IBM introduced Linux-based servers based on its new POWER8 Chip and innovations from the OpenPOWER community.
The new servers join IBM’s Power Systems LC lineup and aim to outperform x86-based servers on a variety of data-intensive workloads.
The servers leverage the IBM POWER8 silicon with NVIDIA NVLink, which is a high-speed, energy-efficient bidirectional interconnect that directly connects the new IBM POWER8 processor with NVIDIA Tesla P100 Pascal GPUs. NVIDIA NVLink is embedded at the silicon level and incorporated into the overall system design.
IBM said this tight coupling of IBM and NVIDIA technology enables data to flow 5x faster than on an x86-based system.
Tencent is testing the servers. The U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) and Lawrence Livermore National Laboratory (LLNL) have also been early testers and the systems will serve as an early-generation test bed for developing demanding applications for Summit and Sierra, the next generation supercomputers that IBM expects to deliver respectively to ORNL and LLNL in 2017.
“The user insights and the business value you can deliver with advanced analytics, machine learning and artificial intelligence is increasingly gated by performance. Accelerated computing that can really drive big data workloads will become foundational in the cognitive era,” said Doug Balog, General Manager of POWER, IBM Systems. “Based on OpenPOWER innovations from partners such as NVIDIA, our new OpenPOWER Linux servers with POWERAccel set a new standard for these workloads compared with x86 processor-based servers.”
http://www-03.ibm.com/press/us/en/pressrelease/50479.wss
OpenPOWER Advances in HyperScale Data Center Race
Since its founding two years ago, the OpenPOWER Foundation, which is an open development alliance based on IBM’s POWER microprocessor architecture, has grown to more than 200 participating companies and organizations. The goal is to build advanced server, networking, storage and GPU-acceleration technology for next-generation, hyperscale data centers.
At the second annual OpenPOWER Summit held in San Jose this week, more than 50 new infrastructure and software innovations, spanning the entire system stack, including systems, boards, cards and accelerators are showcased.
Some highlights:
- New Servers for High Performance Computing and Cloud Deployments – Foundation members introduced more than 10 new OpenPOWER servers, offering expanded services for high performance computing and server virtualization.
- Google is developing a next-generation OpenPOWER and Open Compute Project form factor server. Google is working with Rackspace to co-develop an open server specification based on the new POWER9 architecture, and the two companies will submit a candidate server design to the Open Compute Project.
- Rackspace announced that “Barreleye” has moved from the lab to the data center. Rackspace anticipates “Barreleye” will move into broader availability throughout the rest of the year, with the first applications on the Rackspace Public Cloud powered by OpenStack. Rackspace and IBM collectively contributed the “Barreleye” specifications to the Open Compute Project in January 2016.
- IBM, in collaboration with NVIDIA and Wistron, plans to release its second-generation OpenPOWER server,which includes support for the NVIDIA Tesla Accelerated Computing platform. The server will leverage POWER8 processors connected directly to the new NVIDIA Tesla P100 GPU accelerators via the NVIDIA NVLink high-speed interconnect technology. Early systems will be available in Q4 2016. Additionally, IBM and NVIDIA plan to create global acceleration labs to help developers and ISVs port applications on the POWER8 and NVIDIA NVLink-based platform.
- Expanded use of CAPI for Acceleration Technology – Foundation members, including Bittware, IBM, Mellanox and Xilinx, unveiled more than a dozen new accelerator solutions based on the Coherent Accelerator Processor Interface (CAPI). Alpha Data also unveiled a Xilinx FPGA-based CAPI hardware card at the Summit. These new accelerator technologies leverage CAPI to provide performance, cost and power benefits when compared to application programs running on a core or custom acceleration implementation attached via non-coherent interfaces. This is a key differentiator in building infrastructure to accelerate computation of big data and analytics workloads on the POWER architecture.