REALITY CHECK ON NEW TAPE FORMAT.
Polaroid, which owned the `instant' camera business in the 1960s and much of the 1970s, was smart enough to realize that its major competitor was itself--although it had the patent on the basic technology enabling its products, the company continued to push for newer, better products that leveraged the technology. Similarly, Quantum has been pushing to keep up with, or jump ahead of, competing technologies that promise competitive performance and reliability specifications. Like Polaroid, which developed a color version of its photographic technology because color photography was becoming the dominant competing technology, Quantum has similarly been making ongoing changes in order to meet the ever increasing market demands for higher capacity, higher performance, and lower per-megabyte storage costs. In this article, a few of the most recent DLT challengers, including DLT itself, will be explored.
New 8mm Offerings
During 1999, most of the 8mm contenders announced last year have actually begun shipping. Sony's AIT-2, for example, began shipping mid-year and is available from Sony and a number of its OEM customers.
AIT-2 boasts native capacities of 50GB per tape and a 6MB/sec native transfer rate. With compression, both capacity and performance can be expected to run at approximately 2:1, depending upon the type of data being recorded onto the tape. In addition, the AIT-2 cartridge uses memory-in-chip technology that enables directory information and usage statistics to be recorded onto the cartridge. This feature enables automation devices to read information about the cartridge without having to load it into the drive.
Exabyte's long delayed Mammoth-2 drives are expected to begin shipping sometime in November, according to reports from the company. Mammoth-2 will reportedly have a native capacity of 60GB and native transfer rate of 12MB/sec. Again, with compression, capacity and performance may double.
Exabyte has recently trimmed back its work force and has offered a free upgrade program for some of its tape library systems. Current purchasers of specific systems that use Mammoth drives will be eligible for a free upgrade to Mammoth-2 drives once they become available. Additionally, Exabyte appears to be repositioning itself as a SAN integrator through its NetStorm initiative. NetStorm is designed to sell SAN backup systems that use Exabyte's tape libraries and to provide complete systems on which the somewhat intimidating SAN configurations have already been tested and proven. It also, obviously, is designed to help boost Exabyte's sales of automated tape libraries.
Exabyte recently announced price reduction on its low-end Mammoth drive, the Mammoth-LT. This drive, with a price of under $1000, offers 14GB native capacity and a native transfer rate of 2MB/sec. At its current pricing, the Mammoth-LT is clearly positioned to compete with such lower-end devices as DDS-3 and 20GB Travan drives. Additionally, any capturing of market share away from DDS and Travan can only help boost the financial performance of the company.
A potentially strong challenger to DLT, in addition to its 8mm helical scan cousins, is the Ecrix VXA-l drive. The VXA-1's 3MB/sec native transfer rate puts it more into a league to compete with current generation Mammoth, NIT-1 and DDS 3 drives than with Mammoth-2, AIT-2, or DLT7000, DLT8000, SDLT, or LTO drives. Native capacities of 12GB and 33GB per cartridge also position it roughly in line with Mammoth, AIT-1, and DLT7000.
Perhaps the strongest selling points that Ecrix has made for its technologies are reliability and price. With a suggested price of $1165, the Ecrix drive is priced competitively to DDS-3 and slightly above Exabyte's Mammoth-LT. However, native capacity and performance suggest clear superiority over DDS-3 and DLT4000, and much lower cost-per-megabyte than DLT7000 or Mammoth drives.
Even more important than price/performance may be the claimed robustness of cartridges and drives using Ecrix technology. Unlike other recording technologies that record large frames or blocks of data, Ecrix records data in 64-byte packets. Because the packets are so small, write errors can be easily detected and corrected on the fly, virtually guaranteeing 100% data accuracy, according to the company. Further, Ecrix's read technology uses four heads to read the data. Each packet can be read more than once by one or more heads as the tape moves past the head. Electronics inside the drive reassemble the packets into a continuous data stream.
Ecrix has demonstrated the ability of its drives to read tapes written in other Ecrix drives. Additionally, Ecrix has demonstrated the ability of its drives to read cartridges that have been subjected to such torture tests as freezing, dropping, and even dropping into hot coffee. The ability of the Ecrix drives to read cartridges that have been subjected to hazards that may make cartridges used by other tape formats unreadable may help this company earn market share in the midrange storage market.
DLT And DLT-Alikes
Quantum's DLT is no longer the only DLT kid on the block. Tandberg Data now manufactures DLT 7000 and DLT 8000 drives, under license from Quantum Corporation. In addition to the DLT drives, Tandberg continues to manufacture and sell its SLR family of tape drives.
