NFR Storage: From Hurricane to Drizzle

During the second quarter, vendors of large-scale removable-disc technologies announced that they'd soon be shipping a new generation of storage products that would "take the media-storage industry by storm."

Leveraging new techniques for laser-assisted data recording on magnetic media, these products constituted a "tape killer" technology, according to its proponents. Vendors predicted that large-capacity removable drives would begin to appear in the market during the summer hurricane season. They would "change the land-scape of storage," leveling midrange magnetic-tape-industry profits, estimated by market researcher International Data Corp. (IDC) to total approximately US$3 billion in 1997.

However, marketing, like meteorology, is an inexact science. The technology maelstrom forecasted by TeraStor Corp. (San Jose, California) and others proved to be more a tempest in a teapot than a real-life typhoon.

Though TeraStor, the leading proponent of large-capacity removable storage, demonstrated functional 10- and 20-Gbyte devices as planned in June, the company also admitted that extensive testing is still required. TeraStor's 10G-byte drives won't ship until late 1998, and the 20-Gbyte drives will follow in early 1999.

Despite delays, Richard Walker, TeraStor's vice president of marketing, remains enthusiastic about the new technology. Together with Mike Hardy, senior product marketing manager, Walker claims that the intervening time is being put to good use-forging vendor relationships, identifying target vertical markets, and building the business case for the product. If all goes well, TeraStor's optimistic projections may become reality in 1999.

TeraStor's technology, called near-field recording (NFR), integrates patents secured from Stanford University and Quantum Corp. to deliver dramatic improvements in storage capacities on magnetic-disc media. Central to NFR is a solid immersion lens (SIL), an optical element in the flying head of the NFR removable-disc drive. The SIL is used to focus a laser beam on the medium, heating its surface and enabling extremely small, magnetic, bit-size cells for data storage. With NFR, the recording density of the medium greatly increases.

TeraStor engineers are working to combine SIL technology with fixed-distance, flying-head read/write technology. Combined with innovative media-production techniques, TeraStor's NFR drive will offer a new, high-capacity, removable drive product with faster read/write access than tape products.

Focus groups of IS managers, system suppliers, vertical integrators, and others have declared that they need this technology to redress the issues they have with tape, Walker says. He points out that the NFR drive has "greater longevity and restorability and a lower price-per-gigabyte of storage than tape."

First-generation products write to only one side of the 5.25-inch magnetic-disc media and use a SCSI interface. Subsequent generations will utilize both faces of the media and will support a Fibre Channel interface. Walker claims that the products will play well in the multimedia video- and audio-editing market, "where the rotational access to data is preferred over the linear access provided with tape," and also in archival database applications, "where data access speeds are important." He adds that "significant interest exists in the fields of medical imaging, document management, and digital asset management." Bob Amatruda, senior analyst for tape and removable storage at IDC in Framingham, Massachusetts, agrees that rotating media and random access might deliver some advantages in certain applications, but he questions whether the price-per-gigabyte criteria can be met by NFR.

"The TeraStor product is clearly aimed at the back-end system, as opposed to removable drives from Iomega Corp. and others, which are used at the desktop," Amatruda notes. "The sweet spot for backup and storage products aimed at the back-end system [or midrange] market is 20 Gbytes."

TeraStor promises a drive with this capacity in early 1999 but must compete with digital linear tape (DLT) and numerous 4-mm and 8-mm tape products that play well there, Amatruda says. To do this, the price point for removable NFR media must be competitive with tape media-which is now falling below US$30 per cartridge for 4-mm products.

TeraStor's Hardy agrees that media pricing will be important, but he stresses that the unique nature of NFR technology will make its own business case for product adoption. He points out that business relationships established with current vendors of tape media, hardware, and software will be key to product success. Hardy suggests that this network of partnerships, including Exabyte, Plasmon, and Maxell, portends early success for the NFR drive when shipments begin in late 1998.

However, he emphasizes that TeraStor's relationship with Quantum-which is expected to compete with TeraStor with its own line of NFR drives-is a guarantor of success. That Quantum, the industry leader in DLT market share (according to IDC), would invest in a purported "tape killer" technology such as NFR removable disc, may seem somewhat confusing.

Brody Keast, vice president and general manager for Quantum's DLT division in Milpitas, California, explains that the company supports a diverse array of backup and archival storage technologies to which NFR will be added. "We see NFR technology as important and as a complementary technology to our other storage products, which include optical at the low-end and tape at the high-end," says Keast. "TeraStor's product is closer to optical technology in terms of capacity, media costs, and other characteristics. We expect it will be cost-competitive with optical-storage products and become a viable alternative in the magneto-optical product space."

