- Related Stories
-
New hopes from Sun's idea factory
October 17, 2005 -
FAQ: The Sun-Google partnership
October 4, 2005 -
Sun and Google shake hands
October 4, 2005 -
Galaxy remakes Sun's server strategy
September 12, 2005 -
IBM backs Sun's Solaris, renews Java pact
June 27, 2005 -
Solaris engineers offer personal source-code tours
June 14, 2005
When you're trying to convince the business world you're still relevant, it helps to have a former employee now working in the corner office of arguably the hottest tech company in the world.
In a tightly packed press conference at Silicon Valley's Computer History Museum earlier this month, Sun Microsystems Chief Executive Scott McNealy took the stage with Google CEO Eric Schmidt. Back in Sun's glory days, when it introduced the Java software technology and was making a killing on computer servers, Schmidt was the computer maker's chief technology officer.
Now, as Sun attempts to regain its shine after four years of disappointment, it can't hurt to be associated with Google. And the ever quick-witted McNealy was quick to make a self-deprecating jab at the moment.
What's new:
Sun is in the midst of an effort to recover the prestige and financial health it lost when the dot-com bubble burst.
Bottom line:
In a two-part series, CNET News.com takes a look at the new products and plans that could get Sun back on track, and explains why some other areas may not be so easy to fix.
Will a new plan to embrace rather than do battle with open-source alternatives in software and an aggressive acquisition in storage make things any better?
"It's a natural for me to be sitting here with an ex-Sun employee," McNealy quipped. "We have littered the industry with ex-Sun employees."
The focus of their partnership, interestingly, is software. Schmidt's Google endorsed Sun's software philosophy and pledged to help with several projects. Among them: Java, which lets the same program run on widely different computers; OpenOffice, the open-source software suite that competes with Microsoft's Office; OpenSolaris, the open-source version of Sun's operating system. Meanwhile, Google is paying Sun so its Toolbar software can piggyback on downloads of Sun's Java.
But as has so often been the case with Sun's software plans over the years, the partnership talk was long on vision and potential but short on specifics. Unless Sun and Google can provide better evidence, the deal "is more Sun trying to huddle under Google's current market halo than a definitive business agreement that will result in real dollars for either party," said Gabriel Consulting Group analyst Dan Olds.
Truth is, there's always been potential in Sun's software business. The same could be said of Sun's long-foundering data storage equipment unit. But for one reason or another--programmers bogged down by too many projects, poor execution, or cheaper alternatives--sales in both areas have routinely disappointed.
Will a new plan to embrace rather than do battle with open-source alternatives in software and an aggressive acquisition in storage make things any better? Sun President Jonathan Schwartz, who ran Sun's software efforts before taking the No. 2 slot at Sun in April 2004, certainly hopes so. "Everything that Sun produces will be open source and free," Schwartz said in a September speech, with Sun trying to make money by selling support to big businesses.
Two projects, the Solaris operating system and Java Enterprise System, dominate Sun's effort to expand into the software business. Sun hopes both will remake the industry; the JES server software through unusual pricing, and Solaris by building an alliance with open-source developers.
Sun investors
size up McNealy
JES is a broad collection of server software for handling tasks such as managing the digital identities of computer users or running Java programs. Initially, Sun charged each customer $100 per year for each employee, no matter which JES packages were used. Sun tweaked the pricing in January, selling JES subsets for $50 per employee per year or the whole kit and caboodle for $140 per employee per year.
There is some good news. JES subscriptions jumped by 186,000 to 619,000 in the most recent quarter, and a major contract announced in July with General Motors boosted the tally to nearly 1 million. And the subset approach opens new doors, said RedMonk analyst Stephen O'Grady. "Breaking the JES package up into separate packages allows customers to buy just the functionality they require, at a lower cost," O'Grady said.
And this still has the potential to be a good business for Sun. Between 80 percent and 90 percent of customers renew their JES support subscriptions, Loiacono said.
