The future direction of tape media
We are now eighteen months into the life of sixth-generation LTO-6 Ultrium, the newest version of the most successful tape
format in existence. Over 900,000 HP LTO-6 cartridges have been sold to date, to add to the 68 million HP Metal Particle
cartridges shipped to customers since the format was launched in 2000. For that reason, LTO-6 has already
surpassed its predecessor LTO-5, in terms of adoption rates.
The future role of tape
As we know, data is growing at an unprecedented rate with one statistic saying that 90% of all data ever created was
originated in the past 12 months.
IDC’s Digital Universe survey is estimating a massive 40 zettabytes (40 billion terabytes) of data will be created by 2020,
up from just 3 zettabytes in 2012.
This puts tape and other storage technologies in the forefront for budget and resource planning for companies across
the globe. In a recent study by analysts ESG, 92% of companies indicated they would be increasing or maintaining their
storage and backup budget in 2014. Over 35% identified backup and recovery as a business challenge; 11% said it was
their primary focus.
HP’s Converged Storage vision identifies three broad types of data that need managing and protecting in this rapidly
data for business survival, which tends to have Service Level Agreements associated with it: Recovery Time
Objectives, Recovery Point Objectives and so on.
data that is typically worked on daily.
data, which is data that is not accessed often and may or may not have value but which the business decides,
or is required, to keep.
HP StoreEver Tape is the ideal solution for safe, scalable and low cost archiving of the vast quantities of legacy data
that will be created during the remainder of the decade, all 40 ZB of it! It complements the other pieces of HP’s Storage
portfolio – 3 PAR StoreServ Recovery Manager, StoreOnce De-duplication and HP StoreAll Intelligent Archiving – to address
the challenge of managing data to the end of its life cycle.
The ideal storage medium for the digital universe
With 90% of all data never being accessed again once it’s been archived
, tape represents the ultimate last line of defense
and repository for a business’s deep store of content. Offsite, offline data stored on tape is immune to all threats from
human error, software corruption hackers and viruses. With a shelf life of at least 30 years and minimal costs for power,
cooling and management
, tape remains the supreme archiving technology. And these benefits underpin the impressive
market figures witnessed for LTO-6 Ultrium.
Tape is the cloud – ask Google
In summary, although some commentators speculate as to when tape will be replaced by cloud technology, in HP’s
opinion this is the wrong question. Tape can’t be replaced by cloud because tape is the cloud. Even the most notable
companies associated with cloud services use tape. When Google’s Gmail service was badly disrupted by a corrupt
software patch in 2012, the only way of recovering the accounts of more than 40,000 customers was through backups
stored offline on LTO Ultrium. In the intervening period, and in the years to come, no other technology is emerging that
would provide an alternative to this ‘when all else fails’ scenario.
So what of the most recent tape technology to be launched? Uniquely, LTO-6
is the first tape format to be available in different types: Metal Particle and
Barium Ferrite. HP is the only supplier to sell both types of LTO-6 storage
media and is in the best position to summarize the differences between the
One format, two types of
Until now, 100% of Ultrium data cartridges have been manufactured using
advanced Metal Particle (MP) technology. MP offers robust, tried and tested
data protection and has been the bedrock of successful backup and archiving
for millions of LTO Ultrium customers since 2000. Over 200 million MP data
cartridges have shipped in that time. And since launching LTO-6, during
which time HP has been the market leader for the whole category, 86%
of all HP LTO-6 data cartridges have been of the MP type.
Barium Ferrite (BaFe) is a newer technology that will be used in future generations of LTO Ultrium by some manufacturers.
It was first used for LTO with the launch of LTO-6. Although the BaFe media is manufactured in a similar way to MP tape,
with magnetic particles suspended in a coating on a substrate, BaFe particles are smaller in size than MP particles. That
means they have superior magnetic potential which becomes significant as the areal density – and hence the capacity –
of the media increases. This makes BaFe very well suited to be a candidate for future LTO generations, LTO-7 and LTO-8,
although HP is considering other particle technologies as well.
What is the difference between format and formulation?
