Market Roundup

July 15, 2005

To the Core: IBM and LAMP

HDS, Sun, HP Announce New Storage Array

IBM Announces Enhancements to eServer iSeries

More Bandwidth? How About on Power Lines?

Anti-Spyware, Take Two


To the Core: IBM and LAMP

IBM and Zend Technologies have announced the availability of Zend Core for IBM, a product designed to help developers deploy database applications and services based on the PHP Web language. The companies also announced they were working together to further the adoption and use of PHP. Zend also announced the formation of Zend Networks, a resource for PHP developers that will provide them with enterprise-class updates and support. The Zend Core for IBM integrates IBMís DB2 Universal Database and Cloudscape, the open source database management system with Zendís PHP environment. Zend Core is based on PHP 5 technology and is available as a free download.

The announcement that Zend Core for IBM would be developed was made a number of months ago, and for many it may have been something quite easy to overlook at the time. Zend and the entire Linux-Apache-MySQL-Perl/PHP (LAMP) phenomena was only beginning to make itself felt in the marketís consciousness. Here, just a few months later, we would argue that LAMPís profile is growing, with considerable credit due to IBMís embrace of the technology.

Make no mistake, LAMP and its adherents are a ways from having market-wide recognition of a viable solution for enterprise-class environments. But they are getting closer every day. As we have noted in the past, IBM has a pretty stellar track record when it comes to getting out in front of newly emerging technologies and backing them with the full faith and credit of Big Blue. First was Java, which everyone thought was for designing dancing baloney on Web pages. IBM recognized its value on the other side of the fence as a valuable server-side development environment. Then came Linux, and IBM jumped out ahead of that parade when most companies were dismissing open source as a non-starter. Again, given the clout of IBM, Linux is now an increasingly integral part of enterprise computing environments and shows no signs of receding. Which brings us to LAMP. Based on what we have seen to date, and the overall success of open source projects in the marketplace, we believe that LAMP technology is going to shine brightly in the coming years, and due in no small way to IBMís early investment and backing.

HDS, Sun, HP Announce New Storage Array

This week HDS announced the new Network Storage Controller model NSC55, a lower-cost, scaled-down version of its TagmaStore high-end storage controller. OEM versions of the product were also announced by HP as the StorageWorks XP10000 Disk Array to which HP will add clustering, disaster-tolerant, and fast-recovery software to the HDS product. Sun will also offer the product as the StorEdge 9985 to replace the current 9970. HDS claims the product is targeted at small and mid-size enterprises that have traditionally purchased midrange systems. With the new NSC55, they should be able to enjoy enterprise functions like high-end customers have enjoyed. The NSC55 will manage up to 16PB of internal and externally attached storage, including arrays from other vendors. The array can hold up to thirty-two controllers and cache to ensure data availability and performance. The price is expected to fall at about one-third the price of the original TagmaStore.

The new storage is the latest in the ongoing round of storage array wars between HDS, EMC, and IBM. And as IBM did with the launch of the DS6000, HDS is attempting to follow in the footsteps of the server vendors and drive high-end features downstream in midrange packages. And like IBM, HDS is finding that itís easier to do the engineering than it is to adjust the market mentality. Sun and HP are positioning the new products clearly in the high-end range. (HP in particular has its own successful mid-range products.) IBM still sells the FAStT line renamed as part of the DS family with different numbers, and positions the lower-end box as the 6000 line. The truth is that storage arrays, like servers before them, are transitioning. It is more cost-effective for vendors to develop systems based on the same architecture, and it is more cost-effective for end users to have fewer types of systems to manage. It is also a win for mid-range customers who frequently need the same capabilities in a smaller package and not a watered-down version of enterprise systems.

