Intel’s 72-Core ‘Knight’s Landing’ Xeon Phi Supercomputer Chip Cleared For Takeoff

The supercomputing segment is set to get a big boost from new silicon announced today at Intel. That silicon is a new version of Xeon Phi, otherwise known as Knight's Landing. Whatever you want to call it, the pre-production chip is a 72-core coprocessor solution manufactured on a 14nm process with 3D Tri-Gate transistors.

These aren't CPUs like the kind you drop into your motherboard. They're coprocessors built around Intel's MIC (Many Integrated Core) architecture that, just like it sounds, combines a whole bunch of cores into a single chip, which itself is part of a larger PCI-E add-in card solution for supercomputing applications.

Intel Xeon Phi Die
Intel Knight's Landing Processor Die Map

The add-in cards run alongside engines just like NVIDIA's Tesla GPU accelerators and help with advanced number crunching chores. Where the MIC architecture shines is in offering a high degree of parallelism in a relatively power efficient package. That makes Knight's Landing suitable for a wide range of supercomputing tasks, like climate change simulations, genetic analysis, investment portfolio risk management, searching for new energy sources, and the list goes on.

Knight's Landing succeeds the current version of Xeon Phi, codenamed Knight's Corner, which has up to 61 cores. The new Knight's Landing chip ups the ante with double-precision performance exceeding 3 teraflops and over 8 teraflops of single-precision performance. It also has 16GB of on-package MCDRAM memory, which Intel says is five times more power efficient as GDDR5 and three times as dense.

The new chip is part of Intel's efforts to beef up its Scalable System Framework (SFF) for the high performance computing (HPC) segment.


"We're entering a new era in which supercomputing is being transformed from a tool for a specific problem to a general tool for many," said Charlie Wuischpard, vice president and general manager of HPC Platform Group at Intel. "System-level innovations in processing, memory, software and fabric technologies are enabling system capabilities to be designed and optimized for different usages, from traditional HPC to the emerging world of big data analytics and everything in between. We believe the Intel Scalable System Framework is the path forward for designing and delivering the next generation of systems for the 'HPC everywhere' era."

Pre-production Knight's Landing coprocessors are already being used in several supercomputing class systems. Final versions will be in use by next year.

Via:  Intel
Show comments blog comments powered by Disqus