Amazon Web Services (AWS) uses its annual Re:Invent summit to announce its latest products and services, to the world. This year, the big game-changer was staring its 50,000 attendees, right in the face, when AWS CEO, Andy Jassy, raised the curtain on ‘Glacier Deep Archive’.

Scheduled for launch in 2019, Glacier Deep Archive signals Amazon’s sustained aggression in the cloud market, where the company has staked a 51.8 percent claim in market share.

“It is by far the lowest-cost storage available on the cloud today,” said Jassy, at the launch. “It comes to less than one-tenth of a cent per gigabyte, per month, which pretty much means it’s only a dollar per terabyte, per month.”

What this means is the startlingly simple fact that cloud will now be an undisputed disruptor to the present-day practice of storing voluminous data on tapes. It also means that AWS is leaving no stone unturned in consolidating its position as the leading provider of cloud solutions, with Microsoft (13.30 percent), Alibaba (4.60 percent) and Google (3.30 percent) a far cry from AWS’ 51.80 percent market share. Jassy also reported annualized revenue of 27 billion USD for AWS, with a 46 percent year-on-year growth.

Cheaper Machine-Learning

Disruptive cloud solutions though, wasn’t the only launch at Re:Invent. AWS also launched ‘Amazon Elastic Inference’, in an attempt to help companies cut down inference costs, which Jassy said accounted for a whopping 90 percent of costs attached to machine learning.

“Running inference in production drives the majority of cost for machine learning,” he said. “We train Alexa a few times per week, but think about how many answers are being inferred every minute. Nearly 90 percent of machine learning costs are in inference.”

Autonomous Mobility

Re-Invent 2018 also saw Amazon launch its first-ever fully autonomous mobility device. ‘AWS Deep Racer’ is no larger than a toy car, and yet is fully autonomous in that its mobility comes by way of machine learning.

Running on a Dual-core Intel Atom Processor, DeepRacer comes equipped with a HD camera, an all-wheel drive, suspension, two batteries, an accelerometer and a gyroscope for orientation. The car literally “learns” to drive on its own, using an AI system that “rewards” it for smooth movements, which is a process of inherent machine-learning.

AWS DeepRacer is priced at $399, but is being offered for a limited-period offer of just $249, on the Amazon marketplace.

Have you signed up for Primo, our daily newsletter? It has all the stories and data on the market, business, economy and tech that you need to know.