The annual Fujifilm Recording Media USA (FRMA) Conference in San Diego in late June, 2022 brought together many of the biggest names in IT and cloud storage. What did they all have in common? Heavy usage of tape storage and tape archiving.
“Adding disk to deal with the data explosion makes sense in the short term, but tape is needed at a strategic level due to the volume of data that needs to be moved, stored, and archived securely and cost-effectively,” said Rich Gadomski, Tape Evangelist at FRMA, who emceed the event.
Speakers from the likes of Amazon Web Services (AWS), Microsoft Azure, Meta (Facebook), IBM, and many others extolled the virtues of archiving large quantities of data onto tape.
All except AWS came out openly in favor of tape as the most reliable, cost-efficient, secure, and long-lasting way to archive data. AWS has not publicly commented on the media they use for data storage, but it’s one of the worst-kept industry secrets that they use a LOT of tape.
Others were more forthcoming. “Tape is still the most cost-effective storage medium,” said Nikhil Jain, principal product manager for Microsoft Azure Storage.
Qingzhi Peng, Meta’s Technical Sourcing Manager, noted that his company operates a large tape archive storage team. “Hard disk drive (HDD) technology is like a sprinter, and tape is more like a marathon runner,” said Peng. “We use tape in the cold data tier, flash in the hot tier, and HDD and flash in the warm tier.”
The zettabyte era
Fred Moore, an analyst with Horison Information Strategies gave a detailed rundown of how tape fits in with the zettabyte (ZB) era, which began around 2016. By 2025, he said, about 11.7 ZB will be stored in organizational systems.
A far greater amount of data will be created, but most of it is transient. Moore gave the context of one ZB—the storage equivalent of 7.5 trillion MP3 songs, or 125 million years of one-hour TV shows.
“Object storage is going to be the primary mode of archival storage going forward,” said Moore. “Files won’t go away, but they can’t scale well.”
Moore also commented on the sheer number of data centers operated by hyperscalers. There are more than 600 data centers in the world larger than 100,000 square feet. The biggest is 10 million square feet, and they all store vast amounts of data.
“It is these hyperscalers that are pushing the limits of IT and increasing their usage of tape,” said Moore. “Tape storage is needed to enable growth and control infrastructure costs.”
Further, he recommended tape to combat cybercrime. The air gap (no network connection) provided by tape cartridges plays a central role in any cyber security ecosystem, he added.
Organizations, Moore said, should keep three copies of data on two different media types, with one copy offsite in the cloud or a vault, as well as one copy with an air gap. The latter would primarily be tape cartridges, which require no electronic connection when stored.
“Tape has the lowest acquisition price per TB and the lowest total cost of ownership while using 85 percent less energy than HDDs,” said Moore. “Tape scales without adding energy consumption, whereas HDDs scale capacity by adding drives and raising energy consumption.”
On the last morning of the show, Steffen Hellmold, senior vice president of business development for DNA data storage at Twist Bioscience, revealed an upcoming form of storage: synthetic DNA.
The basic concept is that human DNA is a very efficient storage mechanism for DNA data. By using that same model, synthetic DNA can be used to store data for long-term information retention—very long term according to Helmond.
“Magnetic storage scaling is slowing down and supply can’t keep up with demand,” he said. “DNA data storage is denser by several orders of magnitude.”
But this storage medium is still a long way from prime time. Relatively small amounts of storage have been created using synthetic DNA. It could be several years before we see this technology come of age.
Explore top BI data storage solutions on TechnologyAdvice: 5 Best Data Storage Solutions for BI