It’s important to be mindful that LTO Ultrium is a very tightly defined storage format. The three inventors, or Technology
Provider Companies, of LTO (HP, IBM and Quantum) define the standards and specifications that all companies, both
hardware and media, must abide by. This is at the heart of the open standard philosophy behind LTO Ultrium format: that
products manufactured or supplied by different manufacturers are guaranteed to work together without any compatibility
or interchangeability issues. Only HP, IBM and Quantum can define the future roadmap of LTO Ultrium. All other companies,
like media vendors or other hardware OEMs, are licensees and must follow the path prescribed by the TPCs.
Neither MP nor BaFe are formats, but rather they are two different materials used to manufacture media: providing
technical solutions to achieve the same goal, namely compliance with the LTO Ultrium specification itself. In much the
same way as nylon and cotton are two perfectly acceptable materials to use for making a shirt, so long as the materials
meet the specification, they can be sold as an LTO product.
The benefit of open standards
This example highlights one very important distinction that has been crucial to the success of LTO compared to older
legacy formats like DLTtape and Advanced Intelligent Tape (AIT). The benefit of an open standard is that it gives companies the
freedom to approach the format from multiple directions, based on their technical skills, target audiences, cost structure
and other factors. These are the bedrocks of competition that have enabled LTO Ultrium to become the pre-eminent
solution in backup and archiving. Ultrium is not, and never will be, proprietary technology and not even the TPCs themselves
can change the format to suit their own expertise or bias. Put simply, HP, IBM and Quantum must follow the rules just like
What’s in a name? Nothing unless it meets the specification
Naturally, in an open market, one expects competition and the intervening years have seen very healthy and vigorous
rivalry between some of the world’s leading data storage companies. Brand names like Nanocubic, Neosmart, Dternity and
StoreEver are familiar to IT decision makers. But customers should never lose sight of the fact that ultimately, all these
terms are marketing names to identify a particular vendor’s solution for meeting the LTO Ultrium specification.
In the world of LTO tape media, MP and BaFe are like the nylon or cotton of the earlier example. They enable companies to
fulfill the LTO-6 format and have been marketed using a variety of brand names. But underneath the marketing, they are
all LTO-6 format compliant products.
So which is the better tape? Neither (and here’s the proof!)
The reason why both exist is because different vendors chose alternative paths to achieving compliance. Unlike every other
tape company, HP chose to sell both types of media because we recognized that some customers would choose MP and
some would prefer BaFe. And based on our exhaustive testing, for LTO-6 users, neither version offers any advantages
in terms of performance, reliability, durability or cost.
As an example of just one of those tests, HP conducts Full Volume Life Testing in two different environments to examine
and ensure the performance of both MP and BaFe tapes across a broad spectrum of conditions. Both ambient and
29°C/84.2°F 80% R.H. (hot/wet) conditions are used.
Although HP is aware that some vendors have positioned BaFe as a more desirable technology for LTO-6, our test data –
extensively published in our white paper ‘Best of both worlds’ – shows there is no performance difference when using
either media type. User capacity and transfer rates were consistently excellent throughout the tests (ie Extended Green Tape
Tests that completed using 1,500 cartridges of MP and 1,500 cartridges of BaFe media). There is no hard data that proves
customers using MP media are disadvantaged in drive performance if they choose MP LTO-6 media rather than BaFe
media. The archival life of their data will not be different (See ‘The benefits of tape for archiving’ white paper). Users in the
installed base continue to rely on MP technology because they have LTO-5 and LTO-4 cartridges that need to be used
inside their LTO-6 devices.
Even before LTO-6 was launched, HP had already completed over half a million hours of testing for MP and BaFe media,
pulling over 500,000 miles of tape.
In 2013, we conducted over 2.6 million hours of quality control testing for HP LTO Ultrium branded media.
In a typical year, we use over 15,000 drives and 8,500 pieces of media.
In 2013, we pulled over 200,000 miles of MP and BaFe LTO tape through HP and non-HP devices.
In 2013, we pulled over 100 million yards of MP tape to demonstrate the reliability of the GMR heads in HP StoreEver
In summary, the reasons for promoting BaFe as better, or even as being somehow fundamental to the LTO-6 format, are
entirely down to branding and marketing and not in any way related to the TPC specification or to a customer’s likely
HP feels able to make this distinction because we are one of the TPCs; we have sold more LTO-6 tapes and media than
any other company; and, to re-iterate, 86% of HP LTO-6 media to date has been of the MP variant.
LTO-7 and beyond. What next for tape?
But what of the future? Will we see a continued divergence between MP and BaFe when LTO-7 products begin to appear?
