IBM HAS delivered some significant improvements to its collaborative OpenPower initiative along with its partners, including new technologies that the company said will enable faster and deeper data analysis.
The new offerings bring integration of IBM's open and licensable Power processors with accelerators and dedicated high-performance processors that can be optimised for computationally intensive software code.
The update also provides collaborations and further investments in accelerator-based tools on top of the Power processor architecture. This includes incorporating Nvidia Tesla K80 GPUs to accelerate Watson's Retrieve and Rank API capabilities to 1.7 times normal speed.
"This speed-up can further improve the cost/performance of Watson's cloud-based services. For example, a call centre agent responding to an individual's health and insurance query can leverage Watson's natural language processing technology to obtain an answer in real time even faster than before, potentially increasing customer satisfaction and lowering service costs," explained IBM.
The firm said that the GPU acceleration increases Watson's processing performance by a factor of 10.
IBM has also ported a series of key Internet of Things (IoT), Spark, big data and Power architecture and accelerators.
"The accelerated Power-based offerings come at a time when clients are seeking the ideal server platform for the IoT, machine learning and other cognitive computing applications," added IBM.
"Through an open, collaborative model, IBM and more than 90 members participating in the OpenPower Foundation's Accelerators Working Group are developing and delivering a wide range of accelerator-based solutions."
The update to the OpenPower initiative also brings a multi-year strategic collaboration with Xilinx to jointly develop data centre and network function virtualisation solutions. This will bring together the systems, software and management components around Xilinx FPGA accelerators as they focus on emerging workloads including high-performance computing, cognitive computing, machine learning, genomics and big data analytics.
And that's not all. The OpenPower updates bring a new cluster to the University of Texas in Austin. As a new OpenPower Foundation member, the Texas Advanced Computing Centre will install a Power8-based accelerated computing cluster for academic researchers and developers. IBM said the new cluster is currently running successfully in an early user mode, and will begin accepting requests for access sometime this week.
It was only last month that IBM announced the line of Power processor-based, Linux-tuned machines at LinuxCon Europe.
The Power Systems LC line was introduced by Dr Stefanie Chiras, director and business line executive of IBM scale-out Power Systems, as part of her keynote on the subject of "waitless computing".
The range consists of three products based on Power 8 processors offering scale-out servers in preconfigured, single-click order, and bespoke options to add to the OpenPower servers launched earlier in the year.
The core product is the S812LC, consisting of a single eight-core 3.32GHz processor or 10-core 2.92GHz processor, up to 1TB of system memory and space for up to 14 3.5in drives for internal storage. µ
It's the week in Google
You can probably guess which
GPU is available in Blighty now for £260
Move could bring Halo and Gears of War to the hybrid console