Sun doesn't disclose JES revenue, but 1 million subscriptions at $50 apiece means $12.5 million per quarter--tiny when compared with
See more CNET content tagged:
Sun Microsystems Inc.,
BEA Systems Inc.,
Eric Schmidt,
Scott McNealy,
Sun Solaris
Sun management is talking about services... without doing much about desktop that is requesting those services.
Sun mangment is behind the curve on going direct, and cuting the cost of sale down.
But.. I think SunW broke out of it's trading range and if Java Dekstop (JNLP aka WebStart) gets patched as expected, Sun could own the internet ecosystem.
.V
Sun management is talking about services... without doing much about desktop that is requesting those services.
Sun mangment is behind the curve on going direct, and cuting the cost of sale down.
But.. I think SunW broke out of it's trading range and if Java Dekstop (JNLP aka WebStart) gets patched as expected, Sun could own the internet ecosystem.
.V
>Valley's Computer History Museum earlier this
>month, Sun Microsystems Chief Executive Scott
>McNealy took the stage with Google CEO Eric
>Schmidt.
>
>"It's a natural for me to be sitting here with
>an ex-Sun employee," McNealy quipped. "We have
>littered the industry with ex-Sun employees."
What an _appropriate_ venue.
Sun reminds me of Racad Vadic (remember them?).
When the price of modems went into free-fall,
Racal just couldn't bring themselves to change
their sales model, or reduce their overhead.
For a while, they retained market share by
adding features and by massaging their legacy
customers, but the $200 modem killed them.
Sun has a very hard time changing anything.
It took them 20 years to give up on the idea
of a proprietary microprocessor, and it is
very clear that they will go back to using
the SPARC if they can ever figure out how to
make it go 4 GHz.
Google uses x86 boxes and free software.
I'll believe that Sun has a chance if I see
Google converting their search farms to run
on SPARC hardware, or even Solaris on x86.
Until then, watch the pretty colors:
Sun is setting.
Guest what?
Who every reads this ...we won?t get old enough
to experience ethers MS or Sun's nova
The dynamics are just so huge ?it could even invalidate my statement.
I will admit it , in a sense it?s a roll of the dice !
Whoever knows what is the next big thing to hatch on ..go for it
Just a thought
Regarding his modem comment: isn't it true that Sun just came out with Servers with better performance and lower prices than Dell's and isn't Dell acknowledged as the industry cost leader? Doesn't that seem to weaken his cost argument a bit?
Regarding his "proprietary" SPARC comments: let's see, are Intel chips proprietary? Uh, yup. AMD's Opterons and IBM's Power chips? Yup. So name me a microprocessor design that isn't proprietary? Oh wait, that's right, Sun released the SPARC design to open source some ten years ago. But then, does it really matter whether microprocessors are proprietary? Again, he doesn't really know what he's talking about.
The reality is that Sun has had some great technology over the years but does a poor job packaging it into saleable units and then actually marketing and selling it.
On the other hand, it has a new president now, and despite the Ben-Stein-in-Ferris-Bueller tone to this article, it has a group of products that are vertically comprehensive and technologically pretty darn cool and exciting. On top of that, it's clear that it's approaching it's product line differently now. All you have to do is look at the fact that it'll sell you a computer with Unix, Linux, or Windows on it for confimation.
Can Sun execute? Who the heck knows. Does it have good product? Oh heck yeah. Is Dave Chapman some shill for either a Sun competitor or the immature-like-a-group-of-three-year-olds open source/Slashdot community? Probably. Am I a shill for Sun? No. I'm just another guy like the rest - I don't necessarily like the way Sun's run its business, but I know good ideas and good technology when I see it.
Here?s the site: www.Codonology.com (理子工程学)
Thanks.