While the industry awaits the SuperDLT (SDLT) tape drives, Quantum (and now Tandberg) has been shipping the DLT-8000, a modest capacity and performance-improvement over the DLT-7000 drive. While all three drives, the DLT-4000, DLT-7000, and DLT-8000 use the same DLT-4000 media, each drive has different capacity and performance specifications.
The DLT4000 has a 20GB native capacity, while the DLT7000 has 35GB and the DLT8000 provides 40GB. The transfer rate of the DLT4000, at 1.5MB/sec is more than doubled by the DLT7000, with 5MB/sec. The DLT8000 delivers a 6MB/sec native transfer rate. While the DLT8000 is clearly an incremental improvement, the improvement is only modest, serving as an end of the road for "standard" DLT technology before SDLT becomes available (possibly later this year).
An interesting DLT variant was announced in early September by Benchmark Tape Systems. The Benchmark DLT1 tape drive is read compatible with data written by a DLT4000 drive. The drive provides 40GB native capacity, equivalent to the storage capacity of the DLT8000 drive, but is not read or write compatible with the DLT7000 or DLT8000.
Native performance of the DLT1 drive at 3MB/sec places it on a par with the Mammoth and AIT1 drives, and less than that of DLT7000 and DLT8000 drives. The Benchmark DLT1 drive achieves higher capacities by increasing the number of tape tracks that it can read and write.
The most interesting feature of the Benchmark DLT1 is its price--with an approximate street price of $1299 for the internal model, this positions it slightly below the DLT4000, according to material provided by Benchmark. In essence, the DLT1 drive offers twice the capacity and twice the performance of the DLT4000 drive at a slightly lower price. Media costs will be approximately the same as those for standard DLT4000 tapes. The cost savings are primarily in the drives.
Benchmark will also be offering automation, enabling the use of multiple cartridges with a single drive. The form factor of the DLT1 drive is similar to that of other DLT drives. The use of the same cartridge should enable the use of the DLT1 drive in automated systems designed for DLT drives.
The company will market the DLT1 as having "The Power of DLT at the Price of DDS." In many ways, the description is appropriate and the drive may be attractive to price-sensitive companies wanting DLT and not requiring read and write compatibility with other DLT formats.
Curiouser And Curiouser
I used to think that it was easy to figure out what a first generation device was. This was a device that used newly developed technologies and that was the first of its type. A drive that improved upon an earlier implementation was a second, or later, generation drive.
By that definition, it would seem that the LTO drives, both Ultrium and Accelis, could be considered first generation, since there were no previous LTO drives, and certainly none created through the collaboration of IBM, Seagate, HP, and Fujitsu. By the same definition, SuperDLT would not be first generation because it builds on previous DLT technology. Right? Wrong.
The current, working definition, seems to be "first generation is what we're using and not what the competition is using." Thus, the Ultrium could not be considered first generation because it makes substantial use of technology developed by IBM for its 3480 drives (according to Quantum). Further, according to Quantum, the SuperDLT drive could be considered first generation because it's a major redesign of the DLT drive.
There are some significant differences between DLT and SuperDLT. These include a simplified, redesigned tape path and tape loading mechanism, new head technologies, and an optical servoing mechanism for precisely positioning the tape heads.
What's first generation? What's not? Does it really matter?
Will the newly shipping 8mm drives eat into DLT? Will LTO drives, or, for that matter SuperDLT and DLT1 drives eat into the DLT market share? At the low end, will the Travan 20 or Quarter Inch Ccommittee (QIC) drives eat into the desktop workstation market? Will DDS4, with increased capacity and performance that can finally challenge the bottom of the high end market, eat into the server space?
What ultimately matters is the ability to restore data. A company's decision will be based on many factors: storage capacity, compatibility with prior technologies, performance, robustness, available automation options (if needed), price, and, perhaps most important, the ability to restore from the media onto the drive that wrote the data, as well as any other "compatible" drive, in case of a disaster--another factor that deserves increasing attention of both the drive and its manufacturer.
The need for storage is evolving and continually increasing. In many enterprises, backup is evolving into an activity that is done remotely or on serverless systems. Older technologies will continue to be displaced by newer, more capable ones.
There is significant growth potential for all tape technologies. The question may not be as much whether or not any company can eat into DLT market share--even with a significant drop in DLT's piece of the market, the growth of the tape market may still result in continued growth in DLT unit sales.
As seen from the battlefield, there appears to be the potential for shared victory by most major combatants. Although each may wind up sharing a market because the market continues to grow, most tape formats will see increased overall sales, even if market share remains constant or declines for some, and grows for others.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||DLT tape drives|
|Publication:||Computer Technology Review|
|Article Type:||Buyers Guide|
|Date:||Sep 1, 1999|
|Previous Article:||WAP: New Web Whopper.|
|Next Article:||Online Unified Messaging: Fast, Cheap, On The Way.|