PERSPECTIVES - Vendor viewpoint

As well as negotiating the great divide between high-volume, low-margin, and high-value, low-volume business, vendors have to accept that it's a case of survival of the fittest in the storage market.

There are two notable aspects to the position of the storage vendorrgin, and high-value, low-volume business, vendors have to accept that it's a case of survival of the fittest in the storage market. right now - one is the polarisation of the market, the other is the increasing competitiveness of the high-end sector.

The split between high-volume, low-margin and high-value, low-volume businesses is stronger than ever. Although vendors such as Seagate still lay claim to the mid-range and high-end server markets with their high-capacity units, the reality is that they operate in the volume sector.

Increasingly, the storage put into 'ordinary' desktop servers is being bought and installed at the OEM level.

The real high end is somewhere else, in the so-called enterprise systems offered by companies such as EMC, IBM, Hewlett Packard, DG, Sun Microsystems and Compaq. In the traditional disk drive manufacturing sector, most of the consolidation seems to be over but there are still a number of vendors slugging it out for volume desktop business - Seagate, Western Digital, Maxtor, Samsung and Fujitsu. IBM less so, but it is still influential in the volume market.

The result, according to IDC, is that there has been too much product and not enough demand, and prices have fallen as a result. But many vendors don't recognise this picture.

Brian Stanley, European vice president of marketing at Seagate, says it's business as usual. 'I wouldn't say it was going to be a wonderful quarter, but it'll be flat. We expect it to be slow in the summer, but we are not over-stocked and keep a close watch on what's going on in distribution.

We have worked hard on our time to market, making sure we meet demand when our customers want it, and certainly, with our key OEMs, demand is pretty good.'

He adds that the OEM market is getting stronger, but this is because Seagate is winning share in that sector. He doesn't believe the increasing share of the PC market that the big brand name vendors seem to be taking is having a noticeable effect on traditional channel sales. The split for Seagate at present is about 45 per cent OEM to 55 per cent distribution.

'We are seeing a shift as we gain share in the OEM market and some key integrators are turning into partners for us,' Stanley says.

He is referring to businesses such as Time, which are raising sales and taking more of the home market. Increasingly, Seagate and other vendors are tying up direct deals with such firms and this is taking products out of the channel.

But the OEM business is also under pressure and, sooner or later, the contest must come to a head. Trevor Duplock, European sales and marketing director at Samsung, is in little doubt about what is going to happen.

'While the market for disk drive products continues to expand, increasing competition means margins and profits will dictate that only the strongest will survive. The survivors will be those players that have efficient, modern manufacturing plants with the capacity to grow; the strongest brand image and financial muscle; and focused management that places particular emphasis on the channel.'

Samsung, with its powerful manufacturing base, is one of the chief threats to the established volume players. But they each have their strengths as well. Quantum has DLT and a good high-end reputation; Maxtor has a lot of Dell's business; and Western Digital has made a comeback with its high-end products.

But it is all far from over and, says Duplock, while the OEM market is important, the ultimate winners in the hard disk drive space will be the companies that can own the channel. Samsung has an advantage here over its US rivals, he claims. 'Samsung isn't beholden to shareholders like so many US firms. It doesn't have to flood the channel with products at quarter-end so distributors have a stockpile of product at a protected price.'

This has always been a problem for the top US vendors and is one reason why Seagate, Western Digital and Maxtor are all desperately trying to break out of the low-end volume market by pushing capacities up and using the latest technologies to enhance performance.

But as well as Samsung, there is fierce competition from IBM, which is strong in the OEM market and has some technology edges, and from a resurgent Fujitsu which, with its GMR head developments and manufacturing facilities, is also ahead of most rivals.

Technology will continue to be important in this market and there will be several developments this year, predicts Mike Nelson, technical director at Fujitsu UK. 'We will see densities doubling every year and the continuation of capacity growth, but we'll see a different emphasis now, still on quality, of course, but also on top-level performance. Demand for capacity, although strong, is also slowing a bit, and I think we are providing enough at the entry level.'

