|
IBM and Intel
By Joyce Tompsett Becknell
This week IBM and Intel announced their intention to
jointly design and develop modular server solutions, also known as blade
servers. The companies will focus on three aspects of development: system and
chassis development, networking infrastructure, and systems management. The
two companies aim to produce blade servers that provide scalable performance
and reliability features at a lower total cost of ownership. IBM plans to soon
launch blade servers that will incorporate Intel Xeon family processors as
part of a solution, and Intel intends to provide blades to its OEM customers
based on the Xeon architecture. The companies will be able to offer the
portfolio of jointly developed products to their respective customer bases.
IBM states that blade servers are expected to reduce
costs with improved systems management, simplified provisioning, and
increased reliability. These goals are not particularly surprising, as most
vendors are designing systems and architectures to solve these rather
pertinent customer issues. What is interesting is the bit where IBM argues
that they will not sacrifice performance for increased density. And indeed
poor performance has been the bugbear that has limited customer adoption of
blade systems so far. The first generation of blades has not been able to
maintain attractive price/performance ratios. Customers performing comparison
tests have frequently found that pizza box servers provided better
performance for the price even with server density factored in.
IBM and Intel maintain that IBM’s experience with
architecture and systems design combined with Intel’s expertise in building
modular servers, chipsets, and software optimization will provide the best of
both worlds and produce the new generation of blades that will genuinely
provide the capabilities customers are looking for. Sageza believes that this
is a win for both companies. For Intel, selling more blades means selling
more processors. And creating an OEM system that beats vendor solutions gives
Intel opportunities to sell greater volumes of higher-margin products. For
IBM, this is the opportunity to replicate their success in the original PC
market. Once again IBM can create a system that becomes the standard for the
next generation of modular computing. Sageza also believes that the focus on
improving performance is a win for customers. Without addressing this issue
the blade market will be just another fad that fades when customers find
alternate ways to address the problems that blades were meant to solve. By
focusing on increasing performance as well as better manageability,
reliability, and serviceability, the evolution of blade computing will affect
the evolution of SMP and its role in modular computing as well.
|
|
Microsoft Offers WiFi Hardware for
the Home
By Jim Balderston
Microsoft has announced that next month it will
begin selling Microsoft-branded home wireless networking hardware based on
the 802.11b standard and use the Intersil PRISM WLAN products as its core
technology. Among the items Microsoft will be selling are: a wireless base
station ($149.95), a wireless USB adapter ($79.95), a wireless notebook PC
card adapter ($79.95), and kits including both a base station and adapters
for notebooks or USB connections ($219.95). The company will also offer wire
10/100 Ethernet gear as well, including a base station ($79.95), USB adapter
($29.95), a notebook adapter ($39.95), a PCI adapter ($24.95), and a 5-port
switch ($39.95). The wireless hardware will include a 128-bit encryption that
is turned on by default, a network address translation (NAT), and a built-in
hardware firewall. The company also offers further security through its
browser-based Base Station Management Tool, such as address filtering and
parental controls.
Its déjà vu all over again. Microsoft, the world’s
largest software maker, has never shown any real interest in becoming a
hardware vendor — a la Sun, HP, or IBM — but it has repeatedly dabbled on the
periphery — or should we say peripherals? Branded keyboards, mice, joysticks
have all shared a common strategy: to drive forward the value proposition of
the PC, and of course drive sales of Windows software. This latest foray into
branded hardware seems right on track with those earlier initiatives.
Windows XP has a significant amount of easy-to-use
networking capability built into it. By offering its own branded WiFi
hardware we believe Microsoft is trying to assure that these new capabilities
are actually used, especially in homes and small businesses where the non-alpha-geek user would rather not
participate in the head-scratching (and sometimes fist-pounding) rituals of
trying to make multi-vendor products work together. Utilizing features like
Internet Connection Sharing (ICS), not only will this initiative drive sales
of XP, it will elevate the status of that PC to an Internet server. This will
also make the proposition of wireless handhelds and multiple PC environments
in the home a much more viable and attractive proposition, which of course
means more software sales. Looking a bit farther down the road, establishment
of the PC as the home/small business Internet server will open up a host of opportunities
for connecting and managing non-PC devices that have IP capability. This will
position Microsoft to own the software and networking environment that will
be pivotal in establishing what Bill Gates has called the Internet Lifestyle.
