October 14, 2005
HP has announced new products and services directed at the SMB market. Specifically, the company announced new offerings in its Smart Office portfolio that center on data protection such as backup, recovery, and compliance. HP StorageWorks Data Protector Express is designed specifically for SMBs that will work across Windows, Linux, and Netware environments. The product is designed to be easy to configure and manage and will allow backup in tape, disk, or CD environments, and is priced at $779. The new HP StorageWorks MSA 1510i provides a SAN external array with iSCSI interface and is designed to allow customers to simply create a SAN. Pricing starts at $7,995 for the base configuration. The product will also allow users to take internal disks on HP Smart Array Controllers out of Proliant servers and place them in the MSA1519i enclosure. HP said it will be distributing its products both directly and through its nearly 145,000 worldwide distributors and business partners.
The sleeping giant appears to be awakening. HP is showing signs that a coherent, companywide set of strategic decisions are being made with a real eye to the future. While under previous management the company seemed more interested in competing on the consumer side, we are seeing indications that the company is looking hard at the SMB market, as well they should. The fact that the company plans to move these SMB products through channel partners is an indication the company is aware of how SMBs buy IT products and services. HP is also making services available online, another means by which the company can reach smaller customers who would like to avoid being entangled in HPís bureaucracy.
It should be clear to most observers that HP is not breaking any new industry ground. IBM has been building up a formidable lineup of SMB-directed products and services, while at the same time bolstering their channel partners who have both the relationships with the customers and their market niche expertise. HP would be wise to follow that model, which is apparently exactly what the company is doing. It is of perhaps only passing interest that HP has tagged its SMB products with the Express label, just as IBM did several years ago, if for no other reason than to indicate how much ground HP needs to make up to catch Big Blue in the SMB market. But given the fact that HP has been adrift in the past few years, copying pages out of the IBM SMB playbook is a distinct improvement in how the company goes about marketing and selling its products to SMBs. While the company lacks the breadth of offerings IBM can put forward, HP is showing that it is willing to compete with vendors who have a head start in marketplace. That could well be a very good thing for HP and customers as well.
Mercury Computer Systems has announced the Dual Cell-Based Blade, the companyís first product based on the IBM Cell Broadband Engine (BE) processor. This dual-processor blade comes after the June 2005 partnering with IBM Engineering & Technology Services to integrate Cell technology into a range of products in the aerospace, defense, seismic, semiconductor test, and medical imaging markets, among others. The Cell BE is a multi-core chip based on the 64-bit Power Architecture with eight synergistic processor cores to support massive floating-point operations and is optimized for CPU-intensive workloads such as rich-media applications. A high-speed memory controller and high-bandwidth bus interface are also integrated on-chip. The Dual Cell-Based Blade will run on Linux and Mercury will provide the Eclipse-based open source software framework integrating the compilers, debuggers, math libraries, utilities, and middleware. The availability of the Dual Cell-Based Blade is planned for Q1 2006. Detailed specifications to date have only been released under NDA; however, early-access systems are anticipated in the near term.
To some this announcement may generate excitement because of its Cell processor; others may thrill at the versatility of the BladeCenter platform. For us, we find joy in both. The Cell processor, while unknown to many, is an example of the next generation of processor technologies that has culminated from years of intense R&D that is now starting to make itself known in the marketplace. While this announcement would still be interesting in a more pedestrian server form factor, the fact that it is coming in a dual CPU blade package makes it all the more notable. As we have commented many times before, the blade center is not just a way to reduce cabling and real estate demands for multiple servers; its real proposition is the ability to co-locate disparate resources within a common context. Although the Cell BE architecture is OS-neutral and supports multiple operating systems simultaneously, this blade is targeting Linux, which is a popular platform for many compute-intensive applications. Within the BladeCenter context this could make for some interesting combinations such as x86 based blades delivering the Web front end in conjunction with POWER blades accessing databases and data stores that are the basis of very rich multimedia content dynamically rendered by the Cell BE blade. The potential combinations are largely up to the creativity of the IT professional, and the number of open slots in the BladeCenter itself.