In performance, caches will move up to 2Mb as standard, and rpm speeds up from 5400 and 7200 as standard on entry-level and mid-range performance drives respectively, to 7200 and 10000. Early next year we'll see 15000rpm drives, says Nelson. Seek times will also come down, to an average of five milliseconds, he predicts. But it can go no further because an increase in power consumption would be required and with current PC designs, that's not conceivable.

The importance of storage is being recognised more by the systems vendors.

At Sun, for example, the storage business generates $2 billion of the company's $10 billion annual sales and more than one-fifth of all pure product sales revenues. It is the fastest growing part of the business in revenue terms.

Users have also woken up to the importance of storage and are paying much more attention to it, believes Chris Atkins, storage products marketing manager at Sun. 'They examine storage far more closely than they ever did before. They want to discuss it separately because half of their money is going on storage and most of their expansion will be there.'

That does not mean they will try to separate the purchasing decision from that of the system platform, he adds. 'I think the concept of de-coupling will prove to be a passing trait. It doesn't make sense to do that when you are looking to have an integrated system. It happened in mainframes because we had formally published standards for the attachment of storage. But it hasn't happened with Unix because we don't have those standards - every version is different.'

Donal Madden, UK storage products manager at Compaq Storageworks, says most of its systems go in under NT or Netware, and that it owns 97 per cent of the storage under its platforms. But, he points out, EMC has been very successful as a third-party supplier and it is the large systems that offer the most potential.

'The ratio of storage supplied by a third party is low - probably 80 per cent will just turn to their server vendor - but the higher you go up, the more users will look at storage in a different way. The user may, almost by default, turn to their server supplier for the storage, but if that's not up to scratch they may go elsewhere. Our job is to make sure our product is up to scratch,' says Madden.

Glenn Brackenridge, product specialist at third-party Raid storage vendor Baydel, also believes the systems vendors won't have it all their own way. 'The main suppliers have always considered the storage on their platforms to be their domain. Third parties need to have special functionality or pricing to compete, and as the example of EMC demonstrates, they can flourish.'

Some storage products from the leading vendors are surprisingly bland, he says. Third-party products tend to have more of an add-on sales value or to focus on distinct market sectors. In other words, they are easier to sell.

Derek Warry , strategic business manager at distributor Storm, says decoupling is already happening in the majority of situations at the top end. He claims seven out of 10 companies look at storage separately and that their capacity requirements are going through the roof. The investment required is much more significant and, subsequently, the decision is made much more carefully.

But the use of storage area networks (San) and decoupled storage also gives the customer much more choice, so vendors may have mixed feelings about pushing San forward too quickly. It can be equated with attaching a printer to a Lan, explains Atkins.

'Today, you can go into Dixons and buy a printer and pretty much guarantee that when you plug it into your network, it will work. When we get to that stage with San, people will buy storage in much the same way.'

Brackenridge warns that the development of Sans poses a huge risk to the top vendors because it introduces the concept of competition in their own backyard.

'The result of this has been a round of acquisitions and partnerships as the main players seek to acquire all the elements that will allow them to present a total package,' he says. 'The only problem with this is that the packages currently on offer are close to being proprietary. Either by luck or design, the proprietary nature of these San systems is a method by which the main vendors can actually lock in their customers and fend off rivals.'

All the vendors will, naturally, try to retain as much of the storage business on their own systems as they can. They will try to add as much perceived value as possible, with initiatives such as Compaq's ENSA strategy.

This encompasses network-attachment, San, distance storage, and remote and local management of devices.

Development of San will be evolutionary and it won't supersede all other types of storage attachment. But it will happen, says Atkins, for two reasons. 'One is cost and the other is that it is far more expandable and flexible. All you need to do is attach yourself to the network and you will be able to see all of the storage.'

This does mean that, eventually, prices will fall, but that's all the more reason to get into the market early, says Madden. 'San will follow the same path as all other products and the margins will, inevitably, decrease. It's the early entrants that will succeed and reap the greatest benefit.'