|
|
IBM Introduces the Baby Mainframe:
Small, Powerful and Linux-Enabled
By Joyce Tompsett Becknell
This week, IBM announced the eServer z800 model 0E1,
the company’s newest entry level mainframe computing solution. While the 0E1
has the same architecture as the earlier z800 and z900 systems, its primary
difference is size. The previous entry level model, the 0A1, has an 80 MIPS
engine while its new sibling comes with a 40 MIPS engine and an integrated
facility for Linux (IFL). The 0E1 uses the new zELC software pricing which
provides aggressive pricing for the zOS designed for z800 models, to reduce
the overall cost of ownership of a mainframe for low-end workloads. Additionally,
the system will support the z/VM software to enable logical partitions and
clusters. The z800 0E1 system is upgradeable to the 0A1, and from there all
the way to a z900. IBM has also decided to let customers convert or downgrade
existing 0A1 systems to 0E1 configurations if that is a more sensible
approach for their workload.
The new z800 0E1 is designed primarily for the
installed base of traditional IBM mainframe customers (i.e., S/390 and RS
6000) users. IBM would dearly love to see all these folks upgraded to the
newer zSeries architecture they introduced last year. While many customers
have made that leap, customers with workloads well below the 80 MIPS entry
level of the 0A1 have been hesitant to pay for MIPS they do not need. In
addition to software savings, offering a smaller, less expensive system
serves as a competitive deflector to companies like Sun Microsystems which
have been pitching their UNIX systems as mainframe alternatives. This new 0E1
system ups the competitive ante by providing mainframe power more competitive
price points. In addition, many customers who have sunk significant
investments into traditional mainframe architecture are curious to explore
Linux. IBM is positioning the z800 0E1 as an option that lets them have their
cake and eat it too. By offering both capabilities on a standard system, IBM
hopes to entice more users to experiment with minimum risk.
We have been interested to watch vendors continue to
blur the lines between computing architectures. It is becoming common to find
the mainframe competing with high-end UNIX offerings, and we assume that will
extend to 64-bit Intel architectures as they move into the mainstream. UNIX
OS flavors have been battling for years now, and Linux versus Microsoft has
become a commonplace battle cry in the fight for infrastructure and low-end
application servers. While Sageza believes architectural evolution is a good
thing, we also believe that IT managers and their vendors are entering a
period of great uncertainty, perhaps unknowingly. Not all operating systems
and architectures make sense for all applications or compute environments. Managers
need to make sure they have a firm grasp of what they really need as they
welcome salespeople into their offices. Vendors need to resist the urge to
treat all situations as a nail in need of their unique hammer. Balancing
support capabilities, the in-house knowledge base, costs of software, and
ability to integrate with other applications will become even more important
than hardware and OS capabilities in the near term. Sageza believes that the
vendors who offer highly flexible solutions such as the z800 0E1 that meet
specific client needs are doing what they need to create loyal customers and
slowly win over the confused.
|
|
|
Back in the Old Same Place
By Jim Balderston
IBM announced this week that it has officially
formed an autonomic computing unit, which will be headed by Alan Ganek,
former vice president of strategy at IBM Research. The new unit is designed
to integrate autonomic computing efforts across the company. Among the
various efforts to be mounted within IBM around autonomic computing will be
the development of a deployment model designed to guide customers through the
process of developing an autonomic computing environment. IBM Global Services
will develop a Resilient Business and Infrastructure Solutions Practice, and
the company will develop autonomic computing design centers to help partners
and customers design and test autonomic computing products. IBM also plans to
move its existing technology product portfolio into the autonomic computing
model. The company said WebSphere Applications Server Version 5.0 will
include features that will allow it to automatically monitor and fix performance
problems. It also said that its Tivoli products already contain more than two
dozen autonomic computing features and that its DB2 database also already
includes self-managing and self-tuning features. The company also said its
Enterprise Storage Server, Code-named Shark, will have ease-of-use
technologies built in. Finally, the company said it will continue developing
autonomic computing features for its PCs.
IBM has been talking quite a bit lately about
autonomic computing. The phrase pops up in various press releases,
presentation slides, and company speeches. Before this announcement, IBM was
in a good position to abandon the concept if it did not gain any traction in
the Tower of Babel that is IT marketing. With this step, one thing appears
clear. IBM believes that the “autonomic computing” concept has got legs.
Formalizing its commitment to the concept by naming people who own the idea
inside the company, and telling the world that this is a strategic direction,
makes it a lot harder to walk away from now. Rarely known as a rash and hasty
outfit, we suspect Big Blue is making this decision on very solid ground.
So what happens now? One could argue that naming
existing products with the first iterations of autonomic computing technology
already in place is proof of actual substance to the concept, or conversely,
simply the repackaging of existing product in a new, shinier wrapper. While
certainly a little of the latter is going on, we believe that the next twelve
months will see some meaningful enhancements to the autonomic computing line,
just as the past twelve have. By promising increasingly intelligent network
and computing components — be they Sharks or PCs — IBM is promising a lot.