Entertainment systems, including TV’s, audio systems, digital recording
devices, and interactive gaming, as well as medical monitors, HVAC systems,
security systems, and more will all become part of the network managed
through a Windows PC. Looking forward to living like George Jetson? Let’s
just hope it doesn’t turn into Fahrenheit 451.
|
|
IBM Upgrades Web Services Security
By Jim Balderston
IBM has announced new software to further secure Web
services based on the WS-Security specification. The new software will allow
IBM customers to secure transactions with partners regardless of what type of
web services or security the partners are deploying. IBM’s WebSphere
Application Server Version 5 will support WS-Security in the fourth quarter
of this year and the Tivoli Access Manager will do so early in 2003,
according to IBM. The new capability in WebSphere and Tivoli Access manager
will feature out-of-the-box support for the XML Key Management Specification,
and, in the future, other identity standards including Security Assertions
Markup Language, Kerberos, and XML Digital Signatures. The new software will
also allow WebSphere customers to automate the process of created trusted relationships with
partners, whether they are using Microsoft TrustBridge, Kerberos Tokens, PKI
credentials, or other methods to be developed in the future. IBM also plans
to offer higher granularity authorization for SOAP transactions within the
Web service environment.
Essentially, IBM is enjoying the fruits of being Big
Blue. For big systems vendors like IBM, with its well-entrenched place in the
market, the company can avoid the risks smaller vendors must take when it
comes to betting on a particular standard and its future market share. For
smaller companies, such a bet is often fatal. Here, however, we see IBM
rising about the fray of multiple computer security standards and offering
its customers a product that does not bind them (or their suppliers) to
single standard. Of course, it doesn’t hurt if you are one of the co-authors —
with Microsoft — of the WS-Security specification in the first place. That’s
a privilege reserved for big market dominators as well.
IBM’s customers are the end beneficiaries as they
are theoretically released from the cursed vendor lock-in when it comes to
creating trusted relationships with partners and suppliers. To many
companies, the ability to be able to reach beyond a supplier base artificially
constricted by islands of non-interoperable security schemata could become a
real boon. Creating and managing a series of trusted relationships with the
broadest array of available suppliers offers a much more flexible — and
market driven — supply chain, no doubt of keen interest to many companies
fighting to hold on to margins in a tightening economy. Such margins and
efficiencies will only provide greater benefit as general economic conditions
pick up. In short, IBM rightfully sees no need — or benefit — to itself or
its customers from picking a single security standard and asking everyone to
fall in behind it. Why have an ice cream cone when
you can have the whole store?
|
|
Feds to Drive IT Security?
By Jim Balderston
The White House and FBI released a 65-page draft
report entitled “A National Strategy to Secure Cyberspace” outlining the
federal government’s plans to enhance national IT security in the public and
private sectors. The plan as it is presently laid out is largely voluntary
for the private sector, with government agencies more on the hook to meet
deadlines and security standards, when finalized. As it is now, the plan asks
industry to consider conducting a gap analysis on security research and
development, and ideas surrounding more government-funded research. The
report contains a number of charts showing the increase in attack incidents
and responses in recent years, as well as details on the growth of Internet
domains. The report includes a hypothetical scenario in which terrorists take
over a variety of U.S. infrastructure IT networks, such as the air traffic
control system, and shut them down. The document also outlines overall
strategy and policy. Notable is the statement that formal regulation to
enhance security will not be pursued. The plan lays out threats to the
individual and enterprise environments, and calls for greater awareness of
threats and the need for public-private partnerships.
While the war on terror continues to run its course,
it come as little surprise that the Federal government would propose some
sort of response to the risk of what has been dubbed a “digital Pearl
Harbor.” While this draft report indicates that at least some level of the
federal government is focusing on the potential threats in cyberspace, the
lack of substantive proposals indicates that much any real meat of a national
strategy has yet to be formulated. In short, this document contains little
more information or insight than most mid-level IT managers possess today.
The administration’s insistence on avoiding any sort
of federal regulations for the private sector comes as no surprise. However,
the lack of meaningful, positive incentives for the private sector to up its
evaluation of — and thereby its investment in — IT security would seem to
leave much of the response to a digital sneak attack in the hands of budget-hamstrung
IT administrators. In short, little will change. Perhaps the brightest note
in the document is the possibility of government investment in IT security,
whether through mandated minimum security requirements for federal agencies
or ongoing research in the academic realm. The former will certainly pay
dividends to the struggling IT sector in the near term; the latter may
provide relief in the long term. Yet, as we review the sum and total of this
proposal, we see it more as a fire drill, one in which the alarm is sounded
by little actual fire-fighting takes place.
|