https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Scalable-kernels/m-p/909092#M4555 <P>Robert,</P> <P>What exactly is the bottleneck in the kernel? Memory routines, task switching, I/O, or something else? Can you quantify how much time is spent in the kernel vs. user space?</P> <P></P> <P>Kind regards</P> <P>Thomas</P> Wed, 02 Apr 2008 10:53:08 GMT Thomas_W_Intel 2008-04-02T10:53:08Z https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Scalable-kernels/m-p/909091#M4554 <P>Hi, I've been curious about the current state of scalable kernels. I tried looking around, but didn't find anything besides some research papers, and a few early Linux branches. Even after making our applications as concurrent as possible, the final software bottle-neck is the kernel.</P> <P>Anyone have any pointers?</P> Wed, 02 Apr 2008 04:17:55 GMT https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Scalable-kernels/m-p/909091#M4554 robert_jay_gould 2008-04-02T04:17:55Z https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Scalable-kernels/m-p/909092#M4555 <P>Robert,</P> <P>What exactly is the bottleneck in the kernel? Memory routines, task switching, I/O, or something else? Can you quantify how much time is spent in the kernel vs. user space?</P> <P></P> <P>Kind regards</P> <P>Thomas</P> Wed, 02 Apr 2008 10:53:08 GMT https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Scalable-kernels/m-p/909092#M4555 Thomas_W_Intel 2008-04-02T10:53:08Z