TSI’s Take: “This is AI” on the Discovery Channel

I wonder how many readers of “The NonStop Insider” caught the premier of “This is AI” on the Discovery Channel June 22. I found it to be a thoughtful discussion of the topic of Artificial Intelligence. The documentary avoided dwelling on the “Hollywood-style” negative or malevolent aspects of AI, as portrayed in movies such as “2001-A Space Odyssey”, where HAL takes over the ship and eventually must be reduced to a childlike intellect singing “Mary had a Little Lamb”; or 1970’s “The Forbin Project”, featuring two supercomputers, Colossus and Guardian (NonStop old timers take note of that name), one American and one Soviet, linking up and ruling the world through assured nuclear destruction, or Arnold Schwarzenegger’s portrayal of “The Terminator”. As well, the future of AI was not caricatured, with the likes of “Robby the Robot” from “Forbidden Planet”, or the robot from “Lost in Space” (warning Will Robinson!).

Instead, a reasonable and rational approach to where AI is going was presented. Highlighted were innovations in health care (allowing a blind individual to run), educational advancements in teaching via AI, and of course a discussion of driverless cars. Of particular interest was IBM’s Deep Blue and the famous Chess Game against Garry Kasparov, Watson’s 2011 win on Jeopardy!  and Google’s Deep Mind machine learning approach to playing and mastering the very complex Chinese game “Go”.

But what’s all this got to do with NonStop?

In a word “Data” (and not the android from Star Trek!).  Artificial Intelligence and especially machine learning requires huge amounts of stored data. The documentary made this abundantly clear, and even mentioned Cloud Object Storage as the probable repository for massive databases.

There is no debate, the amount of digital information continues to grow, and the rate of growth is rapidly accelerating! In fact, IDC calculated the total amount of digital information created and replicated at 4.4 zettabytes (one zettabyte is equal to 44 trillion gigabytes), way back in 2013. They figure this number will grow to something like 44 zettabytes by 2020, only two years from now. In another IDC study, commissioned by Seagate, “the global datasphere will grow to 163 zettabytes by 2025”.Another startling fact, over 90% of all existent data was created in the last 2 years!

And while most of this data is created by individuals, enterprises handle about 85% of this information at some point in its lifecycle.This means organizations take on the responsibility for architecting, delivering and maintaining information technology systems and data storage systems to meet this demand. Clearly, NonStop is not and will not be exempt from managing gigantic databases!

Obviously massive amounts of data must and are being archived!

How will NonStop servers interface with Cloud Object Storage directly?

By investing in Tributary System’s Cloud Object Storage Solution, that’s how. As the gateway to cloud storage (public, private or hybrid) for NonStop, TSI’s solution employs advanced COS technology coupled with Tributary’s proven Storage Director®as the “front end”, NonStop customers can transparently take advantage of IBM’s Cloud Object Storage®(COS), formerly known as Cleversafe®, without any changes to their NonStop applications. In addition, Tributary’s software-defined Storage Director high performance front end ingestion and caching solution is now fully certified and approved for use with Hitachi Content Platform (HCP).

Storage Director, is a policy-based, tiered, and virtualized software solution especially designed for backup which can be seamlessly integrated with any media, including tapes, disk drives, virtual environments and NonStop or other proprietary server types, plus open systems. Storage Director can group data into different pools and apply different protection policies at different times across any storage medium simultaneously. Tributary Systems has gained a massive strategic edge as it has entered into synergistic partnerships with both IBM and Hitachi to revolutionize enterprise cloud data backup/restore, archive and DR. Combining the capabilities of Storage Director while endorsing long-term archival to Object Storage is where Tributary believes the data backup and retention market evolving, and sees no evidence to the contrary.

Tributary is the only company in the marketplace that can backup all NonStop mission-critical servers (legacy or current) using a single solution. In addition to Storage Director’s AES 256-bit encryption, data is also erasure-coded in the storage tier.

From a performance standpoint, Tributary’s solution can ingest data at a rate of 12TB per hour and restore at about 8.5TB per hour. Should a flash storage be used in the cache layer, the ingestion rate goes up to 37TB per hour and restores at 35TB per hour; these performance metrics are unmatched in the market.

Tributary’s IP, when combined with leading Cloud Object Storage solutions imparts exclusive cutting-edge data storage and management capabilities that can be extended well beyond public cloud models—into hybrid and on-premise environments—and also offers double-layered security for NonStop clients.

Importantly, the Storage Director Cloud Object Storage solution can be implemented with a monthly fixed cost model unlike all public cloud providers such as AWS who charge customers for accessing their data through additional fees for “puts and gets.” Employing a cloud backup solution where costs vary widely from month to month, based on access to their own data, is challenging for most enterprise customers.

Anyway, that’s our take this month, and by the time you read this many of the valued NonStop Partners will have met in Palo Alto for the 2018 Partner Symposium. Hope to see many old friends this July 9th………..

By Glenn Garrahan, Director HPE Business

www.tributary.com