Hua Fang
(方华 in Chinese)
>Valley's Computer History Museum earlier this
>month, Sun Microsystems Chief Executive Scott
>McNealy took the stage with Google CEO Eric
>Schmidt.
>
>"It's a natural for me to be sitting here with
>an ex-Sun employee," McNealy quipped. "We have
>littered the industry with ex-Sun employees."
What an _appropriate_ venue.
Sun reminds me of Racad Vadic (remember them?).
When the price of modems went into free-fall,
Racal just couldn't bring themselves to change
their sales model, or reduce their overhead.
For a while, they retained market share by
adding features and by massaging their legacy
customers, but the $200 modem killed them.
Sun has a very hard time changing anything.
It took them 20 years to give up on the idea
of a proprietary microprocessor, and it is
very clear that they will go back to using
the SPARC if they can ever figure out how to
make it go 4 GHz.
Google uses x86 boxes and free software.
I'll believe that Sun has a chance if I see
Google converting their search farms to run
on SPARC hardware, or even Solaris on x86.
Until then, watch the pretty colors:
Sun is setting.
Guest what?
Who every reads this ...we won?t get old enough
to experience ethers MS or Sun's nova
The dynamics are just so huge ?it could even invalidate my statement.
I will admit it , in a sense it?s a roll of the dice !
Whoever knows what is the next big thing to hatch on ..go for it
Just a thought
Regarding his modem comment: isn't it true that Sun just came out with Servers with better performance and lower prices than Dell's and isn't Dell acknowledged as the industry cost leader? Doesn't that seem to weaken his cost argument a bit?
Regarding his "proprietary" SPARC comments: let's see, are Intel chips proprietary? Uh, yup. AMD's Opterons and IBM's Power chips? Yup. So name me a microprocessor design that isn't proprietary? Oh wait, that's right, Sun released the SPARC design to open source some ten years ago. But then, does it really matter whether microprocessors are proprietary? Again, he doesn't really know what he's talking about.
The reality is that Sun has had some great technology over the years but does a poor job packaging it into saleable units and then actually marketing and selling it.
On the other hand, it has a new president now, and despite the Ben-Stein-in-Ferris-Bueller tone to this article, it has a group of products that are vertically comprehensive and technologically pretty darn cool and exciting. On top of that, it's clear that it's approaching it's product line differently now. All you have to do is look at the fact that it'll sell you a computer with Unix, Linux, or Windows on it for confimation.
Can Sun execute? Who the heck knows. Does it have good product? Oh heck yeah. Is Dave Chapman some shill for either a Sun competitor or the immature-like-a-group-of-three-year-olds open source/Slashdot community? Probably. Am I a shill for Sun? No. I'm just another guy like the rest - I don't necessarily like the way Sun's run its business, but I know good ideas and good technology when I see it.
Here?s the site: www.Codonology.com (理子工程学)
Thanks.
Hua Fang
(方华 in Chinese)
Google is known for buying cost-effective commodity machines and linking them into vast clusters. Google's purchase of Sun hardware is the best endorsement imaginable for the cost effectiveness and managability of Sun's servers.
Because Sun is exclusively sourcing x86 processors from AMD, it's getting a special deal: Sun will be the first vendor to get speed bumps in the Opteron. Sun will use its special relationship with AMD to make superior servers that steal market share from its rivals.
There's some risk in this strategy: Intel may close the performance gap with AMD in the next few years, plus it means selling machines for lower margins than Sun is used to -- but it's a strategy that keeps Sun in the game, and it just might get some people into Sun's storage options and Solaris 10. (BTW, Sun is happy if you want to run Linux or Windows on a Galaxy server...)
As for SPARC, Sun's plan is not to get SPARC to 4 GHz, but rather to build highly parallel chips that will smoke anything in the x86 space for server workloads -- this may or may not work in real life, but system-on-chip is where everybody is going for silicon performance in the next fifteen years.
