An influential Wall Street analyst warns that unless Sun Microsystems cuts up to 19% of its workforce and tightens its focus it will continue to decline and eventually become a takeover target for its large installed base. He recommends that Sun spin off its Java software and reign in Scott McNealy’s brash tendencies.
I think that what all this fuss has been about is that SUN is suceptable to a ‘hostile’ takeover by shareholders.
A lot of thing are happening here right now and this is very important because SUN is on the verge of helping the open source community to accomplish all of it’s goals and at the same time SUN will be able to profit through volume sales.
The most ironic thing about all of this is that the open source community has won. We have won the war, ‘and nobody even realizes it’.
SUN’s new license levels the field. It brings value at a fair price to the customer for the first time in the history of the computer industry. And the platform (Java) is accessible on Linux for students who need a path to migrate from College to their job.
This crap about Sun being taken over is getting as old as Apple going out of business. Sun isn’t going anywhere. It’s been proven time and time again that Wall Street analysts never know anything about the tech industry.
A man can dream though, a man can dream! 8)
Scott McNealy should ride off into the sunset… He helped build Sun, but now give the reigns to somebody else. Just like at MS. Once you stay too long at a job you loose focus and do not understand what is important.
For example Apple, Jobs had to be kicked out. When he came back he came up with some REALLY interesting ideas. One could argue Jobs did not change, but deep down Jobs knows he changed.
Which would be much more likely, between Oracle and Sun… the combined company could then be a “database solutions” provider and make volume sales of Oracle equipped V440 systems which sell as a turnkey midrange database solution, while slowly reducing support for systems from other vendors trying to coax their customers onto new Oracle/Sun systems. By undercutting the competition in this manner they would force IBM into a pricewar, causing them to lower DB2 prices. IBM would most likely promote DB2 on Linux/x86 in such a case, trying to match the Oracle/Sun price/performance and ease of use of an integrated solution.
Okay, that probably won’t happen, but I do forsee a potential Oracle/Sun merger…
I can say that if that analyst thinks that a company that has dumped so many of it technicians can take another 19% hit, then he doesn’t know what he’s talking about.
Sun already struggles to take care of it’s support contracts. How will that be done if they cut loose even more engineers?
Stupid economic analysts always thinks that getting rid of people helps. Hey, they need *people* to do the job, blockhead!
Let McNealy go instead…
Richard Gabriel, a distinguished engineer and scientist at Sun Microsystems Inc., of Santa Clara, Calif., who said he runs a lab at Sun that studies open-source software for the Unix systems company, said, “We noticed the same thing. The number of developers is quite low. So we try to create markets for both open-source and proprietary software in Java.”
The speakers shared a panel called “Innovation or Commoditization: Can Open Source Meet Technology Innovation?”
…
Sun’s Gabriel offered that Sun has a theme that “innovation happens elsewhere” so “we see open source as an innovation pump.”
Indeed, he said Sun also uses open source for marketing. “It’s our dirty little secret,” Gabriel said. “We tell things to developers on the Sun side so it’s coming from developers instead of from marketing, so it’s more believable coming from geeks.”
http://www.eweek.com/article2/0,4149,1306269,00.asp
The analyst may not be correct in predicting Sun’s imminent doom. However, a broad policy of a corporate power lying to their users, their so-called friends, and their customers is the mark of a company that must be brought to justice.
The technology industry is nearing an inflection point. It is only fear that says Sun won’t make it. From where we are today, it is hard to say who will make it and who won’t. Only time will tell if Sun was aggressive enough in bringing new value-centric technology to market. The old corporate culture of “take as much money as possible from the customer” is one thing that will be gone after the inflection point. So Sun has some choices to make if they are going to survive.
“Stupid economic analysts always thinks that getting rid of people helps. Hey, they need *people* to do the job, blockhead!
Let McNealy go instead…”
Right on !
I know the stock market is the last healthy industry america has left – but it’s like a damned corrupt casino. All wall street analysts care about is the stock price – that can’t be the final criterion.
Stupid economic analysts always thinks that getting rid of people helps. Hey, they need *people* to do the job, blockhead!
Kind of reminds me of that IBM commercial, where the guy is sitting is having his dream of the ‘magic cost cutting sword’ anaylized by a shrink. Then, the shrink responds with: “Maybe there are no more costs to cut. Perhaps you should be more producitve.”
Productivity? Shocking!
Wall Street anaylsts always seem to get fired up when a company goes through layoffs. Conventional wisdom is “oh, look, there taking peoples jobs away. That means profits will increase. Yeah!”
Dumb asses… IMHO, I think the corporate finanical machine has gone un-checked for far too long. Most of these people are so out-of-touch-with-reality(TM), they wouldn’t know a good idea if it slapped them in the face!
Oracle…you’re next!