Meanwhile, IBM is not alone in this latest search for the holy grail, with
HP’s Adaptive Infrastructure and Sun’s N-1 efforts in high gear. However, if
IBM’s delivery of truly intelligent enterprise IT environments continues on
pace, it will secure for IBM a familiar position in the IT vendor space: top
dog. It may seem like years since that IBM has been acknowledged as such, but
those years were measured only in Internet time, after all.
|
|
|
Intel Puts Its Money Where It Hopes
Its Chips Will Be
By Jim Balderston
Intel has announced that it plans to invest $150
million in companies that develop WiFi (802.11) technology. The Intel Capital
Communications Fund will invest in companies that are developing hardware and
software products and services that create an easier to use, more secure
wireless environment, simpler billing procedures, improved network
infrastructure, new ways to connect to high-speed networks outdoors, and ways
to deliver services over the network. Intel said it has already invested $25
million in more than ten companies in this area. The company also said it
would continue to invest in its Banias chip, mobile computing technology that
has been designed from the start to be applied to mobile PC uses and which
contains dual band — both 802.11a and 802.11b wireless capability. Banias is
due out in the first half of next year.
We have had little quibble with Intel’s Banias
initiative; anything that gives us more WiFi capabilities is just fine and
dandy with us. It should come as little or no surprise that Intel is
investing some of its sizable nest egg in companies that will help drive a
market that Intel execs put at 30 million desktops within three years. It’s
not the first time they have made such investments, nor will it be the last.
However, the play will be incomplete if the effort stops at WiFi. For mobile
computing to reach everyone and change the fundamentals of connectivity, it
will be necessary to complete the circle by seamlessly integrating WiFi with
other technologies so that users seamlessly transition between next
generation wide area and local wireless without affecting content streams.
But that’s for another discussion.
There is little argument that much of the WiFi core
elements — like robustness and security — still need some fine tuning. As
this weekend’s WorldWide WarDrive
will no doubt illustrate, there are plenty of improperly secured wireless
LANs out there. As these geeks roam around cities around the world looking to
identify — but not use — these rather wide open wireless access points, it
will have little impact on the increasing momentum of WiFi deployments in the
long run, regardless of the scale of ineptitude demonstrated by the WiFi
network admins in this weekend’s big driveby. As P.T. Barnum once said, “It
doesn’t matter if you write good things or bad about me, as long as you spell
my name correctly.” The name of this game is WiFi and we suspect few people
will have trouble spelling it in the coming years.
|
|
|
Google Suppresses Sites in German and
French Search Results: Drawing the Line on Information Access
By Joyce Tompsett Becknell
This week, a Harvard Law School report indicated
that sites were missing from search results conducted at the French and
German Google sites. According to the report, most of the missing sites were
those that support racism or deny the Holocaust. The reason for these missing
sites is that both France and Germany have strong laws that prohibit hate
speech, and these sites fall under the provisions of these laws. It is common
for the various search engines to run different sites for different
countries, adapted to the native language and currency. Users would generally
not be aware of a missing link unless they were to perform the exact search
at several locations, such as Google U.S. as well as Google France or
Germany. Restricting access to these sites when in violation of a local law
has become a common practice of most major search engines.
While the idea of blocking offensive material
presents a popular topic for freedom of speech advocates, the problems go
well beyond freedom of expression issues. Google, as well as Yahoo!, Amazon,
and Alta Vista all face continuing threats of lawsuits for allowing
questionable material to be procured through their services. German companies
threatened to bring suit to several companies earlier this year when it was
alleged that directions for how to disassemble components of Deutsche Bahn
systems could be found on line. The Church of Scientology is another organization
that has leveraged lawsuits to stop links to sites critical of the church
based on legal reasons such as copyright violations. The Chinese government
is one of several that restrict the type of information citizens have the
right to access. In the wake of the September 11 terrorist attacks last year,
the U.S. Government considered censoring itself regarding the types and range
of information it made available publicly on the Web through its various
agency sites. The argument underlying all of these examples is that search
engines make it considerably easier for dangerous, valuable, or misleading
information to get to the hands of large numbers of people. Before the Web,
all of this information existed but was much more difficult to disseminate
widely or quickly. The ongoing crawl toward broadband, the increasing
sophistication of search engine technology, and the growth of Web access
means that the entire spectrum of human thinking is available to us with a
few clicks of the mouse. The troublesome question is how to regulate that
access.
It is impossible to remove the offending sources
themselves. Like mushrooms they will proliferate overnight despite all
efforts to remove them. The current compromise is to force the search engine
companies — the providers of the linking technology — to police the Web. Mostly
this functions on a reactive level. Someone reacts to the presence of
objectionable material and the link is duly severed. However, at some point,
limiting access to information becomes censorship and the definition of
questionable material becomes highly subjective. Additionally, it becomes
ever more difficult for search engine companies to maintain the
responsibility for managing Web content through their services. Already some
groups argue that the service providers who host the sites where content is
stored should be responsible for this information. We believe that there are
two complex issues at stake here. First is the difficult question of who
should be able to determine what is acceptable and what is not. The second issue
is the question of who is responsible for monitoring and enforcing these
definitions, assuming anyone can agree to them at all. It is hard to draw
concrete geographic boundaries on a logical structure. It is harder to
enforce on country’s set of rules on the citizens of another when the
information is delivered to them by a company based in a third country,
delivered on equipment owned by a company in a fourth country. Institutions
like the United Nations or even something regional like the European Union
are wholly unequipped to tackle these issues effectively. Sageza expects this
to be an ongoing battle pitting localized political power against the forces
of human expression and technology that will make today’s ongoing World Trade
Organization disputes look like child’s play in comparison.
|