Google is known for buying cost-effective commodity machines and linking them into vast clusters. Google's purchase of Sun hardware is the best endorsement imaginable for the cost effectiveness and managability of Sun's servers.
Because Sun is exclusively sourcing x86 processors from AMD, it's getting a special deal: Sun will be the first vendor to get speed bumps in the Opteron. Sun will use its special relationship with AMD to make superior servers that steal market share from its rivals.
There's some risk in this strategy: Intel may close the performance gap with AMD in the next few years, plus it means selling machines for lower margins than Sun is used to -- but it's a strategy that keeps Sun in the game, and it just might get some people into Sun's storage options and Solaris 10. (BTW, Sun is happy if you want to run Linux or Windows on a Galaxy server...)
As for SPARC, Sun's plan is not to get SPARC to 4 GHz, but rather to build highly parallel chips that will smoke anything in the x86 space for server workloads -- this may or may not work in real life, but system-on-chip is where everybody is going for silicon performance in the next fifteen years.
Fundamental changes are in the works. These changes are not just in technology, but also in business practice and business models. Failures to adapt will once again bear witness the carcasses of many major corporations and their CEO/CIO's plugging the sewers. Cast-iron and stubbornness does not any longer guarantee survival forever (hello, AT&T). But an identity and a corporate culture are unique identifiers from company to company. This uniqueness is what distinguishes a corporation from falling into the ?just-like-everybody-else? recycle bin.
Now, I said all that to say this. Some corporations need to diversify in order to stay alive. Others just need to shift their core businesses and include those that fit their competencies. Still others need to just ride out the tides of change (as long as those tides are not a tsunami). SUN can do that. But, they must be ?open? and ?closed?. Sun should continue to develop a proprietary hardware technology for the embedded market. Thus, chip development (microprocessor, FPGA, etc.) and architecture (SBCs, cPCI, etc) are important competencies to promote. Adopting PICMG standards, Fiber channel and Advanced TCA will give SUN in-roads into markets they haven?t aggressively pursued. SUN must economically distinguish itself from others in the primary embedded development groups, by using the power of its marketing name. And as long as AMD keeps the pressure on Intel for dual-core, 64bit supremacy, the SUN relationship with AMD will make it viable in the x86 space for a long time too.
SUN is on the right path. OpenSolaris is a great step forward. Securing deals (like StorageTek) that strengthen their position in the storage space to compete with the likes of IBM, EMC, etc will ensure some interim term survival. Also, exploiting the opportunity to market and implement an integrated cross-platform management solution is a workable sell for the long term. Also, "Windows" enabled SUN workstations, SUNs office suite of products and SUNs compatibility with Active Directory provides access to once thought to be closed environments. From SUNs perspective DELL is no longer the only viable option.
Now, as far as the "nobody else is using it". Well, this is the most challenging part. But it can be resolved with a little dedicated effort. SUN must apply itself to affordable education, training, college curriculum, third party certifications. The key word is affordable. Interested solution providers and students should not have to turn to only SUN for 100% of this training. Spreading this effort will ensure a new and steady stream of talent and an available pool of bodies for development and organizations to pull from. Only then will reluctant third parties be ready and willing to take on the possibilities and challenges that SUN's technology presents to small and medium sized businesses. It is the SMBs where the bulk of SUNs long term successes lay. For example, SMBs readily adapted to open source LINUX because it is scalable and "affordable". And LINUX went off the charts. There's no reason the same can't happen for Solaris-10. But it must be implemented and marketed with an eye on education/training, the embedded space and the future. The Solaris University Challenge is a great start.
Finally, I'm not a SUN or Solaris suck-up. I'm not star struck and I have no loyalty to any software development group. I do not think it is such a great idea, as someone at SUN obviously does, that marketing your top of the line future servers and hardware as being compatible with FORTRAN and PASCAL. What dinosaur thought of that? Sure, it's compatible. But there are other languages and COTS apps to market that will give OpenSolaris greater and wider appeal. Particularly in the data center and the often ignored telecom space.