SUN needs to do something! First thing get rid on Mcneally! then come up with a valid business strategy.
They obviously believe that Solaris is a great product, but bad mouthing Linux is going to do nothing but piss of its stalwart followers. Something McNealy has successfully done! If SUN gave customers a choice of Linux or Solaris then customers would be able to discern for themselves which they want to purchase. Many techno geeks don’t know what differences there are in OSs, they just know that Linux is cool. When SUN bad mouths their favorite OS, they are not likely to buy from them. Instead SUN should say that we prefer Solaris, but the chioce is yours and we fully support both Operating Systems.
McNealy is way too abrasive, he has already become the clown of the IT Industry.
Also, SUN should definately get rid of the cruft, they should have dumped SPARC and hardware development a long time ago and concentrated on Solaris and Java. They could have Fujitsu/Seimens do their hardware. If SUN would concentrate on specific things and make sure they were unstoppable, then they would have a chance. Otherwise, they will be the next dinosaur of computer technology.
He needs to go.
Analysts should be outlawed. They hurt the industry, since mostly they have no clue what they are talking about.
I gotta wonder about people who suggest things like Sun should spin off Java. And that HP should spin off its printer division.
What are they going to say next? MS should spin off the Windows and Office divisions?
I think 19% is way toooo low. Wish Wall Str would lay off 100% of their analysts instead. Anybody remember all the shit they pulled before the bubble burst, hyping stocks & pushing it on the rest of us when they was running the other way!
They don’t know ____ about technology. But thats the world we live in, no nothings can rip into companies whether good or weak. Wall Str is about less & less choice.
my 19c
There is a petition to spin java off as a separate sister company from Sun:
http://www.petitiononline.com/spinjava/petition.html
More information:
http://www.freeroller.net/page/kalimantan/20031003#the_first_brave_…
SUN spins off Java as an IPO and voila, some analysts company makes money out of the IPO. Its the same crap back when AT&T spun off their wireless division. It never made any sense then and still doesn’t make sense now.
They obviously believe that Solaris is a great product, but bad mouthing Linux is going to do nothing but piss of its stalwart followers. Something McNealy has successfully done! If SUN gave customers a choice of Linux or Solaris then customers would be able to discern for themselves which they want to purchase.
You can do that right now. They would *PREFER* you to have Solaris x86 installed, however, if you want Linux, they have no problems getting it pre-installed. Also, SUN is the only vendor out there who are 100% providing a Linux desktop solution. Where is IBM on this front? If SUN’s plan is so bad, why do they have 100 new copies on their waiting list with Telstra, a former Microsoft haven, now migrating to Madhatter. Also, any financial reports that have been released don’t take into account any of the deployments that are either taking place or about to.
Many techno geeks don’t know what differences there are in OSs, they just know that Linux is cool. When SUN bad mouths their favorite OS, they are not likely to buy from them. Instead SUN should say that we prefer Solaris, but the chioce is yours and we fully support both Operating Systems.
Lets all remember that these techno geeks have no say in the procurement process. People who make the decisions are CIOs who look beyond the religious rhetoric and decide on what their requirements are versus what the vendor can provide.
Up until recently, Microsoft has been the only vendor to provide and end to end, desktop to server solution for the enterprise. SUN has now realised this and is doing the same thing using Solaris and Linux. Next year, for example, SUN is releasing an updated version of StarOffice which includes J2EE server integration so that collaboration between employees can be made alot easier, aka, “Sharepoint”.
If SUN can provide a better and cheaper solution than Microsoft, with all the features that their customers want, I see no reason why it wouldn’t succeed.
McNealy is way too abrasive, he has already become the clown of the IT Industry.
How is that any different to what Balmer or Alchin say at their regular hypefeast, ala MSDN Conference which has a regular “bash the competition”, or the regular “press conferences” where various managers from Microsoft jump on the “bash Linux” bandwagon.
Also, SUN should definately get rid of the cruft, they should have dumped SPARC and hardware development a long time ago and concentrated on Solaris and Java. They could have Fujitsu/Seimens do their hardware. If SUN would concentrate on specific things and make sure they were unstoppable, then they would have a chance. Otherwise, they will be the next dinosaur of computer technology.
UltraSparc is a great architecture, however, they fail for these reasons:
1) TI is still their main supplier. Get rid of them and get UMC and TSMC to frabricate the chips and motherboards.
2) Outsource server and workstation assembly. Why assemble workstations servers in the US, the 3rd most expensive country to do business in when you can get the systems assembled in china and the chips/motherboards created in Taiwan?
