I've been using the 2012 SDK for Linux for several months now with a P8400 processor on Fedora 17 and haven't hit any major issues so far. Today I went to the webite to get the 2013 version and I see the attached image. It seems that non-Xeon processors are not supported under Linux for the 2013 SDK and curiously the 2012 SDK no longer supports Linux(although the link is still there if one tries to download 2012).
My question is, what is Intel's long term strategy for OpenCL on Linux? Do they plan to support non-workstation processors like i5 and i3 under Linux? Thank you.
Intel continue the plans to support OpenCL with Linux our CPUs. With SDK 2013 XE, to run OpenCL 1.2 on the CPU you must have a system supporting the Intel® Streaming SIMD Extensions 4.2 (Intel® SSE 4.2) or higher (Intel® Advanced Vector Extensions (Intel® AVX)). This includes both Xeon and Core based processors. The table attached presents the main lead platforms for our different set of SDKs. More details on each SDK is included in the relevant release notes. We will update our documents respectively.
If your system doesn't support SEE 4.2, you will need to continue use the SDK 2012.
Thanks for the response. The 2012 SDK is working well for me.
My company issues laptops to all employees so it is very helpful to be able to compile and run OpenCL applications on such systems. I realize there are not infinite resources so deprecating older processors is understandable but it makes development easier when functional debugging can be done on any Intel platform.
Thanks for the feedback
You can even enhance your debugging experience on every supported systems. With the new SDK XE 2013 for Linux, you can use GDB to debug your code on the CPU. If your company laptops are windows based, you actually can debug the your kernel code directly with Visual Studio debugger and the Intel OpenCL debugger plug-in.
Will be great to get your feedback on those tools.