Anyway, I too, at first glance, looked at the once-upon-a-time master of their domain hooking on to the coat-tails of the young upstart, as a move laced with desperation. But on second thought, for Scott McNealy and SUNs long term survivability, hooking up with old protégées, like GOOGLE CEO Eric Schmidt, is "good business practice". Especially since up to this point, trying to go it alone against the likes of DELL, HP and IBM has not been working.
Fundamental changes are in the works. These changes are not just in technology, but also in business practice and business models. Failures to adapt will once again bear witness the carcasses of many major corporations and their CEO/CIO's plugging the sewers. Cast-iron and stubbornness does not any longer guarantee survival forever (hello, AT&T). But an identity and a corporate culture are unique identifiers from company to company. This uniqueness is what distinguishes a corporation from falling into the ?just-like-everybody-else? recycle bin.
Now, I said all that to say this. Some corporations need to diversify in order to stay alive. Others just need to shift their core businesses and include those that fit their competencies. Still others need to just ride out the tides of change (as long as those tides are not a tsunami). SUN can do that. But, they must be ?open? and ?closed?. Sun should continue to develop a proprietary hardware technology for the embedded market. Thus, chip development (microprocessor, FPGA, etc.) and architecture (SBCs, cPCI, etc) are important competencies to promote. Adopting PICMG standards, Fiber channel and Advanced TCA will give SUN in-roads into markets they haven?t aggressively pursued. SUN must economically distinguish itself from others in the primary embedded development groups, by using the power of its marketing name. And as long as AMD keeps the pressure on Intel for dual-core, 64bit supremacy, the SUN relationship with AMD will make it viable in the x86 space for a long time too.
SUN is on the right path. OpenSolaris is a great step forward. Securing deals (like StorageTek) that strengthen their position in the storage space to compete with the likes of IBM, EMC, etc will ensure some interim term survival. Also, exploiting the opportunity to market and implement an integrated cross-platform management solution is a workable sell for the long term. Also, "Windows" enabled SUN workstations, SUNs office suite of products and SUNs compatibility with Active Directory provides access to once thought to be closed environments. From SUNs perspective DELL is no longer the only viable option.
Now, as far as the "nobody else is using it". Well, this is the most challenging part. But it can be resolved with a little dedicated effort. SUN must apply itself to affordable education, training, college curriculum, third party certifications. The key word is affordable. Interested solution providers and students should not have to turn to only SUN for 100% of this training. Spreading this effort will ensure a new and steady stream of talent and an available pool of bodies for development and organizations to pull from. Only then will reluctant third parties be ready and willing to take on the possibilities and challenges that SUN's technology presents to small and medium sized businesses. It is the SMBs where the bulk of SUNs long term successes lay. For example, SMBs readily adapted to open source LINUX because it is scalable and "affordable". And LINUX went off the charts. There's no reason the same can't happen for Solaris-10. But it must be implemented and marketed with an eye on education/training, the embedded space and the future. The Solaris University Challenge is a great start.
Finally, I'm not a SUN or Solaris suck-up. I'm not star struck and I have no loyalty to any software development group. I do not think it is such a great idea, as someone at SUN obviously does, that marketing your top of the line future servers and hardware as being compatible with FORTRAN and PASCAL. What dinosaur thought of that? Sure, it's compatible. But there are other languages and COTS apps to market that will give OpenSolaris greater and wider appeal. Particularly in the data center and the often ignored telecom space.
Anyway, I too, at first glance, looked at the once-upon-a-time master of their domain hooking on to the coat-tails of the young upstart, as a move laced with desperation. But on second thought, for Scott McNealy and SUNs long term survivability, hooking up with old protégées, like GOOGLE CEO Eric Schmidt, is "good business practice". Especially since up to this point, trying to go it alone against the likes of DELL, HP and IBM has not been working.