible in iOS and OS X. 

Output on iOS 

<blockquote>
 ARM CPU Device
</blockquote>

Output on OS X

<blockquote>
 Radeon HD 4670 
 Intel(R) Core(TM) i3 CPU 540 @ 3.07GHz
</blockquote>

Source with error checks excluded. Using OpenCL headers available(<a href="https://github.com/linusyang/opencl-test-ios/tree/master/include/OpenCL">1</a>) and linking OpenCL from (/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs/iPhoneOS8.1.sdk/System/Library/PrivateFrameworks)

<pre><code>#include "OpenCL/cl.h"
#include &lt;iostream&gt;

cl_context_properties prop[] = { CL_CONTEXT_PLATFORM, 0, 0 }; 
//get num of platforms
cl_uint num;
cl_int err = clGetPlatformIDs(0, 0, &amp;num);


//get each platform
cl_platform_id *platforms = new cl_platform_id[num];
err = clGetPlatformIDs(num, platforms, &amp;num);

//create context for platform
prop[1] = (cl_context_properties) platforms[0];
cl_context context = clCreateContextFromType(prop, CL_DEVICE_TYPE_ALL, NULL, NULL, &amp;err);


//get num devices
size_t numDevices=0;
clGetContextInfo(context, CL_CONTEXT_DEVICES, 0, NULL, &amp;numDevices);
cl_device_id *devices = new cl_device_id[ numDevices ];

//get every device
clGetContextInfo(context, CL_CONTEXT_DEVICES, numDevices, devices, 0);

//get info of every device
for( int idx=0; idx &lt; numDevices; ++idx) {

 size_t bufSize=0;
 clGetDeviceInfo(devices[idx], CL_DEVICE_NAME, 0, NULL, &amp;bufSize);
 if( bufSize &gt; 0 ) {
 char* devName = new char[ bufSize ];
 clGetDeviceInfo(devices[idx], CL_DEVICE_NAME, bufSize, devName, 0);
 std::cout &lt;&lt; "Device Name: " &lt;&lt; devName &lt;&lt; '\n';
 }
}
</code></pre>

Suggestion: As of now, we will need to use either OpenGL(<a href="https://github.com/BradLarson/GPUImage">2</a>) or Accelerate framework(<a href="https://developer.apple.com/library/ios/documentation/Accelerate/Reference/AccelerateFWRef/_index.html">3</a>). Still not sure, for what reason/purpose OpenCL is copied in as private framework on iPhone.

I found this thread on the forum <a href="https://stackoverflow.com/questions/3258257/are-either-the-ipad-or-iphone-capable-of-opencl/5182480#5182480">Are either the IPad or IPhone capable of OpenCL?</a> but is it quite old. Also, what I can gather that OpenCL is available to system libraries of iOS but not to public. Is there more info in this regard or any update ?

Is available OpenCL on iOS

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

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

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

我在论坛上发现了这个帖子，但是它是不是很老了？此外，我可以收集到的是，OpenCL可用于iOS的系统库，但不能用于公共库。在这方面有更多的信息或任何更新吗？

问在iOS上是否提供OpenCL
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

问在iOS上是否提供OpenCLEN