Does this announcement portend a massive uptake of Cell BEs as the Linux blade platform of choice? No, at least not yet. We see this as leading-edge technology being targeted at applications that would reasonably take advantage of the capabilities offered. This will by no means become a mainstream home or office technology overnight; however, the lessons learned from deploying the initial solutions will likely make their ways into other solutions overtime. Given Cell BEís aptitude, we probably would find it coming to a home near you in the future, not as a PC or handheld technology, but rather operating a service provider media feed or gaming console, or an amalgam of both. Nevertheless, this announcement is another example of the innovation that can happen when an industry and its players focus not on shrinking the business to profitability, but rather investing in valuable R&D in order to drive the technological curve and market opportunity forward.
Microsoft announced this week it is changing how it licenses its Windows 2003 server to accommodate virtualization technologies. Effective December 1, companies running Windows Server 2003 Release 2 will be allowed to run up to four instances of the software on a single physical server. The next version of the server will have a license allowing for unlimited instances to be run on a single server. Microsoft said it is making the change to encourage companies to deploy its Virtual Server technology. Presently, users are required to pay a license fee for each instance of the server running. The new licensing structure will also apply to SQL Server, BizTalk Server, and Internet Security Accelerator Server.
Letís face it: Microsoft has never been particularly generous in its licensing schemes. It has been quite aggressive in enforcing its EULAs and gone even farther in prosecuting software piracy (as well it should). In many ways, the announcement to allow the latest version of the Windows 2003 server to range more freely in a virtualized environment is quite a turnaround for the company, even if it is also a ploy to encourage users to upgrade to the second release version or to Longhorn, the next version due from the company sometime next year. What a difference a couple of years can make.
It was a little over two years ago that the term virtualization began making headlines and capturing mindshare, if among only the most forward-thinking in the IT community. Original reports on the technology contained some ďgee whizĒ discussions but most concluded that virtualization was a white lab coat technology that had little perceived value to customers. Now, companies like EMC and IBM are making huge inroads to the market with virtualization technology, largely driven by the needs of enterprises of all sizes to consolidate their IT deployments, whether in storage or simply their server farms. Of particular note is the swift uptake of virtualization in smaller firms, where the need to consolidate is driven by tightly stretched IT management assets. Consolidating multiple operating environments and applications on few servers brings real and immediate value to these customers. At the same time, this technology will be disruptive to existing business models, something that Microsoft is apparently aware of. Other ISVs are going to have to also jigger their licensing strategies as the failure to embrace or accommodate virtualization will be seen as a competitive disadvantage. Literally, not virtually.
Microsoft and Yahoo have both announced that they are going to make their Instant Messaging services interoperable with each other. By next year, subscribers to either service will be able to IM, add friends from either service, and make PC-to-PC voice calls. This move challenges AOLís 56% IM market share.
As weíve mentioned before, having interoperable Instant Messaging makes sense. Email that can only be sent to others in the same email paradigm is almost useless; the same goes for Instant Messaging. For example, if youíre the only person you know that has a telephone, then the telephone is just a nifty box thatís hanging on the wall. If the next person that gets a telephone is someone you donít like and donít wish to speak with, then the phone is still a useless bit of decoration. However, if suddenly everyone has a telephone and you can call your friends and family indiscriminately, then the telephone becomes an essential component of communication. The value of the telephone is dependent upon how many people you can call and speak with. Now imagine that you have AT&T, your best friend is on Cingular, and the two networks canít communicate with each other. Do you buy a Cingular phone, or does your friend buy AT&T? The same game is most likely happening on PCs, with one person subscribing to several services. One Instant Message service wonít do if you want the ability to IM all of your friends and family.
If Instant Messaging becomes universally interoperable, will email go the way of snail mail? Mostly likely not. Despite the fact that corporations can now track and save interoffice IMs, email will probably maintain a niche in the corporate world. A person can send emails out to an almost unlimited number of people at the same time; emails have more space than IMs; itís easier to attach and send presentations through emails rather than IMs. Pre-computer-era, the invention of the telephone did not replace the U.S. Post Office; it merely facilitated communication and forged another link between people. We believe that interoperable Instant Messaging will have the same effect of being another useful tool, not replacing anything but rather enhancing productivity.