The straightforward answer to that question from HP’s perspective is ‘no’. To meet the extended capacities set out by the LTO
roadmap, 16 TB for LTO-7 and 32 TB for LTO-8, it will be necessary to reduce the size of the particles used to manufacture
the recording surface of the tape. Shrinking MP particles further isn’t the right way forward because they would lose their
coercivity and magnetic strength, which could ultimately lead to data corruption.
Tape drives work by magnetically changing the position of the particles in a pattern that corresponds to computer bytes;
the north or south polarities of the particles represent the 1’s and 0’s of binary data. In simple terms, therefore, the particle
count is a cornerstone of how much data can actually be stored on a tape. Increasing their quantity has a direct bearing
on capacity. The more particles you can change, the more bytes you can store. This is known as areal density.
But packing more particles into a physical space that is defined by the LTO format specification brings technical challenges.
Because of the need for backwards compatibility, the tape cartridge cannot increase in size and this in turn limits the amount
of tape that can be wound onto the reel inside the shell. One method for increasing capacity is to make the particles
smaller, which in turn requires ingredients with higher levels of coercivity. BaFe is one such substance and the one with
the greatest public profile at the time of writing, but HP is investigating others that may be better suited to even
Coercivity – and why it’s important
So what is coercivity and why does it matter? Coercivity is the resistance of the particles to having their position changed.
A higher coercivity is needed for higher capacity tapes because as the bits get more tightly packed, the magnetic forces
between them become more intense, which can lead to their individual signal strength becoming weak and blurry.
Because it is an oxide, BaFe and similar compounds are less susceptible to the demagnetization that can occur between
these densely arranged particles. This ensures that data is not only saved with a low and reliable error rate, but also that
once it has been captured, it stays captured, exactly as it was written.
Compared to MP, BaFe media has better magnetic properties, which means improved Signal to Noise Ratio (SNR) and Bit
Error Rate (BER). Although this is irrelevant to customers buying LTO-6 today, for LTO-7 and beyond, compounds like BaFe
will be a more effective and reliable choice for achieving the specifications of next generation tape archiving. Particles
being developed for LTO-7 and LTO-8 will need to have significantly smaller volume and higher coercivity for MP to
be a viable option.
In conclusion, therefore, tape demand is extremely robust. LTO-6 has ramped faster than its predecessor and the amount
of capacity being shipped on tape products is at an all time high.
As the market leader for branded LTO tape drives, low and mid-range automation, as well as storage media, HP has put
its StoreEver portfolio at the heart of its vision for Converged Storage. This is the infrastructure that companies are
migrating to in order to manage the demands of an explosion of data.
Although MP technology has been an outstanding solution for companies seeking the most reliable backup and archiving
products, LTO-6 will be its swansong. But it is going out on a high note, with performance and reliability metrics that
deliver the same outstanding customer performance as the newer, emerging particle technologies.
Although some companies are already promoting their BaFe products as the future direction of the LTO format, in HP’s
opinion, it’s too early to say with certainty that this will be the only choice. Although there will likely be BaFe LTO-7 and
LTO-8 tapes, other technologies may emerge to compete against BaFe, giving customers the same choice for future
versions of the LTO format as they had between MP and BaFe during the LTO-6 era.
How are decisions about the future of LTO made?
What is clear is that the TPCs have defined the specifications of the LTO Ultrium format for the licensees, such as media
companies, to follow. And nowhere in the LTO specification does it stipulate the use of one media formulation versus
another. So while it is highly unlikely that anyone will be able to purchase an LTO-7 tape cartridge made from cotton
or nylon, it’s entirely possible there will be other choices beside BaFe!
Backwards compatibility guaranteed regardless of media type
Also, it is important to remember that since HP and the other TPCs control the format, backwards compatibility (one gen
write, two gen read) is assured for investment protection. The material used to make the media has no bearing upon this
and no media company or single supplier can ‘break’ backwards compatibility if they still want to use the LTO logo. That
means LTO-7 and LTO-8 drives will need to work with the millions of MP LTO-5 and LTO-6 tapes that will still exist in the
future when new models come to market.
In HP’s opinion, as the market leader in LTO Ultrium and as a TPC, this freedom at the heart of the format is not just good for
innovation and good for tape, it’s squarely in the best interests of the customer. The future is bright; the future is tape.