At the same time, it would behoove HDS, Sun, and HP to coordinate their marketing a little better. There is some confusion as to what this product is ó whether the product is meant to be high-end or midrange. It doesnít really matter whether a vendor positions it one way or another ultimately, as customers will gravitate to the price and capacity they need regardless of what itís named or how itís positioned. On the other hand, storage is the part of hardware architecture where vendors still seem to sell mainly products rather than solutions. The real value, of course, isnít the storage;itís the data inside the storage that matters, and understanding the value, flow, and movement of data ó ILM as itís known ó is where customers find real value, and not in purchasing another array, disk, or tape library. The fact that HDS, Sun and HP canít come to some common agreement on marketing and positioning could give EMC and IBM a lead if they can figure out how to take advantage of it and start selling solutions rather than more disk and tape.

IBM Announces Enhancements to eServer iSeries

IBM has announced enhancements to its eServer iSeries that include new offerings for the model 520 and new pricing flexibility for the model 570. The i520 now offers a new 2400/60 CPW performance point for the Express and Value Editions and is targeted at SMBs with fifty or fewer concurrent users that do not expect to grow beyond a one-way system. In addition, IBM has announced the Solution Edition for the i520 that features solutions from select ISVs offering the combined experience and expertise of IBM and its partner ISVs. The Solution Edition is based upon the i520 1000 CPW Enterprise Edition and is targeted at SMBs with fifty or fewer concurrent users that do not expect to grow beyond a two-way system. In addition, the Solution Edition can be upgraded to the 1000 CPW Enterprise Edition at no charge and then upgraded to 6000 CPW. IBM has also modified its software maintenance licensing of i5/OS so that all iSeries systems may purchase the OS license on a per CPU basis; the minimum number of CPU licenses will now be one. The i520 Solution Edition is priced starting at $39,500 when combined with a validated $25,000 minimum purchase of a supported ISV solution. The 520 Express Edition Turbo #6, 2400/600 CPW with 4GB RAM is priced starting at $42,000; the 520 Express Edition Turbo #7 with RAID, 2400/600 CPW with 4GB RAM and 40MB disk write cache is priced starting at $48,000. The i5/OS software maintenance agreement license starts at $1,200/year/CPU based upon system capacity.

As we noted last week, IBM continues to be investing in the iSeries, and the latest announcements are yet more proof of Big Blueís iSeries spirit. What strikes us most about these announcements is that IBM has continued to simplify the pricing of the iSeries to make it more consistent with licensing models available with pSeries and xSeries while at the same time packaging offerings that make it easier for customers and IBM business partners to deploy iSeries solutions. The Solution Edition for the i520 along with the new CPW price point for the i520 is well positioned to encourage more ISVs to support the platform while potentially broadening the user base for existing iSeries ISVs. In addition, the investment protection inherent in the i520 Solution Edition through the free upgrade to Enterprise Edition status when seeking to upgrade the CPW should provide some peace of mind to those organizations who may be attracted to the off the shelf approach of Solution Edition but concerned that their needs may soon eclipse the editionís shipping workload capacity.

At the same time, we see the simplification and unification of the i5/OS licensing model as a good move for customers and for IBM. Customers rarely, if ever, complain when the base cost of licensing declines; however, IBM has also made it easier for end users to consolidate UNIX and other workloads on the iSeries by reducing the minimum cost of deploying an iSeries with sufficient capacity to handle the consolidated workloads without requiring more i5/OS capacity than needed. This improved flexibility should help the iSeries become a more attractive consolidation platform for those with AIX 5L, Linux, Windows, and i5/OS applications. Overall, while to some these announcements may not seem earth-shattering, they are reflective of incremental improvements in value delivered while maintaining a customer-centric focus. To our way of thinking, these are the kind of improvements that ultimately build customer loyalty and promote long-term revenue growth.

More Bandwidth? How About on Power Lines?

Recent reports indicate that broadband penetration in the U.S. home market stands at just over 57% as of March 2005, up three-quarters of a percentage point from February. Broadband penetration at work stands at slightly more than 81%, unchanged from the previous month. These numbers continue to remain higher than European countries, which, while experiencing larger growth spurts in market penetration, still lag behind the U.S. with overall market penetration rates in the mid-to high 30% range. Meanwhile, both Google and IBM have made commitments to continue to experiment and test the feasibility of broadband delivered over electric power lines (BPL). IBM announced it will be working with a Houston energy firm to deliver BPL in the Houston area.

