The “gray ceramic” storage medium: a pioneering innovation in data preservation (Courtesy of Cerabyte).
A German startup, Cerabyte, is poised to disrupt the growing market for storing secondary and archival data by leveraging a groundbreaking approach: utilizing lasers to engrave data onto glass substrates coated with ceramic material. According to the company, its “gray ceramic” expertise provides a cost-effective and durable storage solution that is essentially perpetual, following the recent establishment of two US offices.
“As a tidal wave of data looms on the horizon, innovative solutions for knowledge storage are urgently needed to meet the pressing demands of scalability and affordability.” “Our vision is to achieve a price point of $1 per petabyte per month, representing a 1000-fold reduction in costs within the next 20 years.”
According to Fred Moore, founder of Horizon LLC, writing on behalf of Cerabyte, laborious-disk drives (HDDs) currently store approximately 85% of the world’s digital knowledge and account for around 60% of secondary storage capacity, with data persisting on HDDs. As the COVID-19 pandemic’s brief surplus of HDDs momentarily drove down HDD-based storage prices, this phenomenon is unlikely to persist, leading to a renewed need for more affordable and reliable storage solutions to cope with the ongoing deluge of digital data.
According to Moore, present knowledge development projections suggest a doubling of stored information every three years, with a predicted annual growth rate of approximately 25%. As traditional storage options continue to evolve, HDDs and magnetic tape are poised to engage in an ongoing struggle for dominance as the go-to solution for large-scale secondary storage, each offering its own unique strengths but neither truly surpassing the other.
In 2000, Cerabyte was founded with the goal of creating a more financially viable and durable storage solution than traditional tape-based systems. Munich-based company has successfully patented their innovative technology, leveraging a laser to inscribe encoded binary data onto ultra-thin glass sheets, precisely coated with a microscopic layer of ceramic measuring just 500-1,000 atomic layers in thickness. The corporation leveraged exclusively best-of-breed solutions to scale its platform.
Our company excels in the development of Ceramic Nano Reminiscence technology.
The writing course employs ultra-short laser pulses in conjunction with off-the-shelf digital mirror units, commonly found in video projectors and head-up displays. The combination of the laser with a digital micromirror device (DMD) yields a laser beam matrix that thoroughly vaporizes the ceramic nanolayer, enabling the writing of up to two million bits per pulse in parallel, with extremely high repetition rates operating within the kilohertz range. This enables potential writing speeds of over 1 gigabyte per second, leveraging mere 1 watt of power – a remarkable three to four times faster than current LTO tape and HDD technologies.
While pursuing the study program, the same DMD machine is utilized in conjunction with a high-resolution image sensor capable of analyzing more than 500 frames per second. The information is decoded by leveraging a reconfigurable digital circuit known as a Subject Programmable Gate Array (FPGA).
As soon as manufactured, the glass-ceramic plates exhibit immediate resistance to a multitude of environmental hazards, including floods, high-temperature fires up to 1,073 degrees Fahrenheit, and electrical surges. According to Moore’s projections, Cerabyte’s ceramic media is anticipated to last for at least 1,000 years, with a potentially longer lifespan compared to the typical five-year shelf life of HDDs and the 30-to-50-year durability of LTO tape cartridges.
A ceramic-coated storage medium will likely be stacked within a cartridge, paired with a piezo-driven reading head that enables random access through scanning technology. Robotic systems will autonomously relocate cartridges within a storage library, mimicking the functionality of traditional tape libraries. This technique is designed to facilitate writing as soon as, learning many, in various data-intensive sectors including finance, insurance, media, entertainment, healthcare, science, business, government, and consumer markets.
While Cerabyte’s expertise remains focused on innovation, the company has thus far developed a single functional prototype, its flagship demonstration system. The proposed system features a solitary write and browse head unit, designed to achieve read and write speeds of 100 MB/s and store up to 1 petabyte (PB) per rack, leveraging advancements guided by Gordon Moore’s Law.
By 2025, Cerabyte’s roadmap demands a flagship enterprise system capable of processing and analyzing vast knowledge repositories at speeds of 500 megabytes per second, with scalable capacities of 5 petabytes per rack. Additionally, more extensive storage solutions will be available for cloud-based data centers by the following year. By the end of the last decade, corporations aimed to develop programs that could seamlessly transfer knowledge at speeds of up to 2 gigabytes per second (GB/s), while also boasting storage capacities of approximately 100 petabytes (PB) per standard industrial rack.
Cloud giants, major consumers of vast data storage, enable the cost-effective preservation of massive archives by offering affordable pricing for such extensive knowledge repositories. Over the past two decades, researchers from esteemed organizations such as IBM, HP, and Quantum – all members of the LTO Consortium – have achieved remarkable breakthroughs in tape storage technology, yielding substantial efficiency gains. Despite progress, researchers may be nearing the limits of magnetic tape’s capabilities; consequently, the LTO Consortium has revised some of its original storage density and speed targets outlined in earlier roadmaps.
If Cerabyte successfully meets its ambitious goals, then tape may finally have found a worthy competitor. Despite progress, numerous challenges remain to be overcome, a sobering reminder of the fate that befell other once-hyped long-term storage technologies, like Blu-ray, which two decades ago was touted as the go-to medium for storing data over the long haul. Since then, researchers have attempted to store vast amounts of knowledge in various mediums, including DNA, with mixed results.
Clearly, the upside to decreasing the price of long-term knowledge storage is great, and is sufficient to justify the R&D expense that corporations like Cerabyte are keen to spend to seek out alternate options. As the sheer volume of data continues to grow exponentially over time, existing storage systems will become increasingly inadequate, hampered by limitations in capacity, speed, and affordability that hinder their ability to sustain future expansion. As we contemplate the future of data storage, a pressing inquiry emerges: what methods will enable us to effectively store massive knowledge units 50 years hence, and how can we strategically position ourselves for success?
