December 3, 2004
From China, close to the heart of the microprocessor community, IBM announced Power.org, an open standards community for the development and advancement of the Power architecture. The community starts with fifteen members who represent a broad swath of the microprocessor industry, and IBM has stated that more members are expected to join in the next few weeks. The community aims to establish specifications through pooled resources. The first two focus areas will be bus architecture and high-volume servers. Bus architecture will make “system on a chip” possible by making integration of products from different vendors easier. The high-volume server work is focused on low-cost Power-based servers for multiple system vendors in both UNIX and Linux environments. IBM also announced details on the forthcoming Cell chip with Sony and Toshiba. The chip is a multicore semiconductor composed of several processors that work together to handle multiple tasks simultaneously. The jointly produced chip will be used to run next-generation computers, game consoles, and televisions. The chip is believed to be significantly more powerful than conventional chips and able to move large chunks of data over broadband networks, thus proving particularly useful to network-connected consumer electronics. IBM has announced plans to start pilot production of the chip in the first half of 2005. Toshiba and Sony have announced products using the chip with availability in 2006.
IBM first announced the Power chip in March of this year and has been aggressively pursuing a strategy of Power everywhere since. The company has launched servers, blades, and storage with Power5, and in several cases has priced them aggressively to move 64-bit computing into new spaces, particularly with Linux-based servers and storage, suggesting that IBM is actually prepared to practice what it preaches. While much of the processor wars this year have focused on Intel versus AMD in the 64-bit extension of the 32-bit space, IBM has been actively positioning itself as the standard for 64-bit computing. The move appears to be working. IBM is on its fifth generation of the Power chip for its servers, and Intel continues to back down on promises of Itanium as more and more products (such as HP workstations) are revised not to include the Itanium architecture.
The game in processors is volume. Intel could write the textbook on the classic volume marketing play, and AMD has struggled despite having first-to-market status because it has not been able to drive volumes quickly enough. Not content to compete with Intel just at the server level, IBM has taken a broader look at Power. For IBM, Power is far more than Power5 for the Systems Group products. Power encompasses the entire ecosystem, from developers to licensees to customers and users. IBM understands that achieving volumes now means many companies must drive the technology rather than one. IBM intends to act as just one participant within Power.org and not rest as the ultimate voice of the specifications that the group develops. If the community can drive specifications that lead to widespread adoption of Power, and if projects like Cell lead to large product volumes from consumer product industry giants like Sony and Toshiba, then Intel should be nervous: the future of “industry standard computing” may no longer be code for “Microsoft on Intel.”
Sun has announced that it will acquire privately-held SevenSpace (SS) for an undisclosed amount, in an all-cash deal expected to close by Q1 of 2005. SS will expand Sun’s managed services offerings to include heterogeneous environments with support for HP-UX, IBM AIX, Microsoft Windows, and Linux. SS technology monitors and manages enterprise applications, including SAP, Oracle ebusiness suites, PeopleSoft, and Siebel; databases, including Oracle and Microsoft SQL Server; operating systems, including Solaris, Windows, and Linux; and network devices. SS technologies and processes enable system event identification, acknowledgment, assessment, and resolution, as well as change control and system management. SS has a low equipment and personnel footprint that provides remote monitoring and management. SS will be integrated into Sun’s Services organization and sold as part of the company’s services portfolio to traditional Sun markets. Sun plans to develop new services, joining SS technologies and processes to the Sun Connected Customer offering, to target new market opportunities. SS will be sold through direct sales and partner channels. SevenSpace was founded in 2000 and has approximately 100 employees and more than 100 customers in financial services, government, manufacturing, and retail.
The annual growth rate of managed services is one of the few bright spots in the otherwise cautious IT industry. Customers for managed services and technology include service providers, outsourcers, and enterprises. Managed services offer companies an opportunity to focus on core competencies, while ensuring reliable and cost-effective IT infrastructure management. Businesses are increasingly turning to managed services to ease the pressure on IT resources, rapidly deploy new systems and technology, and implement service level agreements, monitoring, and performance. Managed services are a hot market that major vendors are anxious to tap into and Sun is no exception.
SevenSpace expands Sun’s managed services vision beyond what has been primarily Sun-centric. SS remote management capabilities will enable customers to outsource system management functions, yet retain IT asset control, while supporting a range of heterogeneous management capabilities from the physical through applications layers. Heterogeneous management services, besides representing an incremental Sun revenue stream, are synergistic with Sun’s open strategy and initiatives, and may well benefit from this. SS technologies will to some degree help the company further expand outside of traditional Solaris markets and make Sun managed services offerings more competitive against those of IBM, HP, and Cisco. HP’s managed services recently grew 35% quarter-over-quarter and we suspect that Sun would like to replicate this type of growth in its managed service business. The addition of SS should make Sun more effective in Sun accounts with heterogeneous management requirement, which is especially appealing in light of the erosion in Sun’s traditional customer base.
