|
GM Chooses IBM Regatta for Global Design
Solutions
By Charles King
IBM announced this week that General Motors has
selected IBM pSeries 690 (AKA Regatta) servers as
the basis for a global supercomputing infrastructure the company will use for
vehicle design applications. The infrastructure will include multiple systems
(including the auto industry’s most powerful supercomputer) at GM facilities
in Detroit, Michigan, Russelsheim, Germany, and Trollhatten, Sweden. According to IBM, the new
infrastructure will increase GM’s supercomputing capacity by a factor of
four, and will qualify as one of the ten largest supercomputers in the world
and the largest in the automotive industry. The IBM infrastructure will be
used to run sophisticated crash simulations, as well as computer-aided
engineering (CAE) applications to analyze structural integrity and quality,
and reduce vehicle noise and vibration. The servers in the infrastructure
will be 32-way 1.3GHz p690s with 2GB of memory per CPU, with a total
processing power of 2.3 teraflops or 4 trillion calculations per second. No
information regarding a deployment schedule, the cost of the deal, or the
total number of systems planned for the infrastructure was included in the
announcement.
While the GM deal qualifies as a major customer win
for IBM, it also illuminates more elemental changes that are worthy of
discussion. In one sense, the agreement demonstrates yet again high
performance computing (HPC) systems’ importance to the automotive industry.
While a great deal of media ink has been spilled on the part high-performance
visualization and CAD applications play in auto body design, less attention
has been paid to more (literally) nuts and bolts CAE applications. As is
mentioned in the IBM announcement, these solutions are being used in part to
reduce vibration and noise, notably improving the driving experience. That is
good news for both automobile customers and manufacturers, because along with
creating a more comfortable ride, these technologies are also dramatically
reducing the time and money it takes to bring cars to market. But it is the
use of these infrastructures in crash simulations that may be the most
revolutionary element of these technologies. The fact is that
government-mandated physical crash tests are complex, costly, lugubrious
processes that require months of preparation and more months of post-event
analysis, factors that can slow the overall effort to improve vehicle safety.
A few years ago, performance and TCO issues made effectively utilizing
computers for crash simulations a pipe dream, but the new IBM infrastructure
could offer a powerful, cost-effective solution for accelerating the
development and delivery of ever safer GM cars and trucks.
The other side of this deal is in how it could
represent a sea change in how and with whom automotive and other
manufacturers pursue their high performance IT needs. Until recently, SGI and
Cray were among the biggest traditional players in automotive CAE, but the
companies’ fortunes have faltered as increasingly sophisticated engineering
applications have become available for other UNIX platforms. Another obvious
player here is HP, but it is unclear at this point is how the company will
proceed with integrating newly acquired Compaq assets into its HPC and
supercomputing efforts. IBM is moving quickly and aggressively into this
space (to the discomfort of both traditional players and current rivals) and
the company’s efforts in areas such as grid and autonomic computing should
play very well in HPC. We were also interested to note that while GM’s new
infrastructure is the first commercial deployment among the world’s top ten
supercomputers (the others are at research labs and universities), that
commercial supercomputers make up nearly half of the world’s 500 largest
supercomputing installations. What does this mean overall? Simply, that as
hardware performance continues to climb and hardware costs continue to fall,
IT solutions that were once out of reach to all but the wealthiest few will
be enjoyed by increasing numbers of enterprises. From what we can see, that
could mean that GM’s new IBM supercomputing infrastructure might eventually
be remembered as a driver for a new business revolution.
|
|
¿Donde
Está el Netscape?
By Clay Ryder
WebSideStory has announced
its latest Web browser usage data which reports that the percentage of
Netscape browser usage is now at it lowest point in several years. As of Aug.
26, 2002, Netscape’s global usage share had dropped to 3.4%, down from 13% at
this time last year, according to WebSideStory’s StatMarket report, which aggregates data from millions of
Internet surfers per day. Microsoft Internet Explorer now has a global usage
share of 96%, up from about 87% a year ago. Netscape fares better in some
countries such as Switzerland, where its usage is double the global average,
as well as Germany, Canada, and the United States.
