Analysis

posted on 25 August 2008 13:31

Quantum is emerging from five years of transition with ADIC technology at last bearing fruit that could ripen into enduring profits. The maturing, read declining tape revenues, at last have a counterbalance in the DXi deduplicating drive array range.

After Rick Belluzzo joined Quantum in September 2003 the runes his strategists read all started pointing the same way. Its proprietary DLT tape automation technology was going to be obliterated by an unstoppable transition to LTO. Worse it became apparent that tape was no longer the bastion, the foundation and the be all and everything of data protection, but would be come a niche protection technology as disk-based protection in various guises became more and more popular as cost/disk GB fell, capacities rose and software managed to fool backup software that a drive array was a tape library.

Cheap SATA drives and VTLs were two nails in DLT's coffin and they hammered down the LTO lid on that now fading technology. Of course tape technologies don't actually die. They go on and on and on as the installed base moves with glacial speed to embrace replacement technologies over multiple decade-long transitions. Quantum had plenty of time to regroup. It decided to do three significant things.

It bought the Seagate Certance tape automation business and thus became an LTO adherent with equal billing to HP and IBM. It bought ADIC and, as well as consolidating the tape automation market it gained disk data protection products and a deduplication technology. This was the magic key that increases effective disk backup capa ity by 20:1 or more and cements disk backup's appeal for combining disk speed with tape-like capacity. But it requires a lot of priming and aiming before its promising silver bullet hits the target.

Third Quantum decided to become a systems company and move away from being a vertically-integrated tape automation manufacturing company.

Eighteen months ago the ADIC acquisition completed and Quantum had $400 million of debt which it was bound by contract to pay off before it could strengthen shareholder value by buying back its own shares. The tape market declined faster than Quantum could earn money from disk products and its software portfolio. Its first generation deduplication product was not that successful and a focus on it, particularly in North America, led to mis-steps in tape automation sales. Quantum made losses and became a distressed stock.

The recent results showed a company in an apparently poor position. We reported 'For its Q1 fy089 Quantum reported a ten percent year-on-year fall in revenues to $222 million from $202 million. There was a GAAP net loss of $14 million ($0.07/share), an improvement of 4 percent on Q1 fy08.' The company has revealed more information in an earnings call that revealed the self-inflicted wound in North America and a very promising start to its second generation dedupe products.

The revenue fall was driven by unexpectedly lower tape automation product sales in North America, and compounded by costs from a dedupe patent dispute with Riverbed (legal fees of $2.8 million), a $50 million ADIC debt payment leaving $290 million still to pay, and a $1.1 million cost to restructure the US sales team with five new area directors, a changed compensation scheme and more channel-friendly activities.

The big sales boost came from the DXi range with a new 9TB entry-level product, a core to data center offering, all based on its new policy-driven dedupe engine, and the first pleasing but not yet material revenues from EMC which is licensing the software. Quantum thinks it has a winner on its hands here. It will, in effect, stay out of EMC's large accounts and let that company tear into Data Domain, the main dedupe competition, there while it takes on Data Domain in its own accounts and through its channel.

Belluzzo said Quantum was going to focus more on its ADIC-acquired StorNext software for file clustering and data management. So far Apple has been the only effective licensee although it is now on HP's price list. The jury must still be out and, until it comes back in, we can't be certain whether this is a promising asset or an also ran. Certainly other companies are making far more of file clustering, such as Panasas and Ibrix in the software space and Isilon and others in the SW+HW system space. With HP having its ExDS9100 highly scalable clustered file storage product in the wings it's hard to see HP pushing StorNext very strongly.

Quantum is also rebalancing it revenues by moving away from low-margin OEM sales and more towards higher margin branded goods sales.

Also Quantum is no longer developing its own tape drive technology. The LTO R&D deal with HP was a means of outsourcing tape drive development to HP and so saving cost and moving away from inhouse HW R&D and manufacturing.

Belluzzo said nothing in the earnings call about Quantum's removable, and deduping, disk cartridge GoVault system. ProStor has 90 percent or more of the removable disk market with its RDX system and the EMC-owned Iomega will be pushing its own Rev technology. GoVault looks to have missed the boat so far. Perhaps a recent price cut of up to 30 percent will do something but Belluzzo said nothing about GoVault in the earnings call. It looks as if it is largely irrelevant to Quantum's future when compared to the DXi products and StorNext.

Even StorNext is a sideshow. The main events are continuing to milk the declining DLT and flat LTO base, paying off ADIC, and pushing DXi sales very hard indeed. Its path to tape option, which half of Quantum's customers are buying, suggests that it will help keep the tape base happy as well as directly boost Quantum's revenues.

It could be another 18 months before the ADIC debt is paid off. Oddly ADIC is both lifebelt and millstone. The ADIC technology-inspired DXi is the lifebelt while the $50 million a quarter debt repayment is a millstone dragging profits down. The trick is pump more buoyancy into the lifebelt - sell a bundle of DXi product - and hope, really, really hope, that the lawyers are right and the Riverbed case is winnable. If it's lost then much will hang on how much it costs  in legal fees, license fees and damages. The case goes to court in March, 2009. Riverbed could certainly do with a revenue boost of its own.

Quantum's progress on its road back to profitability should be much clearer at the end of its next quarter (Q2 fy09) in October. A full quarter of EMC dedupe revenues and hopefully rebounding tape automation revenues in the USA should provide a net profit for the quarter, granted no recession in the USA. DXi revenues are the key. Go baby, go!

^{[Chris Mellor.]}