IBM innovates tape storage to tame data tsunami

And Big Blue’s Tucson engineers also protect their customers’ data in the ever-changing and increasingly dangerous world of cyberspace.

One example is the recently introduced IBM Diamondback Tape Library, a high-capacity archival tape storage system designed for organizations that need to securely store hundreds of petabytes of data – each equivalent to more than 220,000 DVD movies.

According to IBM, Diamondback’s key benefits include low power consumption and high-capacity storage at about a quarter of the cost of spinning disks.

But its contemporary selling point includes an inherent ability of removable tape media: separating data from the Internet and private networks, which are vulnerable to online attacks.

The ability to physically air-gap data can help protect large users from ransomware — when a hacker blocks access to networks or data and demands a ransom to unlock it — and other cyber threats, says IBM.

Calline Sanchez, IBM’s vice president responsible for developing the Tucson storage system, said the company worked with some of its “hyperscale” customers — companies like cloud computing services that store massive amounts of data — to refine the design to support the Diamondback.

“We decided to really work with our partners, our customers and users and discuss what we want to do from an air gap perspective or a cyber resiliency perspective,” said Sanchez, a UA graduate student who also works at IBMs State Executive for Arizona and her native New Mexico.

With data breaches and ransomware attacks becoming a constant threat, IBM is finding that its big data customers are increasingly turning to tape for data resiliency.

“From an air-gap perspective, it detaches somewhat from the actual baseline infrastructure of a data center so that it’s self-sustaining,” Sanchez said. “It’s really this idea of ​​sustainable block storage of data where it’s not easily accessible from the outside at all.”

While data storage on tape was viewed years ago as a fading technology as faster hard drives and solid-state “flash” storage advanced, sales of tape have skyrocketed amid seemingly never-ending data growth.

Sales of tape media for what is known as “Linear Tape Open” or LTO Ultrium – an industry standard tape technology developed by a consortium of IBM, Hewlett Packard and Quantum Corp. deployed — grew by a record 40% to 148 exabytes of capacity in 2021 (one exabyte is roughly equal to 1000 petabytes, or one billion gigabytes), according to the Silicon Valley-based LTO program.

LTO tape is arguably the cheapest and easiest way to recover from ransomware attacks, said Phil Goodwin, research vice president at IT research firm IDC.

“Ransomware and malware are threats that will not go away,” Goodwin said as part of the LTO program’s sales report. “Magnetic tape is an established, well-known, and proven technology that can be an invaluable tool in combating ransomware.”

Sanchez credits tapes’ resurgence to its sustainability and energy efficiency, as well as its ability to ward off cyber threats.

“It’s less expensive than other storage media like floppy disks or hard drives, as well as flash drives,” she said. “And that’s one of the reasons why tape now primarily feeds the back-end infrastructure of cloud environments.”

Like most IBM data storage systems developed over the past four decades, Diamondback was developed on the company’s multi-building campus at UA Tech Park on South Rita Road.

IBM built what is now UA Tech Park in the early 1980’s to house storage systems design and manufacturing facilities that employed approximately 5,000 people in initially 10 buildings covering more than 1.3 million square feet.

As part of a major reorganization in 1988, IBM moved the manufacturing portion of its operations and nearly 3,000 jobs to San Jose, California, but kept its storage research and development units operational.

The company sold its sprawling campus on South Rita Road to UA in 1994, but remained there as a tenant.

Today, the storage development labs are still in full swing, with approximately 1,000 employees spread over more than 600,000 square feet in four buildings.

IBM’s local operation — which won Large Company Innovator of the Year at last week’s Governor’s Celebration of Innovation Awards — generates more than 400 patents annually, Sanchez said.

At the Tech Park, IBM develops and tests a range of storage systems and media, including disk, tape and flash systems.

At the Tech Park’s reliability testing lab in building 9032, IBM engineers performed final tests on brand-new Diamondback tape library systems, about the size of a slim refrigerator, with palm-sized tape cartridges on aisles of shelves.

With a constant hum of clicks and whirrs, a robotic shuttle moves back and forth in the Diamondback cabinet to pluck and assemble barcode tape cartridges according to a preset program.

Each cartridge can store 18 terabytes, or 1,000 gigabytes, of data, and a full Diamondback packs 1,548 LTO cartridges for a total storage capacity of 27 petabytes, which can be operational within 30 minutes of delivery, said Shawn Nave, an IBM mechanical engineer who helped has developed the Diamondback.

A number of Diamondbacks were tested last week and like all lab tested products they are being put through their paces.

“As part of that testing, we’re going to run it until the wheels fall off and see what happens,” said Nave, one of many UA alumni at IBM’s local operation.

A nearby building houses labs that subject IBM hardware to extreme temperatures and other environmental conditions, Sanchez noted.

As the world generates more and more data, engineers and scientists at IBM’s Tucson site will continue to innovate to keep up, said Sanchez, whose 22 years at IBM includes a year in 2008 for Nick Donofrio, a well-known tech personality who led IBM’s technology and innovation strategies for a decade before retiring in 2008 after 44 years at Big Blue.

Sanchez compares IBM’s storage systems and services to the Wizard of Oz, who hides behind his curtain, remains invisible and sets everything in motion.

“That’s what IBM is offering from a back-end infrastructure perspective — we’re making it easier for these companies because it’s so easy to leverage big data on these critical systems,” she said.

“Where as now in our lab, we’re not just talking about petabytes. We’re talking about how to deploy exabytes (1,000 petabytes each) as well as zettabytes (1,000 exabytes).”

“Because we just know that in a short time, research institutions are going to come up to us and say, ‘Hey, Calline, our two petabytes of research data that we have in all US research institutions, universities, we need you to start building on the base starting from exabytes or possibly zettabytes.’ And to me that’s crazy.”

IBM engineers are working to increase data storage and efficiency using artificial intelligence, which essentially programs machines to “think” like humans, but Sanchez doesn’t like the term.

“I like more ‘augmented intelligence’ for the benefit of humanity, and the great thing about working in data storage is that you keep thinking along those lines,” she said. “Because you have to figure out the physics of how to store an immense amount of data created by people on this earth and with devices like this.”