Intel Unveils Knights Landing Processor and Omni Scale I/O Fabric

Intel announced microarchitecture and memory details of its next-generation Intel Xeon Phi processor (code-named Knights Landing) and an end-to-end interconnect fabric optimized for fast data transfers, both aimed at powering High-Performance Computing (HPC) systems.

Knights Landing, which is expected to debut in the second half of 2015, will include up to16GB high-bandwidth, on-package memory at launch – designed in partnership with Micron – to deliver five times better bandwidth compared to DDR4 memory, five times better energy efficiency and three times more density than current GDDR-based memory. Knights Landing will be available in a PCIe-based card option, or mounted directly on the motherboard socket to remove the bandwidth bottlenecks of data transfer over PCIe, common in GPU and accelerator solutions.

Powered by more than 60 HPC-enhanced Silvermont architecture-based cores, Knights Landing is expected to deliver more than 3 TFLOPS of double-precision performance and three times the single-threaded performance compared with the current generation.

Intel's new Omni Scale fabric, which is built upon a combination of enhanced acquired IP from Cray and QLogic along with tis own in-house innovations, addresses the I/O requirements of the HPC sector.  Intel Omni Scale Fabric is an end-to-end I/O solution encompassing adapters, edge switches, director switch systems, cables and management software. Intel intends to replace traditional electrical transceivers in the director switches with silicon photonics, enabling increased port density, simplified cabling and reduced costs. Intel Silicon Photonics-based cabling and transceiver solutions may also be used with Intel Omni Scale-based processors, adapter cards and edge switches.

Intel noted continued market gains in powering the HPC sector.  The company said its current generation of Xeon processors and Xeon Phi coprocessors powers the top-rated system in the world – the 35 PFLOPS "Milky Way 2" in China, and its designs account for 85 percent of all supercomputers on the 43rd edition of the TOP500 list.

http://newsroom.intel.com/community/intel_newsroom/blog/2014/06/23/intel-re-architects-the-fundamental-building-block-for-high-performance-computing