This story is punctuated by a set of realities, many
of which have been historically pooh-poohed, but which have once again proven
their endurance. Reality number one is that Netscape, as a software vendor,
was never a long term contender in the market — its product was simply too
important to not become part and parcel to operating environments for both
the client and server. Reality number two is that operating environments are
largely controlled by Microsoft, and it should be no surprise that such an
operating environment extension would be cast under the spell of the Redmond
Giant. Reality number three is that AOL-TW has become the poster child for
lack of clarity, purpose, and execution. Reality number four is that despite
Sun’s protestations, iPlanet was simply a
convenient way for the company to gain access to software to bolster its
operating environment and platform.
So where is Netscape? Although AOL’s release this
week of an updated Netscape browser and the company’s suggestion that the
latest version of its software will replace the IE-based browser with a
Netscape derivative, this seems to come years too late for meaningful market
share impact, and also begs the question of whether or not AOL was all that
concerned about which browser its customers used. Since AOL is all about
content and $21.95/month connections, like many other aspects of the AOL-TW
merger it remains unclear what the corporate leverage for a Microsoft-free
browser would actually be. So while some may dig in their heels, “fight the
good fight,” and refuse to ever use IE, the reality is that this assemblage
of like-minded citizens will rapidly cease to have any commercial impact.
Just as in the early days of broadcasting, when RCA owned radio stations and
three television networks, and also manufactured radios, we are witnessing
the reality that in a standards-based marketplace, whose radio (or browser)
one uses to access content is largely moot. It’s what’s on the radio that
counts. AOL seems to understand this. But do not shed too many tears, as
Netscape is alive and well, as a brand, portal, and reminder of good old
times gone by. One can still find its software and descendants on Sun
equipment (an always amicable relationship), but we believe it is wise to
face facts and admit that Netscape is simply not a commercial software
company anymore.
|
|
Déjà vu All Over Again
By Jim Balderston
Elitegroup Computer
Systems of Taiwan has unveiled the latest attempt at low-cost computing: its Desknote line of portable computers. The Desknote line looks and acts like a portable computer,
with one notable difference: it has no battery and therefore must be plugged
in to be used. Other interesting elements of the system include its chip
selection: Elitegroup uses desktop chipsets from
Intel and AMD. Elitegroup offers external battery
packs and has priced the Desknote computer at
around $1,200, substantially less than portables using mobile processors.
The impetus behind this idea lies largely with the
idea that most notebooks spend a vast majority of their working time on
desktops, docked into a bay with power, peripherals, network connections, and
monitors. Perhaps so, but an awful lot of computing is still done on
airplanes, in waiting areas and away from a convenient power source. The real
value of today’s high-performing laptops is the inherent choice they give the
user about where and when the computing gets done. Further raising our sense
of skepticism is the whole idea behind decontented
computing form factors that, by doing away with capabilities, undercuts the
price thresholds for similar computing experiences. The problem is that they
don’t actually create the same experience.
Remember the much-hyped Network Computer? It was
going to have minimal onboard computing power, virtually no storage, come in
at the magical $500 price point, require permanent Internet connections, and
would be sold by ISPs. Consumers were supposed to flock to it in a time when
desktops were hovering at near twice the price. But the NC was slow to
market. So slow that regular, full-featured PCs began to drop through the
$1000 barrier, then through $700 and so forth. Worse than the pricing issue
was the fact that NCs did not actually provide the
same computing experience as the desktops they were supposed to replace.
Consumers and corporate users shunned the idea of being relegated to
second-class general-purpose computing. While low prices may initially
provide some appeal for these portable desktop units, we suspect that history
will repeat itself, and something in this price range — likely full-featured
laptops — will render this product largely moot in the market. Consumers
don’t flock to products that take valuable capabilities away. They become the
drivers of markets around products that are new and innovative, providing
them with something they perceive as valuable at a price they can afford.
|
|
HP, Dell Announce WordPerfect
Agreements
By Charles King
In separate and unrelated announcements, Corel
revealed that both HP and Dell have chosen the company’s WordPerfect 10 word
processing package and Quattro Pro 10 spreadsheet solution to be included as
part of the WordPerfect Productivity Pack on some models of the two
companies’ desktop and laptop PCs. Starting in September 2002, WordPerfect 10
and Quattro Pro 10 will be pre-loaded on Dell’s recently announced Dimension
2300 desktops and Inspiron 2600 notebook computers. The two applications are
currently available on the company’s SmartStep 200N and 250N notebooks. Also
starting in September, the WordPerfect Productivity Pack will come pre-loaded
on every HP Pavilion computer sold in North America. No licensing or cost
issues were addressed in the announcement.
