IBM plans Asian rollout for POWER-powered cloud

Source: itnews

IBM plans Asian rollout for POWER-powered cloud

AIX users score a cloudy upgrade path as Big Blue takes on AWS and pals with AI-tuned services.

IBM’s cloud will soon offers more IaaS and AI services powered by its own POWER9 CPUs.

Big Blue has previously offered POWER CPUs for infrastructure-as-a-service, but limited their availability to just one data center in Dallas, Texas, and focused on a rent-a-server option, plus a POWER-powered AI service offered by partner Nimbix.

But at the company’s Think conference in San Francisco last week, IBM announced the debut of its POWER9 platform with a new “Virtual Server on IBM Cloud—available in select IBM Cloud data centers [northern] spring 2019” plus the intention to run its own POWER AI service.

As IBM had previously only offered POWER in its Dallas data centres and not flagged the use of others, iTnews inquired which data centers would get the new offering and was told the platform will land “in Dallas and Washington Data centres and then expanding to Europe and Asia later this year.”

“The time frame for global roll out is being finalized and we will be working to release more widely later this year (second half of the year).”

This news will be welcome by users of POWER systems. IBM continues to evolve the platform and wins greenfield sales for POWER, which has some advantages over x86 in some roles. But the bulk of POWER users run legacy applications and have just Big Blue to deal with. Third-party clouds have shown little interest in POWER-as-a-service, leaving such users with few options other than trying to negotiate hard on whatever prices IBM offers.

IBM expanding the footprint of its POWER-as-a-service footprint will therefore be very welcome news, as it represents a cloudy opex option IBM has previously not offered.

News that “The IBM Cloud and IBM Power Systems teams are working hand-in-hand to deploy the raw performance of POWER9-based virtual servers with NVIDIA V100 GPUs on Linux for machine learning and artificial intelligence workloads” is also interesting. Teaming with NVIDIA suggests IBM has taken note of specialist AI clouds from the likes of AWS, Google and Microsoft, and decided it needs a customized offering too.