I’ll put money on it if they scaled down their US presence to one office, outsourced all their hardware assembly and ripped the cossie contract of TI, I think you would find that SUN’s costs would drop dramatically. I would then move to have a look at the possibility of scaling down the US call centre and locate it in a region with cheaper labour and English as the first language, either Australia or New Zealand, I personally would go with New Zealand as they don’t have the same union issues that Australia has. After the slash ‘n burn I propose, you would essentially have in the US a software and hardware development.
As viable alternatives develop for those, essentially all that would be left in the US is a skeleton operation providing services and maybe a small inventory for customers who may require hardware replacement within 24hours.
Remember SGI (now sgi). They sold their 3D products in low volume for high margins. They produced truly revolutionary technology. Now they are all but dead. Brought down by the commoditization of their technology (by 3dfx, nVidia, and ATi).
Now we have Sun. They sell their servers in low volume for high margins. Their sales are crumbling as Intel/Dell eat away at their low-mid range. Linux is quickly catching up with the capabilities of Solaris.
If AMD can get people buying Opterons, I think they pose a huge threat in the mid-high range given the 64-bit capability and good multiprocessor scaling of the Opteron architecture. Intel could even be a threat in the mid-high range eventually if they start selling lower-cost Deerfield or eventually add 64-bit to their x86 line.
Sun isn’t dead yet, but their current strategy isn’t working. They are currently competing on too many fronts with “commodity” technologies. Sun is holding on too tight to their own technologies. They currently have a large customer base, but aren’t providing them with the products they want/need. Embracing non-Sun technologies like Linux or non-SPARC processors is dangerous for Sun. Unless executed well, it will only hasten their demise. However, the alternative is a slower but even more certain death.
Now we have Sun. They sell their servers in low volume for high margins. Their sales are crumbling as Intel/Dell eat away at their low-mid range. Linux is quickly catching up with the capabilities of Solaris.
How can they be a high end, low volume business when their Solaris x86 based servers are cheaper than what Dell or IBM sell? their dual UltraSparc III is competitive to Dells dual configurations. SUN’s products and pricing don’t suck, their marketing does. When was the last time you saw an advert in a magazine typically read by management?
If AMD can get people buying Opterons, I think they pose a huge threat in the mid-high range given the 64-bit capability and good multiprocessor scaling of the Opteron architecture. Intel could even be a threat in the mid-high range eventually if they start selling lower-cost Deerfield or eventually add 64-bit to their x86 line.
Again, how is Opteron a threat when SUN has made it well and truely known that they’re behind Opteron 100% and that they are porting their 64bit version to Opteron and will be available next year and again I stress, will be cheaper than Red Hat Linux Enterprise.
Sun isn’t dead yet, but their current strategy isn’t working. They are currently competing on too many fronts with “commodity” technologies. Sun is holding on too tight to their own technologies. They currently have a large customer base, but aren’t providing them with the products they want/need. Embracing non-Sun technologies like Linux or non-SPARC processors is dangerous for Sun. Unless executed well, it will only hasten their demise. However, the alternative is a slower but even more certain death.
Their strategy works, their marketing sucks. If more people know of their Solaris x86 offerings, their future Solaris x86 plans and what their partners were doing, I think every man and his dog would by a Solaris x86 equiped server off SUN.
I like SUN’s new strategy because it accomplishes many of the goals of open source and delivers value to the customer. The problem is that the open source community doesn’t realize that they won. SUN has began a course of action that will level the industry and protect Linux from interference by the state, but nobody knows that we won yet.
And if SUN falls prey to a hostile takeover bid by shareholders, than the industry could be trapped in it’s current state, with products being severely overpriced and customers having no choices.
I think Sun lost its way a long time ago. McNealy et al don’t seem able to decide if they’re a hardware company or a software company. The idea, pre-Java, was that Solaris shifted the Sparc boxes.
However, now we have tight hardware margins, cheap commodity (intel) boxes muscleing in, and nowhere left for the dot in dot com to go.
There seems to be no clear strategy for Java (branding everything hardware and software as the Java Platform just isn’t making sense to me), no clear strategy for Linux (talk it down one week, offer it to your customers the next), no clear strategy to deal with the the threat from commodity boxes and Linux on Intel (buy Cobalt and end of line its products).
The writing was on the wall back in 1998 when Sun purchased Forte Software in a multi-million dollar deal. Anyone who knows what Forte was will realise the potential there was to avoid the inelegent mess that J2EE has turned into.
Forte was a system for rapidly building distributed systems, with failover and loadbalancing working out of the box. The gem in its crown was a technology called Application Partitioning which allowed developers to write and test code on one machine and partition (distribute it) to all machines in the target environment with one mouse click. In many ways it was ahead of its time.
I think Forte is called Sun Unified Server these days and it’s due to be end-of-lined within the next two years. Sun are hoping that it’s customers will “upgrade” (downgrade) to J2EE. They’ll probably move to .NET…
Really, Sun needs a clear strategy and clear direction.