HP announced the OpenView Automation Manager at its Software Universe event this week. HP claims the product will automate the link between business processes and information technology, so that IT resources can be shifted to meet changing business needs. Automation Manager is the first result of the combined business intelligence technologies of HP Labs along with capabilities from its earlier acquisitions of Novadigm and Consera. The product fits within existing OpenView change and configuration management options, and helps establish server management policies that determine when resources need to redeployed to meet changing environments.
For HP customers and shareholders, these kinds of announcements are good news. They imply that not only has HP spent its money wisely in acquiring companies (reassuring to shareholders) but it also means that the technology is composed of worthy bits from which they actually understand how to drive tangible customer benefit. Of course, that is not really a surprise, as HP’s OpenView has been a market leader for years for enterprise management functionality. The interesting bit is how HP has repositioned OpenView of late. As the industry began to get excited by various forms of virtualization (read: automation) and making technology drive business value, HP announced the Adaptive Enterprise as their version of the future of computing. Along with this vision, the company began mumbling about the UDC: the Universal Data Centre. While occasional glimpses of the beast were rumored, the product never took off and nobody seems to have missed it. Thankfully, HP has decided to put at least some of those features into OpenView, which is the important part for customers.
The ability to automate resource change is all the rage among marketers these days, with every vendor offering some variation on the theme with its own product set, and sometimes even with those of their competitors and partners. This is crucial to making data centers more effective, because there are just too many variables for a group of administrators to manage effectively and things are better handled automatically. The problem is that in open systems there is no brain. Intelligence is distributed in various ways depending on the architecture or architectures that the IT department has chosen. This means that there is no one, simple way to identify, prioritize, and control those resources. As a result, every vendor seems to have a different way to approach the problem, based upon the solutions it offers. IT managers don’t want to have competing managers set up in their data center, but more than one solution is generally required to really get a grip on available resources, their usage, and state of health. We expect to see more announcements like this one from HP that focus on a specific set of tasks for a specific set of technologies. As products are designed to work with management, we anticipate that the market will evolve and more complex interrelated products will emerge. Until then, each vendor will be required to better manage at least their own products, and if they want to maintain a strategic presence, they had better learn to manage other products as well.
Citrix announced this week that it has entered into a definitive agreement to acquire Net6 in an all-cash transaction of approximately $50 million. Net6, which was founded as San Jose, CA based WebUnwired in September 2000, produces the Net6 Gateway that securely connects users to applications, voice resources, and collaboration tools from any location or device. Net6 currently has partnerships with Avaya, Cisco, and Nortel amongst others; a reseller relationship with SBC, Verizon, Bellsouth, Sprint, BellCanada, and BT as well as others; and more than 150 customers. Upon closing, the transaction will result in approximately a $1 to $3 million charge for in-process R&D with the remaining purchase consideration to be allocated amongst acquired tangible and intangible assets and goodwill. The deal has been approved by the board of directors of each company and is subject to customary closing conditions with the transaction expected to close before end of year.
Citrix has had the good fortune of being an anomaly in the post-dotcom bust as a software company whose revenues continued to grow and one that remained profitable. While originally focused on remote display technology, the company has now embraced remote and local access control as its mantra. Although the company has made complementary acquisitions before, the focus has remained on software. This is one of the striking differences with the acquisition of Net6, which is clearly a hardware supplier, and one with a notably different customer base. It is this difference that catches our attention most firmly, and begs questions about the challenges this would pose on a formerly software-centric firm.
By acquiring Net6 Citrix gains a hardware-implemented SSL-based access gateway solution that acts likes an appliance with the goal of providing a secure, always-on, single point-of-access to corporate network data including IP voice and data applications from a variety of access devices such as IP phones, PDAs, and of course the existing laptop, desktop, etc. user base. While potentially simplifying the setup of a Citrix access solution by giving the customer an appliance-like device, we wonder if delivering a hardware product through existing software channels will be as simple. Granted some channel partners deliver both hardware and software, but others will require some assistance in understanding why they should now deliver a box as well as a CD. The complementary nature of Net6’s customer base should give Citrix some exposure to new customers especially in the carrier space, but at the same time will require Citrix corporately to adjust its view of its customer and channel base. At the same time, we wonder just what the networking and switching companies like Cisco have in mind in the long haul. With their view of controlling access through intelligence in the network interconnects, there is threat that the value delivered by Net6 could be rendered redundant by the underlying networking technology. Nevertheless, we see the addition of IP voice as just another application that users need to remotely access as forward-thinking, bolstering the notion that the network provides services in many forms. We shall watch with interest to see how well Citrix responds to the challenge it has given itself.