Market Roundup February 27, 2004 HP and AMD Announce Opteron
Partnership/Products |
|
IBM announced this week that it has signed up more than
200 ISVs to its ISV Advantage Initiative in the first year of its existence.
The initiative is designed to help ISVs integrate their offerings with IBM
technology, including middleware and hardware, in an effort to produce more
comprehensive offerings to customers. The ISVs include a number from various
specific industries including retail, telecommunications, healthcare, and finance
as well as providers of cross-industry applications. The ISVs IBM identified
are from across the U.S. and around the world. IBM noted that more than 70% of
the ISVs signed up to date are actively supporting IBM software on Linux. The
company also indicated that its ISV Initiative was designed to target the SMB
market, which the IBM said was worth some $300 billion, and noted that many of
these ISVs were historically supported only the Microsoft platform. The
announcement comes on the eve of the company’s PartnerWorld event.
When IBM announced in 1999 that it was leaving the application
business, observers scratched their heads and wondered how the company would be
able to compete with the like of Oracle or Microsoft, not to mention the host
of ISVs of all sizes servicing a host of different horizontal and vertical
markets. Partnering was seen as one means to do this, but one that might have
inherent risks. A close partner may become someone else’s acquisition target,
leaving the jilted partner faced with two not particularly pleasant options:
seeking a replacement for the lost partner or doing business with a competitor.
In IBM’s case, it decided to do neither.
Instead, IBM bet that it would be more effective — and potentially profitable — to offer ISVs the opportunity to build applications on IBM products, specifically middleware on which the vertical applications customers are demanding can be run. As a result, IBM is allowing the ISV to leverage its vertical expertise and relationship with the customer; things IBM never really owned nor realistically could attain to the degree owned by ISVs, especially in the SMB space. By positioning itself as a valued partner and business ally of the ISV community, the company also can distance itself from other providers of applications, such as Microsoft and Oracle, noting that those companies often end up entering the individual ISV’s market as competitors. Going forward, we expect IBM to continue tuning its ISV strategy, offering more co-marketing, technical training opportunities, and targeted products as a means to reach into the SMB market in a fashion that is familiar and comfortable to SMBs, who for the most part have been seeking technology guidance and deployment from these same ISVs and IBM business partners for years.
HP and AMD Announce Opteron
Partnership/Products
HP and AMD have announced an expanded collaboration that
will result in the introduction of AMD Opteron-based
systems in HP’s ProLiant server family. According to
HP, the new servers complement the company’s existing ProLiant,
Integrity, and NonStop solutions, and will be of
particular value in high performance technical computing (HPC), government, and
financial services applications. HP plans to deliver its first Opteron-based systems in the first half of 2004, and
described the two-processor ProLiant DL 145 as being
ideal for HPC, Web serving, security, and streaming media applications, while
the four-processor ProLiant DL585 would help
performance in memory-constrained applications including database and Microsoft
exchange environments. HP also plans to ship dual processor Opteron-based
ProLiant blade servers during the second half of
2004, products it said would help maximize performance in dynamic datacenter
and HPC deployments. In addition, HP stated it plans to deliver one- and
two-way ProLiant servers based on Intel’s new Xeon
processors with 64-bit extensions later this summer, and plans to offer four-
and eight-way servers based on the same technologies in 2005. No specific
pricing or availability information was included in the announcement.
Coming hard on the heels of Intel’s subdued announcement
of 64-bit extension technologies for its venerable Xeon line, HP’s plans to
deliver Opteron-based ProLiant
servers sounded laissez faire to the
point of being a bit cheeky, given the close strategic and tactical
relationship HP and Intel share. So what really is going on here? It’s easy
enough to argue the benefits of extending the capabilities of x86 solutions and
HP’s own solutions. However, while HP carefully wrapped the announcement with
pro-Itanium gauze, it is hard to discern how the company could realistically
see Opteron as “complementary” to IA-64. By that same
token, Itanium, a stand alone 64-bit platform, is not particularly complementary
to any other chip architecture, be it x86 or, more appropriately, IBM’s POWER,
Sun’s SPARC, or even HP’s venerable (and soon to be abandoned) Alpha and
PA-RISC. This is not to say that Opteron is a fully
capable replacement for Itanium, which has been designed for the rigors of
enterprise datacenters, but overall Opteron is far
more likely to disrupt Itanium’s adoption than complement it. The reason Opteron caught the market’s and media’s attention in 2003
was because for the first time, the legions of businesses that rely on 32-bit
x86 servers had a clear and easy path to many if not most of the benefits of
64-bit computing that avoided the costs and complexities of dedicated 64-bit
solutions. As a result, Intel’s announced 64-bit extensions for Xeon stand as a
practical acknowledgement of those market dynamics, as well as a profound
admission of strategic errors in its declared Itanium migration path.
