Power servers

With IBM POWER9, we’re all riding the AI wave

Share this post:

There’s a big connection between my love for water sports and hardware design — both involve observing waves and planning several moves ahead. Four years ago, when we started sketching the POWER9™ chip from scratch, we saw an upsurge of modern workloads driven by artificial intelligence and massive data sets.

We are now ready to ride this new tide of computing with POWER9. It is a transformational architecture and an evolutionary shift from the archaic ways of computing promoted by x86. POWER9 is loaded with industry-leading new technologies designed for AI to thrive. The architecture is expected to shake the foundation of computing across a wide range of industries, and transform data from a liability into a weapon.

POWER9 is looking great, and we think it will blow you away. We have systems booted and functioning and delivery is on track. The crown jewel for IBM and its OpenPOWER members is the U.S. Department of Energy’s Summit and Sierra supercomputers, which are equipped with POWER9 processors and NVIDIA’s newest Volta-based Tesla GPU accelerators. They will push the frontiers of scientific computing, modeling and simulation. Early benchmarking shows they could be two of the world’s fastest supercomputers when they becomes fully operational in 2018.

Powering the era of AI with POWER9

With POWER9, we’re moving to a new off-chip era, with advanced accelerators like GPUs and FPGAs driving modern workloads, including AI.  We see that the era of the on-chip microprocessor — with processing integrated on one chip — is dying as well as  Moore’s Law lapsing.  POWER9 gives us an opportunity to try new architectural designs to push computing beyond today’s limits by maximizing data bandwidth across the system stack.

In a way, POWER9 is a modern tribute to classic computer designs, where a bunch of chips interacted to handle CPU, graphics and floating-point operations. The bedrock of POWER9 is an internal “information superhighway” that decouples processing and empowers advanced accelerators to digest and analyze massive data sets. The result? You are hammered nonstop with amazing performance.

POWER9 will be the first commercial platform loaded with on-chip support for NVIDIA’s next-generation NVLink, OpenCAPI 3.0 and PCI-Express 4.0. These technologies provide a giant hose to transfer data.

The next-generation NVIDIA NVLink and OpenCAPI provide significantly faster performance for attached GPUs than the PCI-Express 3.0 in x86 systems. The PCI-Express 4.0 interconnect will be twice the speed of PCI-Express 3.0, a technology that is more than a decade old and to which current x86s are committed.

The open POWER9 architecture is engineered to allow customers to mix and match a wide range of accelerators to meet their AI computing needs. OpenCAPI is expected to unlock coherent FPGAs to support a wide range of accelerated storage, compute and networking workloads. In a world where choice is king and computing is more open, we are against locking users into proprietary systems. We anticipate that a significant number of the 300+ members of the OpenPOWER Foundation and OpenCAPI Consortium will also see the benefits and launch products and solutions for POWER9.

Accelerating the future with POWER9

The boundary-pushing bandwidth and performance capabilities of POWER9 are poised to provide scientists with revolutionary ways to interpret data. The last era of Top500 was characterized by benchmarks like LINPACK, but IBM’s AI architecture — consisting of POWER9 and PowerAI software – has the potential to fundamentally change supercomputing with cognitive methods for optimized performance and improved insight.

IBM is heavily invested in cognitive systems, and IBM Research labs are developing a plethora of amazing technologies for POWER9. IBM researchers have already cut deep learning times from weeks to hours with the PowerAI Distributed Deep Learning Toolkit, and those offerings will only get better.

We’re excited to bring you POWER9 systems, and announcements are just around the corner. Stay tuned. And for more, be sure to join our webinar to take a look behind the scenes of the CORAL systems we’re building with POWER9.

Statements of direction represent IBM’s current intent, are subject to change or withdrawal, and represents only goals and objectives.

Add Comment
No Comments

Leave a Reply

Your email address will not be published.Required fields are marked *

More Workload & resource optimization Stories

Linux on IBM Power Systems: Support for cognitive computing

Throughout its more than 100 years of history, IBM has continuously reinvented itself to adapt to an ever-changing world. Likewise, the IBM Power Systems architecture has evolved as enterprise-class systems that are core to our business. Over the last decade, many organizations have consolidated workloads using Virtual I/O Server (VIOS) technology, and IBM Systems Lab […]

Continue reading

Reaching the Summit: The next milestone for HPC

The high-performance computing landscape is evolving at a furious pace that some are describing as an important inflection point, as Moore’s Law delivers diminishing returns while performance demands increase. Leaders of organizations are grappling with how to embrace recent system-level innovations like acceleration, while simultaneously being challenged to incorporate analytics into their HPC workloads. On […]

Continue reading

IBM Power Systems offers choice for SAP HANA with RedHat

In just a little over two years nearly 1000 clients have chosen the IBM Power Systems  platform for their SAP HANA workloads. Why are they choosing the Power Systems platform? We believe it’s due to flexibility and choice.  The flexibility to virtualize up to 8 production SAP HANA LPARs (logical partitions), to run mixed workloads, […]

Continue reading