Though not the biggest IT news to come down the pike
of late, these two announcements point to a pair of deeper issues we think
are worth a few words. First, notice the two vendors who have decided to
replace Microsoft’s venerable Works suite with Corel Productivity Pack. By
dint of its merger with Compaq, HP recently took the lead in PC sales away
from Dell, which had somewhat less recently taken it away from Compaq. While
the percentages required to accomplish such a change are tiny, first place
bragging rights are deemed crucial by all players involved. Good enough but
remember also just how thin the profit margins have become in the consumer PC
biz. Dell’s ascendancy has been fueled largely by the company’s remarkable
ability to continually squeeze nickels out of its supply chain to an almost
magical degree. The continuing dicey economy and a bruising merger battle
have left HP counting its pennies. The roughly $30 difference in price
between MS Works and the Word Perfect Productivity Pack is unlikely to break
anyone’s budget, but is a notable sum to trim from the asking price or add to
the bottom line of millions of $500-800 home PCs.
So much for the monetary side of this equation. At
the same time HP and Dell are looking for whatever financial/market advantage
they can find, Microsoft has been smarting from highly publicized battles
resulting from its ongoing anti-trust wars and controversies surrounding the
company’s new software licensing policies. In other words, it is not a bad
time for PC manufacturers to offer non-Microsoft options to customers who
want them. Does HP and Dell’s embrace of WordPerfect have any chance to
injure Microsoft substantially? Pardon us if we giggle. From a purely
practical standpoint, these deals amount to small change on a per unit basis.
But in a market this tough, any and every edge is a good edge for HP and
Dell. On the other side of the aisle, Corel stands to dramatically increase
its market share simply by not being Microsoft. Not a bad week’s wages for a
company many have long considered down and out, if not forgotten.
|
|
What Are the Odds?
By Jim Balderston
News reports indicate that a number of cash-strapped
states are considering — seriously considering — expanding legalized gambling
opportunities as a means to generate more revenues. New York has joined a
regional lottery, and is allowing casino gambling at a handful of local
racetracks. Both Pennsylvania and the state of Washington have joined
multi-state lotteries while Midwest states are raising taxes on existing
riverboat casinos. Arizona, Tennessee, Idaho, and Nebraska have ballot measures
scheduled for this fall that would expand existing gambling options. States expect
budget deficits to continue for the coming year; this past year forty-six
states had deficits totaling $40 billion.
The expansion of gambling — especially casino
gambling — is no new phenomenon. But with state budgets mirroring the deficit
spending of Washington D.C., it appears that various forms of gambling
establishments are going to make headway in states that once turned up their
noses at such activities. While the present considerations and ballot
initiatives focus on physical sites, we can’t imagine that online gaming will
remain off limits indefinitely to many of these cash-starved states. We
predict that state-operated online casinos can’t be all that far off in the
future. Instead of just collecting taxes, why wouldn’t these states want to
reap the full benefit of the house’s odds? While casino developers will lobby
furiously for laws banning such enterprises, we suspect that the online
casinos — complete with the real-life stimulation of wagering and in most
cases losing hard earned cash — will not siphon off much business from their
real-world counterparts. Instead, it’s likely to bolster and encourage those
who are mathematically challenged not only to visit real casinos, but to do
so more often. Online gambling — like lotteries and real casinos — are
essentially a tax, and one that more and more people seem ready and willing
to allow. Using online adjuncts to the physical casinos will in all
likelihood cement or strengthen the bond between the state government and its
voluntary tax base, bringing more revenues into the physical casinos than
otherwise might appear. Can we expect such clarity of thinking in the
gambling industry — or statehouses for that matter? Probably not. But who
knows? With tighter state budgets acting as a coercive incentive, the odds
could change.
|