IBM targets AI workloads with POWER9 systems; claims to be faster than x86

Soma Tah

Speed to insight is going to emerge as the key competitive differentiator for businesses, as they start stepping into the era of compute-and-speed-hungry artificial intelligence(AI), and deep learning workloads.

IBM, recently announced a new line of accelerated IBM Power Systems Servers, keeping this new requirement of businesses in mind. The systems are built on its new POWER9 processor, which reduces the training times of deep learning frameworks significantly from days to hours and allows building more accurate AI applications in considerably less time.

“The era of AI demands a tremendous amount of processing power at unprecedented speed,” said Monica Aggarwal, Vice President, IBM India Systems Development Lab. “To meet the demands of the cognitive workload, businesses need to change everything right from the start- the algorithms, the software, and the hardware as well. POWER9 systems bring an integrated AI platform designed to accelerate machine learning and deep learning with both software and hardware that are optimized to work together.”

The POWER9 systems were designed to drive considerable performance improvements across popular AI frameworks such as Chainer, TensorFlow, and Caffe, as well as accelerated databases such as Kinetica.

Unlike x86, the open architecture also allows customers to mix and match a wide range of accelerators to meet their AI computing needs, said Aggarwal.

The new POWER9 processor delivers nearly 4x deep-learning framework performance over x86, said IBM, which exited its x86 business in 2014. The combination of NVIDIA NVLink, PCI Express 4.0 and OpenCapi 3.0 can accelerate data movement considerably, calculated at 9.5x faster than PCI-E 3.0 based x86 systems.

Building and training deep learning models will become easier on POWER9 systems, said Anand Haridass, Chief Engineer - Cognitive Systems Development at IBM India.

The AI-optimized POWER9 systems will give a boost to the modern AI, high performance computing(HPC) and accelerated database workloads and support new use cases across government HPC labs, and different industry verticals such as BFSI, telecom, retail, automotive, life sciences and healthcare, etc.

“For example, detecting fraudulent transaction or spotting an anomaly in the spending pattern will be easier if banks can quickly analyze the users’ location data, ATM and online transactions data over a certain period of time. Similarly, analyzing the tones of the customer interactions, businesses can detect the mood of the customers during a helpdesk interaction and address the issues in best possible manner,” explained Haridass. The systems will also help in harnessing image recognition capabilities and targeted marketing campaigns, said he.

POWER9 has already been the building block of the US Department of Energy’s “Summit” and “Sierra” supercomputers which are touted as soon-to-be most powerful data-intensive supercomputers in the world.

On the other hand, Google, being a founding member of OpenPOWER Foundation, is also building a POWER9 server called Zaius for use in its data centres.