Intel (NASDAQ:INTC) has shed further details on its second-generation Xeon Phi CPU, known as Knights Landing.

The processor features several technical achievements, starting with a 14nm manufacturing process, which is a first in this series. Designed to offer high-performance computing, Knights Landing differs from other server-based CPUs in that it uses lots of low-energy cores to run parallel tasks, whereas offerings from IBM (NYSE:IBM) or Oracle (NYSE:ORCL) use fewer but more powerful cores.

Built on Intel’s MIC (Many Integrated Core) architecture with a total of 8 billion transistors, Knights Landing runs a modified version of the Atom Silvermont x86 core in a tile configuration, with a single tile featuring two cores and vector execution units along with shared L2 cache as well as a circuitry that connects the tile to the rest of the mesh network. Intel has mentioned that each Knights Landing package would include a processor with 30 or more tiles and eight on-chip memory modules. Another major highlight with Knights Landing is that it would be able to function as a host processor, meaning that it can boot and run x86 operating systems and application code without any need for recompilation. It can also act as a co-processor.

Talking about memory, the chip vendor has announced that Knights Landing would feature eight 2GB stacks of memory, totaling up to 16GB. The chip is manufactured at Micron, and looks to be a variant of the manufacturer’s Hybrid Memory Cube, which involves stacking memory and using an embedded logic chip to deliver higher bandwidth at a lower power. Micron has mentioned that its HMC modules will be able to transfer data 15 times faster than a standard DDR3 module, while utilizing 70% less energy. Along with on-chip memory, Knights Landing will come with six memory channels that can connect a total of 384GB DDR4 memory.

The result of the new manufacturing process, core design and memory is that Knights Landing will offer three times the performance as the current-gen Knights Corner, with Intel claiming 3 teraflops double-precision and 6 teraflops single-precision performance. That number is close to the 7 teraflops figure Nvidia (NASDAQ:INTC) touted during the launch of its latest video card, the Titan X.

It is no wonder, then, that Intel is aiming for the same use-cases as Nvidia for Knights Landing, with the chip vendor stating that the CPU can be used for deep learning and data analytics. Nvidia, however, has invested significant resources in its platform, and is offering tools such as the Digits software framework. Even if Intel does not manage to successfully challenge Nvidia in the Knights Landing, it is witnessing a great amount of demand, with over 50 companies set to sell server systems with the CPU as the host.

The post Intel Gunning To Challenge Nvidia At HPC With ‘Knights Landing’ Xeon Phi Processor appeared first on VR World.

Today, Intel made two fairly large announcements simultaneously with the announcement of their Omni Scale Fabric and the integration of it into their next generation of Xeon Phi chips. Additionally, Intel has worked with Micron to enhance Knight’s Landing with high-performance on-package memory. This on-package memory is also known as Hybrid Memory Cube or HMC.

The Knight’s Landing next generation Xeon Phi product announced today will use Intel’s Silvermont CPU architecture which is modified (or as Intel says, enhanced) for HPC. The expectation that Intel is setting is that these cores will deliver three times the single threaded performance of the previous generation and still be Intel Xeon Binary compatible.

Knight’s Landing – MCDRAM

As for the on-package memory itself, the HMC, it will support up to 16GB at launch while only taking up 1/3 the space (compared to Knight’s Landing vs Knight’s Corner) compared to GDDR5. They are also claiming five times the bandwidth compared to DDR4 using the same amount of memory. However, to be fair, this 5x bandwidth comparison versus DDR4 isn’t necessarily a fair one since DDR4 is still in its infancy and a more appropriate comparison would be DDR5 since DDR3 in servers is fairly slow. Intel and Micron are also claiming 5x the power efficiency when compared to GDDR5, based upon comparisons between Knight’s Landing and Knight’s Corner (the previous generation). But do keep in mind that Micron says they will only be shipping 2GB and 4GB parts this year, making the 16GB parts for Knight’s Landing 2015 parts.

Knight’s Landing – MCDRAM

The Omni Scale Fabric from Intel is designed to deliver maximum bandwidth and scalability between Intel’s future Xeon and Xeon Phi products. With interoperability between Knight’s Landing and Xeon processors coming with the 14nm generation of Xeon server processors. This Omni Scale Fabric will have PCIe adapaters, edge switches, director systems, Intel’s own silicon photonics and open software tools. All of these are designed to make the upgrade to Intel’s Omni Scale Fabric less painful.

Omni Scale Fabric

In terms of performance, Knight’s Landing will deliver “3 TFLOPS of peak theoretical double-precision performance” which is based on preliminary and “expectaionts of cores, clock frequency and floating point operations per cycle.” Which really means that this hasn’t really been benchmarked exactly quite yet and we will have to see what the end product delivers when Intel actually starts shipping commercial systems in the second half of 2015.

There is also already an operational supercomputer using Knight’s Landing in their system and that’s NERSC’s Cori supercomputer which already employs Knight’s Landing and has already benchmarked the performance at over 3 TFLOPs per Knight’s Landing node.

NERSC

The post Intel's New Knight's Landing Xeon Phi Combines Omni Scale Fabric with HMC appeared first on VR World.