Google is exploring BPL as a means to get around having to have third parties deliver its ever growing stable of content and services. IBM would appear to be doing so as yet another effort to explore a technology that could someday make the company a great deal of money. Given that broadband providers ó most notably the cable and phone companies ó are sharply discounting their broadband offerings in attempts to gain more of the uncommitted market share, one would suspect that they would be very suspicious if not outright hostile to a viable broadband delivery system that would not require huge capital outlays for its infrastructure.

But even if the BPL supporters are able to circumnavigate the opposition of the incumbent broadband providers, there will be other thorny issues to untangle before BPL goes mainstream. While some of the technical issues have been addressed, others remain unresolved. One, for instance, is the question of how much radio frequency emissions will increase with the deployment of BPL and what parts of the already crowded radio spectrum will be impacted. Perhaps even more daunting than the technical issues, however, will be those of regulatory oversight. Will local Public Utility Commissions have the power (pun intended) to regulate rates? Will the Federal Energy Regulatory Commission take this role? Or will it be the FCC? These questions are far from answered, and in fact may not be in the process of being answered. Before we see BPL make an impact, a fair bit of legislative soul searching will have to occur, and that, quite frankly, could take years if not decades.

Anti-Spyware, Take Two

In April of this year, the Center for Democracy and Technology spearheaded an effort to unite the IT industry in fighting spyware by forming the Anti-Spyware Coalition (ASC). Some of the IT members of ASC are AOL, Symantec, EarthLink, HP, McAfee Inc., and Panda Software. Public interest members include Samuelson Law, Technology & Public Policy Clinic at Boalt Hall, UC Berkeley School of Law and The Canadian Internet Policy and Public Interest Clinic. The ASC is an effort to replace the Consortium of Anti-Spyware Technology Vendors (Coast) which fell apart in February of this year for several reasons, one of them being an ongoing debate over whether to include in their membership companies that actually make spyware. The first mission of the ASC is to define exactly what spyware is so that they can more effectively fight it. To this end, ASC is inviting public comment on their Web site.

To defeat the monster, we must first identify the monster. This is a bit more problematic than it at first seems. For example, look at the problems courts have had when they tried to define pornography. ďI know it when I see itĒ doesnít work as a definition of pornography and wonít do as a definition of spyware, either. To that end, the ASC has tentatively defined spyware as any software that is installed on a userís computer that is not asked for, not known about, and not easily uninstalled. But this definition can be changed, partially because one of the objectives of the ASC is to have consensus upon every point. Good luck with that.

The ASC has written a thirteen-page paper about the problem of spyware and has invited public comments. Part of the problem with spyware is that it is simply technology, and as with all technology, it can be used for either good or not so good purposes. Just banning technology is not an answer; for every ďdarkĒ purpose there is also a ďlightĒ purpose. It just depends on what side of The Force the user is on. For example, the tracking software that a shady company uses to illegitimately share personal information could share components that are found in software that parents use to monitor their childrenís Internet behavior. So where is the line drawn and how is that line defined? Wherever that line in the sand ends up being drawn, we anticipate that it could affect every IT company in some way with the potential of new laws, regulations, or controls on their products. Much as farmers petitioned the USDA to officially define the tomato as a vegetable despite its scientific characteristics indicating that it is indeed a fruit, we suspect that many vendors of spyware would lobby to be defined as personal assistance, search, or preference enhancement software or other less nefarious sounding moniker. Since everyone is affected, everyone believes they have a say. However, as when voting on an issue of great importance, one canít count on a great turnout. It will be interesting to note exactly how many consumers will say anything at all.

The Sageza Group, Inc.

32108 Alvarado Blvd #354

Union City, CA 94587

650∑390∑0700†††† fax 650∑649∑2302

London +44 (0) 20∑7900∑2819

Milan +39 02∑9544∑1646


Copyright © 2005 The Sageza Group, Inc. May not be duplicated or retransmitted without written permission.