However, now we have tight hardware margins, cheap commodity (intel) boxes muscleing in, and nowhere left for the dot in dot com to go.
The only place x86 is “muscleing” in is on the low end (although “muscleing” in on whom is another matter, as Sun doesn’t compete in the low end market) For midrange servers (what I’d consider midrange… Sun would consider the low-end) the V440 has Dell beaten on price/performance, and on the high end there really is no x86 competator.
Linux zealots spewing anti-Sun FUD certainly gets boring and repetative…
It seems like they are begrudgingly offering x86 and Linux. Are they offering 64-bit support on Opteron (not being sarcastic, I just don’t know)? They also seem to be waffling on Linux. I just don’t get the feeling that they are really behind anything but Solaris on SPARC. You are right that this is more of a marketing issue, but marketing is a really importation part of their “strategy” (remember Digital – great technology, poor marketing).
It doesn’t take a rocket scientest to know that IT shops are crippled, at one point in time there was a lot of money to throw around. Training, trips, large hardware purchases, etc. Now, it seems difficult to even get replacement servers for equipment that is EOL. Intel has a reputation now as cheap and reliable server hardware. SUN can’t even compete on 2-4 way systems from Intel, its last bastion of hope is large 16-128 way cpu systems. Unfortunately, no one has that kind of money.
SUN woke up from their dream to realize no one could afford what they were pushing, yet they insisted on doing things the same old way. What they should have done was push SPARCs harder, spend more money, spend less time evangalizing, crying, and pointing fingers and more time pushing out as many SPARCS as possible (for a loss if needs be.) It seems to me that the only companies that understand the benefits of market saturation are the companies that are going to survive. They should have pushed SPARCS like crazy, lowered their cost and increased saturation.
In their current position their only hope is to push their OS and Java as hard as they can. This means drop the hardware gig, you lost, its over. Unless their is an amazing SPARC within the next year that blows everything away (INT and FP) in dual and quad configurations for CHEAP, then they are lost. It seems likely that this would be highly unlikely if not impossible.
Given that, why not understand what is going on with the market and stop developing your beaten hardware and start porting Solaris to every damn platform out there. Have an Alpha, and Itanium, a PowerPC version, no problem. We have an OS and apps that will run on all of them, across the enterprise… I think they missed the boat, if they would have done this a year ago and they were releasing versions of Solaris right now on every platform you could think of, I think they would be in a lot better position. What hardware manufacturer wouldn’t support Solaris if it wasn’t a niche OS? If they have the mindshare with Solaris then JAVA would be the next logical step for an IT shop. As it stands .net is looking more attractive to CIOs, in particular after the progonstication of their imminent death on Wall Street.
SUN can’t even compete on 2-4 way systems from Intel
FUD FUD FUD… please show me an x86 vendor that can offer a system equivalent to the V440 that has better price/performance:
http://store.sun.com/catalog/doc/BrowsePage.jhtml?cid=104994
(and let me tell you already that Dell is not the answer)
I’m so very tired of this drivel… back up your arguments with numbers, please. Anything else is mere FUD.
In their current position their only hope is to push their OS and Java as hard as they can. This means drop the hardware gig, you lost, its over. Unless their is an amazing SPARC within the next year that blows everything away (INT and FP) in dual and quad configurations for CHEAP, then they are lost. It seems likely that this would be highly unlikely if not impossible.
You’re describing the Niagra, an SMT multicore SPARC processor.
As for pricing, Java Enterprise provides a complete IT solution for any company at $100/employee/year. Given the current cost of service contracts/on-site support, hardware and software upgrades, etc. this seems like quite a bargain. Can you show me an x86-based configuration which would come to such a price, including the cost of service contracts, which comes to the same price.
To Bascule:
Now that is easy to find. Just go to
http://colfax-intl.com/jlrid/SpotLight_more.asp?L=71&S=1&B=624
and you can confdigure a 4-way opteron with 16GB ram for only 16541$. At least as fast as the sun box. And that includes Two hot-pluggable 64bit-133MHz PCI-X slots & three 64bit-66MHz PCI slots. I could not find out what kind of pci the sun box had, but it is NOT pci-x so the opteron also have a faster extern bus.
Martin
The way I see it it, Sun was a victim of their own success. Unfortunately their success wasn’t as big as the cash-beasts Intel and Microsoft. Sun’s market has never been consumers, the people who read email and browse websites and play games, but always professionals: developers, graphics artists, government, data centers, ISPs, infrastructure. They never had any desktop solution, only a workstation solution.
Unfortunately what they didn’t see is how big the market for desktops and consumers would become, and how much money there was to be made there. Intel and Microsoft capitalized on the (relatively) low margin, but vastly larger consumer markets while Sun enjoyed success with infrastructure. The CPU battles starting around 1997 between AMD and Intel set off one of the fastest growth periods for microprocessor innovation ever. Intel had huge cash reserves and lots of experience, different constraints, and fierce competition that drove them to become very efficient, very agile, and very very smart.