Quantum intends to "productize" the TeraStor NFR drive in its own storage libraries, says Keast, and the company is reluctant to respond to TeraStor's earlier "tape killer" claims.

"Let's just say that we see tape as moving to the high-end," Keast says. "There's a role for both DLT and NFR well into the future."

Early marketing hyperbole about the "tape killing" characteristics of NFR removable disk seems to be fading. At TeraStor it is being replaced by more pragmatic messages. According to Walker, the challenges ahead are practical ones. "We're still being compared to Iomega's and Syquest's removable products," he says. "These are desktop storage, while we're clearly aiming at the back-end market space. We're also sometimes classified as another magneto-optical drive, which is not the case. We use technologies that are closer to hard-disk technology than magneto-optical technology in several key areas, such as read/write access. We use hard-disk rechannel technology, for example, to get data off the disc. Magneto-optical drives do not use hard-disk rechannel. We are working to establish this technology as an entirely new product type for the midrange market."

Walker says that other important efforts at present are to develop vertical application-focused solutions and stronger OEM and distributor channels. Taken together, the quest at TeraStor seems less about issuing hurricane warnings to the tape industry than making every move that can be made in advance to ensure that sunny days are ahead for NFR.

(Jon William Toigo is an independent writer and consultant specializing in business-automation solutions. He may be reached at Linear Tape Technology: It Ain't Dead Yet!

Between forthcoming SuperDLT and LTO products, near-field-recording storage will have a tough market row to hoe.

With all of the talk about near-field-recording (NFR) media and digital-linear-tape (DLT) leader Quantum Corp.'s investment in NFR storage developer TeraStor Corp., it may seem that tape technology is fast approaching its zenith. But Geoff Hogan, senior marketing manager for SuperDLT at Quantum, thinks otherwise.

Quantum plans to extend its DLT market dominance with its SuperDLT technology, which uses new tape media and combines optical and magnetic technologies to deliver vastly enhanced storage capacities and access speeds.

The base technology employs laser-guided magnetic recording (LGMR) to increase capacity and performance. Data is recorded on one side of the SuperDLT cartridge, while optical targets are implemented on the back for optical servo tracking and positioning.

"This method is very precise and enables us to increase the track count enormously," Hogan claims. Native storage capacities reportedly range from 100 to 500 Gbytes, with native transfer rates ranging from 10 to 40M bytes per second.

SuperDLT products are expected to ship from Quantum by mid-1999. Within the same year, another high-capacity, high-performance tape technology-named for the multivendor linear-tape-open (LTO) initiative-should also begin to bear fruit.

The LTO initiative, driven by a core vendor alliance of IBM, Hewlett-Packard, and Seagate Technology, is defining specifications now with the goal of expediting time to market. According to Sharon Stone, product-line manager for LTO within IBM Corp. in Tucson, Arizona, "The LTO specifications are a response to growing needs for increased data-storage capabilities in the midrange market."

LTO drives will be delivered in two formats, utilizing two different cartridge-media types. Stone explains that one implementation, known as Ultrium, is optimized for high capacity, and the other, called Accelis, is optimized for fast access. Ultrium, a derivative of IBM's 3590 tape cartridge in some respects, is similar in other form factors to Quantum's DLT tape cartridge. It also comprises a single-reel medium with a storage capacity of 100 Gbytes in native format (200 Gbytes with data compression). Accelis, the second media-format specification, is a fast-access, dual-reel implementation of LTO technology that enables data retrieval in less than 10 seconds and supports midpoint loading. Accelis also bears "IBM fingerprints," according to Stone, since the dimensions of the cartridge are identical to IBM's current Magstar MP 3570 tape cartridge.

The first-generation Accelis product is expected to support 25 Gbytes of data in native format (50 Gbytes with data compression). Initiatives like SuperDLT and LTO underscore a market trend.

Though revenues from tape and archival technologies are expected to decline on the whole by 1 percent between 1997 and 2001, according to analysts, the midrange market is expected to grow at a robust 28 percent in the same time frame. Market researcher International Data Corp. attributes this trend to increases in midrange system deployments, as well as the use of these systems to host applications hungry for storage.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about DLTExabyteHewlett-Packard AustraliaIBM AustraliaIDC AustraliaImmersionIomegaMaxellPlasmonQuantumSeagate TechnologyStanford UniversitySyquest

Show Comments