So what do HP’s planned Opteron solutions really mean? For HP and its myriad x86 customers, the announcement is likely to be good news. Until recently, IBM has been the largest vendor of and benefactor from Opteron-based solutions, followed by Sun’s announcement and aggressive pricing of Opteron-based servers; and it is never good news when competitors have hot new products that you do not. The deal also works well for AMD, who in the past two weeks has seen a tripling of its list of major OEM partners. Interestingly enough, while HP’s Opteron plans follow the same HPC-focused path as IBM, at least rhetorically, their solutions fall more in line with Sun’s promotion of Opteron as a general-purpose server platform. And therein lies the rub, at least for HP and Intel. While both companies are carefully positioning x86 64-bit extensions for specific enterprise cubbyholes, IT customers are notorious for ignoring vendor directions and using technologies, often quite successfully, for a host of jobs they were never intended for. If Opteron’s popularity and usability continue to expand, and ISVs deliver increasing numbers of applications for the platform, Opteron could well become the x86 64-bit migration solution of choice, pushing Itanium into a rarified HPC niche rather than the industry-dominating platform Intel and HP have always envisioned.
ClearCube
and IBM Japan Announce Distribution Deal
ClearCube Technology, which
develops and delivers PC Blade computing solutions, has announced an agreement
with IBM Japan to market and distribute ClearCube’s
products to customers and business partners in Japan. IBM Japan will distribute
ClearCube’s entire solution, including hardware,
management software, and 24/7 support and maintenance from local and remote
locations. The companies claimed that customers using ClearCube’s
solutions would enjoy stronger security, higher availability, and improved
manageability, and could realize up to 40% reduction in support costs. No
pricing or availability information was included in the announcement.
PC blades are the latest iteration of an IT model that has
been around for years, and described variously as thin clients, network
computing, and x-terminals. Though each of these solutions offers discretely
different options, all utilize remote servers to run programs for desktops, as
opposed to conventional networks of standalone PCs. The benefits server-based
desktops offer are substantial: simplified management and maintenance, higher
systems security and availability, and lower support costs. Despite this and
the continuing enthusiasm of various IT vendors, these solutions have had
difficulty moving beyond limited, usually specialized markets, largely due to
comparatively high initial costs and deployment complexities.
However, we believe the growing acceptance and success of blade server solutions could help to change this. The compact form factor, integrated management processes, easy scalability, and affordability of blades make them a natural choice for server-based desktop solutions. While ClearCube, which has been delivering blade-based PCs since 2002, was an early entrant to this market, larger vendors are beginning to follow its lead. For example, HP announced plans in December 2003 to deliver its Pentium processor blade-based Consolidated Client Infrastructure (CCI) for enterprise desktops early in 2004. ClearCube provides a greater range of solutions than CCI, including both Pentium-based desktops and dual Xeon blades for workstation applications, but the attention and interest of behemoth vendors such as HP and IBM can change the landscape of a market quickly and substantially. For that reason, we see ClearCube’s alliance with IBM Japan as particularly significant. The deal should not only open substantial numbers of enterprise customer doors to ClearCube, but will also offer IBM a chance to evaluate the technological strengths and market potential of blade-based desktops without financial or strategic risk. Such a partnership offers obvious benefits to both companies, and provides the foundation for a possibly deeper relationship in the future.
Speaking at the RSA Security Conference this week,
Microsoft Chairman Bill Gates outlined a number of new security features that
will be included in the next major update of the Windows XP operating system. Gates
stated that the update — Service Pack 2 — will contain a number of features to
improve desktop computer security, including an improved firewall, easier
maintenance of security software such as anti-virus software, and ad-blocking
software. He also told the audience that the company was working on technology
to help fight unsolicited email — or Spam — which he indicated was not only a
security threat but also a drain on resources. Gates further indicated that the
company would begin testing a “caller ID for email” on its Hotmail system later
this year and predicted that Spam will become a lessening problem in the coming
years as new technologies are deployed against it, including the caller ID
technology.
There is no question spam continues to be a significant
problem for end users, ISPs, and network operators of all sizes. End users
complain of the torrent of unsolicited mail filling up their inboxes, some of
it pornographic, some containing viruses, and all of it unwanted. Network
administrators and ISPs complain that the sheer volume of spam threatens to
overwhelm their networks. Recent legislation, most notably the CAN-SPAM act
passed by Congress, has yet to really take hold. According to news stories, no
prosecutions under the law have been initiated, and many spammers are moving
their operations offshore to avoid US laws now in effect. Spam still reaches
its targets, but from places beyond the reach of domestic law enforcement
agencies, demonstrating both the scope and reach of the Internet and the need
for further anti-spam efforts.
We believe that any meaningful techniques to combat spam will come with trade-offs, some which may be less than palatable to some Internet users. Take for example the email “caller ID” technology now under development at Microsoft. While it may go a long way toward diminishing the amount of spam on networks and in people’s in-boxes, it will come at a price: the ability to operate anonymously online. The old saw, “On the Internet no one knows you’re a dog,” will become a quaint memento of the early days of the Internet to future generations. Another anti-spam measure being bandied about is the idea of charging micro-payments for each email sent. The cost associated with each email would be minimal, and not a burden on individual users, its boosters note, but would become a financial burden of significant weight to spammers who send million of emails daily. While the initial costs of such a system in all probability be picked up by ISPs who see a return on such a cost in the form of less spam clutter on their systems, the possibility of a toll booth for individual email would, in our mind, lead to the institution of email postage of a more significant nature, a cost that would eventually be borne by end consumers. For all of its early years, the Internet has proffered the ability to communicate in any guise one chooses, and for a cost that, beyond access fees, has been consistently at zero. Such an environment has led to the plague of spam that so many decry. To do away with spam, we believe, means eventually doing away with these two original, yet clearly not trenchant, properties of the Internet.