Sun had it so good for so long (the dot com boom) that there was a lull in pushing the state of the art. That lull more or less coincided with the early part of the serious CPU battles between AMD and Intel. While Sun was making money hand over fist on huge machines that were getting bigger and bigger, AMD and Intel were making money hand over fist (well at least Intel was) beating the hell out of each other with faster and cheaper processors. Sun’s microprocessors languished. Software development languished. Whatever the strategy was, it didn’t matter. Sun was basking in, well, the sun.
Of course there were always newer processors in the pipeline at Sun, but without pressure for them to succeed, there were several F***-ups that left Sun at a strategic disadvantage in terms of microprocessor performance. This has left them with microprocessors that have performance laughable in comparison to x86 processors that cost 1/10th the price.
Another big problem is that Sun doesn’t have enough volume to bring down their per-CPU manufacturing costs. They essentially have only one customer for their microprocessors–themselves. And they have only a tiny tiny fraction of the volume that x86 does. Intel has lots and lots of customers to pay for its R&D costs, and their fabrication plants are state of the art. Sun has to pay people like Ti and Samsung to manufacture their chips–both of which don’t (to my knowledge) have the technology that Intel has or the volume required to cover the cost of state of the art technology.
Sun has a HUGE disadvantage when it comes to microprocessor performance. This was a problem 5 years ago that went undetected because of such strong sales. The result is that today there are seriously behind in single CPU performance. Of course they build much bigger systems to compensate for this, which then come at a much higher price. But the psychology of how slow each individual processor weighs heavily on people’s minds, whether it is a real issue or imagined.
So now you see a company with lots and lots of brains, but many still have the delusion that the things will somehow go back to how they used to be, and that huge servers will sell like hotcakes again. Finally–finally, the strategy has changed. People at the top finally realized that the desktop matters, and that selling other people’s hardware can actually reduce costs. Software matters, desktops matter, users matter. PDAs and cellphones matter. Finally Sun is getting it.
You are just seeing the first of what I hope will be a revolutionary new strategy that will transform Sun into a major contender. Finally they have a full solution, desktop to server, with all the software there. And they have the simplest pricing and licensing model that can be found. They are rapidly turning into a real solutions provider, instead just a big server company. I think that they will find it is easier to go from building big things to building small things than vice versa.
I feel that they might be at a disadvantage right now and be getting beaten up pretty badly, but with Solaris x86, Opteron, and the Java desktop, the puzzle is finally complete.
Some analysts get paid to sound like they know what they’re talking about and spit out meaningless numbers even though they have no idea what the underlying technology they talk about does.
For SUN, throwing away its SPARC division would be akin to committing suicide.
No offense but SPARCs are just CPUs, processors do the same freakin’ thing no matter who makes them. It would cost SUN dramatically less to act as a middle man then further develop their very low performance SPARCs. They should send their engineers over to Fujitsu. They can buy their processors and hardware from Fujitsu and develop ports to at least the Itanium for Solaris. This would sure future-proof their OS which seems to be tied to a dinosaur that they aren’t will to let go. They really need to embrace other CPUs or they are dead. (I know some of your turning over in your graves, but at least SUN would be around… SPARC really has fallen down, and hard.)
They still have some clout on the high end, if their was a Solaris port for Itanium and PowerPC I feel that in and of itself would save SUN. Having multiple ports would make their OS available on any hardware, if they could maintain software across those platforms (as soes SuSE) then they have a shot at competing with Linux. But they need to wake up and realize that Linux is here to stay.
Intel shouldn’t be their enemy, they went toe to toe with Intel and they lost, they should accept that at this point stop skulking. Microsoft is who they should be going after. IBM is pushing Linux, SUN should be pushing Solaris and Linux. They are competing with each other, but honestly if they threw their full weight behind Java and Solaris they may accomplish a great deal. Also, imagine a Solaris AMD64 bit version, it could run x86 apps and x86-64 apps. They could also FULLY support their software on LSB and compete head-on with IBM.
To Bascule:
Now that is easy to find. Just go to
http://colfax-intl.com/jlrid/SpotLight_more.asp?L=71&S=1&B=…
and you can confdigure a 4-way opteron with 16GB ram for only 16541$.
