Controlling Costs in an Era of Mushrooming Data Storage

March 6th, 2024 by By Carlos Sandoval, Product Manager for Enterprise Tape at IBM Worldwide

The current era is marked by the creation of more and more data in tandem with the deletion of less and less of what we create. That adds up to a headache for those in charge of storage, and of course, it generally equates to soaring costs. So, what can be done to keep storage costs under control?



Frozen Data Explosion
Let’s do a quick recap on data growth projections. According to Furthur Research forecasts, enterprise data is categorized into multiple layers based on its access frequency:

-Hot data has an access frequency ranging from nanoseconds to milliseconds.
-Warm data is from milliseconds to seconds.
-Cool data is from minutes to 24 hours.
-Cold data goes from days to weeks
-Frozen data is from weeks to years.

The overall amount in all layers will continue to expand in the coming years. However, the data distribution percentages will vary. According to Furthur Research, the amount of data retained in hot and warm layers will shrink from 25% of total storage in 2020 (8% hot and 17% warm) to roughly 20% in 2030 (7% hot and 13% warm). Meanwhile, the cool and cold data layers are expected to remain fairly constant at roughly 20% and 25%, respectively.

The big change, though, shows up sharply in the frozen data layer. The Antarctic ice shelves may be shrinking, but the amount of frozen data to be stored is set for an unprecedented era of expansion. While the overall amount of data stored across all layers will skyrocket between 2020 and 2030, the percentage ending up in the frozen category is expected to grow from 30% to 35%. Unless frozen data costs are effectively contained, enterprises can expect some colossal bills from their cloud providers, endless POs for yet more disk arrays, and soaring costs for power and cooling.

Lifetime Costs of Digital Storage
The best way to understand storage costs is to look at all costs, from acquisition to end-of-life and disposal. According to a 2022 analysis by backup firm Arcserve, the average cost of storing a single terabyte of active file data is $3,351 per year. But that is only the beginning, as storage is just one component of the total cost. This may sound expensive; after accounting for the universal expense of data governance, physical storage is only the beginning as storage is just one component of the total cost. When you take all costs into account – hardware, software, media, floor space, CO2e contribution, command charges, operations and maintenance, egress charges, storage charges, telco, energy, and thermal you come up with the following comparison of lifecycle costs per TB:

-$26.81 for tape
-$108.45 for hard disk drives (HDD)
-$201.98 for cloud

As a result, a typical 20 PB archive of data on tape growing at 5% per year would gain a 75% reduction in total cost of ownership (TCO) compared to archiving data using HDDs. Why? Lifecycle costs for tape hardware, software, and maintenance are considerably lower than HDD or the cloud. Though media costs for tape are a little higher than HDDs, that has little impact on the overall lifecycle gains available through tape.

What this effectively means is that tape is four times more cost-effective for archiving than HDD and 7.5X more cost-effective than Amazon Glacier when you consider the entire lifecycle and all costs.

Tape Storage Opportunity
Long-term storage on tape is forecast to grow to over 1.8 ZB annually by 2030. Tape is predicted to have a significantly higher compound annual growth rate than flash and disk. Furthur Research predicts that HDD shipments will grow 25% annually compared to 37.4% for tape through 2030.

As data continues to grow and storage costs in active tiers remain high, managing data costs is more critical to the bottom line than ever before. Managing cost requires that two areas be carefully considered:  accurate estimates of data growth and mitigating costs with intelligent tiering to place the right data on the right tier at the right time.

By embracing tiered storage, organizations will realize significant savings, especially when a lower-cost storage technology like tape is used. The best way forward is an active archive consisting of a hybrid arrangement of HDD and tape technologies.

You can find out more about emerging use cases for active archives by listening to the recordings of the 2023 Active Archive Virtual Conference.


Alliance Members & Sponsors