Apparently you fail to understand the concept of a system vendor… this Colfax place appears to be an assembly service, purchasing and assembling off-the-shelf parts. They offer no warranty on the systems at all, and instead leave it to you to deal with the individual warranties offered from the individual component manufacturers:
(from http://colfax-intl.com/cfx/TermsAndConditions.asp)
Colfax installs and assembles component parts produced by unrelated manufacturers. Colfax makes no warranty, express or implied, as to defects in material or workmanship of those components. Colfax shall not be liable for any inability or failure on the part of manufacturers to make good on their warranties. Colfax warrants that any assembly and installation services that it performs (“Services”) shall be free of defects in workmanship for a period of 30 days from the date of delivery, provided that the Products are operated and maintained in accordance with all instructions provided, are operated under normal use, and the proper Return Merchandise Authorization (“RMA”) procedures are followed in accordance with Paragraph 5.
Colfax warrants that any assembly and installation services that it performs (“Services”) shall be free of defects in workmanship for a period of 30 days from the date of delivery, provided that the Products are operated and maintained in accordance with all instructions provided, are operated under normal use, and the proper Return Merchandise Authorization (“RMA”) procedures are followed in accordance with Paragraph 5.
I can’t magine anyone who would be willing to make a $16,000 hardware investment that comes with virtually no warranty.Perhaps you should find prices from a vendor willing to stand behind their products. From that description they could ship you a system with faulty processors and RAM and not be accountable for it… you would have to deal directly with the vendors of those individual components.
Sun offers a standard 3 year warranty on all parts with a free parts replacement service… call them up and request a replacement and it will be there the next day. Sun also provides on-site service as part of the warranty, for both hardware and software.
Unfortunately Sun has not yet published SPEC CPU2000 numbers for the V440, and I can’t find any for Opteron 840 based systems (which is what you must’ve used for the $16k quote), but I have a very hard time believing a quad 1.4GHz Opteron system could outperform a quad 1.28GHz UltraSPARC IIIi system.
Let’s not forget the poor state of affairs of Linux on Opteron at the present point in time… for database use the V440 would crush this system if it was used in 32-bit mode. As for using a (currently rather immature) AMD64 build of Linux, Oracle has not yet released 9i for Linux/AMD64 (although there is a “developer preview” version) Compare this to Oracle/Solaris, which is a decade mature 64-bit database solution.
Find me numbers for a place that offers at least a year long warranty that offers parts and on-site service… either that or find me someone who has purchased such a system from Colfax with absolutely no warranty on parts.
Anybody that loves OpenSource and hates Microsofts needs to support SUN. They have done more to fight MS than all other proprietary groups combined.
OK… got this right from spec.org
highest posted scores for each cpu on FP2000 Rate Result
(Floating point is the strenght of SPARC, Itanium, and other 64 bit CPUs, INT scores are generally less competitive)
This is in a 4-way configuration ONLY
Itanium 2 – 1.5Ghz – 82.7
Power 4+ – 1.7Ghz – 68.5
Opteron 844 – 1.8Ghz – 49.2
SPARC 64 V – 1.1Ghz – 41.4
Guess not, Opteron kind of wastes it…but look at Itanium fly! btw… Opteron REALLY creams it at INT2000 RATE, so most applications would work much better under Opteron.
Guess not, Opteron kind of wastes it…but look at Itanium fly! btw… Opteron REALLY creams it at INT2000 RATE, so most applications would work much better under Opteron.
Using a 1.8GHz Opteron in place of a 1.4GHz brings the total system cost at Colfax to ~$22,000, as opposed to $26,000 for the V440. Fudging the numbers for the 1.28GHz UltraSPARC IIIi processors in the V440, you can expect a score of approximately 48, compared to 49.2 for the 4 * Opteron 1.8GHz. Now factor into account the fact that an extra $4000 gets you 3 years of parts and onsite labor, Solaris, a decade mature 64-bit operating system and also includes software support for Solaris, and I think you’ll see the V440 is competative in price/performance with systems in the x86 world.
>>Anybody that loves OpenSource and hates Microsofts needs to support SUN<<
Completely disagree. Sunw is supporting scox, in scox’s efforts to ruin linux as we now know it. Scox is claiming that all linux have to pay scox a fee of $699 per CPU – soon to $1499 per CPU. According to scox, scox has grounds to sue any linux user who does not buy protection from scox. Unless you buy from sunw.
Sunw is making a big fuss about this, constantly saying: “buy from sunw, or scox will sue you.”
Sunw is no longer a friend to OSS.
Roy (IP: 12.147.66.—) – Posted on 2003-10-03 19:19:42
It seems like they are begrudgingly offering x86 and Linux. Are they offering 64-bit support on Opteron (not being sarcastic, I just don’t know)? They also seem to be waffling on Linux. I just don’t get the feeling that they are really behind anything but Solaris on SPARC. You are right that this is more of a marketing issue, but marketing is a really importation part of their “strategy” (remember Digital – great technology, poor marketing).
Is this funny: http://www.eweek.com/article2/0,4149,1155172,00.asp
“eWEEK: Has Solaris plateaued, and should Sun ramp up investment in Linux and get it running on SPARC as soon as possible?
Schwartz: No. AIX and HP/UX have plateaued, given IBM’s and HP’s inability to ship them on x86, so Solaris is consolidating Unix in the RISC arena, and Solaris on x86 is now the only data-center Unix on x86, as well. One million new copies [were deployed] in the past year. It’s really down to Windows and Linux otherwise—and we ship two out of the three.
On middleware, Sun leads on messaging, directory, calendaring, identity and clustering for Unix. Where don’t we lead? App servers, but we’ve got great momentum with our [Sun ONE Application Server] 7.0 product, as customers figure out it’s one-tenth the price of its competitors. And, of course, we lack a database.
We believe we win by getting the complexity out in delivering this as a systems stack. And by driving the price way, way down.
Oh, and here is *ANOTHER* quote:
“eWEEK: What’s your gut reaction when your competitors call Solaris a “proprietary” Unix?
Schwartz: Our Unix runs on 32-bit x86 systems, and supports standards that run on all platforms. Last [time] I checked, you could only run proprietary AIX systems on IBM’s proprietary systems. Ours features IMAP, LDAP, ICAP, J2EE, JSP, on and on. AIX, again, doesn’t.
As another example, we support MySQL—IBM’s still trying to promote their proprietary database, which won’t run on MySQL. Seems like a lock-in strategy to me.”
It seems they’re PRAISING Solaris running on x86! I am sure if I spent more time searching, I could dredge more quotations out of the backwaters of the internet.
SUN’s mission is to sell an end to end solution, OUT OF THE BOX. The only people who have a problem with this are IS geeks who are paid by the hour and are pissed that instead of spending 24hours to setup and test a server, they will only need to spend less than 4hours. The only people have have a problem with SUNs Linux strategy is Walterbyrd. For all intensive purproses, every time there is SUN pushing Linux as a desktop solution we have him pop up slamming SUN. Can anyone say, “Windows advocate under the cloak of Linux-hype”?
I would *LOVE* just *ONE* person here to point out an server application that is available on Linux but not on Solaris x86. It seems that people are more wrapped up in the Linux “hype” than looking at the straight out facts.
Martin Tilsted (IP: —.bnaa.dk) – Posted on 2003-10-03 19:59:50
To Bascule:
Now that is easy to find. Just go to
http://colfax-intl.com/jlrid/SpotLight_more.asp?L=71&S=1&B=…
and you can confdigure a 4-way opteron with 16GB ram for only 16541$. At least as fast as the sun box. And that includes Two hot-pluggable 64bit-133MHz PCI-X slots & three 64bit-66MHz PCI slots. I could not find out what kind of pci the sun box had, but it is NOT pci-x so the opteron also have a faster extern bus.
It is only available in the US thus making it completely useless for deployment. Show me a high profile, *INTERNATIONAL* company offering Opteron solutions *CHEAPER* than SUN.
The only place x86 is “muscleing” in is on the low end (although “muscleing” in on whom is another matter, as Sun doesn’t compete in the low end market) For midrange servers (what I’d consider midrange… Sun would consider the low-end) the V440 has Dell beaten on price/performance, and on the high end there really is no x86 competator.
Linux zealots spewing anti-Sun FUD certainly gets boring and repetative…
Oh, please! If pointing up trivial spelling mistakes and resorting to “Linux Zealot” name-calling is the best you can do, you need not have bothered to reply.
The central point, which you didn’t/couldn’t address is that Sun is now lacking a clear strategic direction under McNeallys leadership.
Sun purchased Cobalt and they’ve virtually closed those products down – the source code for the Qube has been released under a BSD licence. Most of the Raq range has been discontinued.
They purchased Forte Software, rolled the product into Sun ONE Unified Development Server, benefitted from the high revenue streams, couldn’t decide how to make it play with Java and so now that product is be end of lined too.
There’s been an on-off-on affair with Linux – which is confusing for customers, partners and (dis)interested observers – to say nothing of the Wall Street “analysts”
I rather like Sun and I want them to survive, but in order to do so, they’ve got to set a clear direction and stick with it. A big part of that is deciding if they’re a hardware company or a software company.
Don’t bother pointing out the spelling mistakes.
The central point, which you didn’t/couldn’t address is that Sun is now lacking a clear strategic direction under McNeallys leadership.
They have a clear vision now, Solaris x86 and SPARC on the server, Linux on the desktop and SUN One Server Stack software running on either Solaris x86 or SPARC. All this can be combined into an end to end solution for around $125 per user, per year, of it you have over 100,000 users, there is flat rate agreement.
Up until now, they had no clear vision. There was a swing to Solaris, then to Linux then to something else. However, I will continue with this “theme down the post”.
Sun purchased Cobalt and they’ve virtually closed those products down – the source code for the Qube has been released under a BSD licence. Most of the Raq range has been discontinued.
Talk to any SUN Employee and even they’ll tell you it was a stupid idea. If I was in charge, I would still have bought Cobalt BUT continue to allow to run it as a seperate operation off the side.
Cobalt had a completely different sales model to SUN. SUN deals via their partners where as Cobalt was you typical small vendor model where they communicated directly with the customer. If you wanted a piece of hardware, you simply approached them. SUN and Cobalt sales model just didn’t marry up as the two customer bases were too different to begin with.
Cobalt was successful when it was a stand alone operation. Had they kept it seperate, SUN would find that they would have a VERY successful Linux unit make decent profits for the parent company.
They purchased Forte Software, rolled the product into Sun ONE Unified Development Server, benefitted from the high revenue streams, couldn’t decide how to make it play with Java and so now that product is be end of lined too.
They never purchased Forte Software. Forte Software is SUN Workshop rebranded. The only thing they did purchase in terms of development was Netbeans which was later opensources under the SUN Community License.
There’s been an on-off-on affair with Linux – which is confusing for customers, partners and (dis)interested observers – to say nothing of the Wall Street “analysts”
Can’t ague with that. First it was Solaris on x86, then Linux on x86 and now it is Solaris on x86. They’ve got a good OS, Solaris, why not bloody well use it? it is cheap, easy to administer and the only thing that is lacking is hardware support, which IMHO something not too hard to fix.
I rather like Sun and I want them to survive, but in order to do so, they’ve got to set a clear direction and stick with it. A big part of that is deciding if they’re a hardware company or a software company.
They have a clear vision right now, the question is now whether they will stick with it and build upon it. Only time will tell.
Sorry, I wasn’t trying to comment on the price/ performance. Just the all out performance. The cost of hardware for x86 hardware from HP would waste anything that SUN offered, and in turn Dell would totally wipe them out. (I am not a fan of Dell hardware personally.) Also, the Sparc IIIi is actually a slower CPU then the SPARC64 V, so I doubt even at 1.4 Ghz it could match the Opteron. Further still, integer performance on the Opteron is ALOT better than then the SPARC. Applications rely more on integer performance than FP. SPARC does not even remain competitive with the Itanium and PowerPC for high-end databases or scientific applications, the Itanium is scaling WAY better than SPARC (Probably due in no small part to Alpha engineers.)
Sorry, I wasn’t trying to comment on the price/ performance. Just the all out performance. The cost of hardware for x86 hardware from HP would waste anything that SUN offered, and in turn Dell would totally wipe them out. (I am not a fan of Dell hardware personally.) Also, the Sparc IIIi is actually a slower CPU then the SPARC64 V, so I doubt even at 1.4 Ghz it could match the Opteron. Further still, integer performance on the Opteron is ALOT better than then the SPARC. Applications rely more on integer performance than FP. SPARC does not even remain competitive with the Itanium and PowerPC for high-end databases or scientific applications, the Itanium is scaling WAY better than SPARC (Probably due in no small part to Alpha engineers.)
Again, does it need to be repeated. There have only been 16,000 Itanium CPU’s sold in the 2-3 years they have been available. Who cares if it is faster, “cooler” or “more trendy”. If the applications aren’t available, the operating systems are poorly tuned and the system availability is terrible, is there any doubt why only 16,000 has been sold.
Worse still, look at the so-called “companies” who support Itanium. HP being the most vocal. Just look at their pathetic legacy of hyping something then dropping it like a hot potato a year later, and worse still THEY HAVE NO MIDDLEWARE. They have no end to end solution for the customer to integrate into their organisation. HP is at the mercy of third party vendors.
As for “SPARC64 V”, it doesn’t exist! and it won’t exist until next year. The Opteron performs adequately, cheaper and most systems sold today are 2 and 4 way configurations. Who gives a shit whether Itanium can scale to 2trillion NUMA configuration, ultimately, if it only serves an ultra-niche market, does it really matter? no.
Yes, SPARC64 V does exist and is released by Fujistu Siemens. And in regard to the Itanium AND PowerPC trouncing the SPARC, this makes the SPARC IIIi even more pathetic, since this is the market it is *supposed* to be targeting and the only place it was once competitive in 16-256 way systems! (Even Pentium XEONS beats it below 16 way, and only because it doesn’t exist in that configuration!) Fujistu doesn’t even bother to post results from their own chips, they post SPEC scores on Pentium 4 systems now.
Just to give you a taste of the current integer performance of the SPARC chip (the fastest one out there), its FASTEST integer score ever posted on a 4 way system is 30.5. Now the fastest 4 way Opteron score posted for integer performance is 48.5, that is only 63% as fast as the Opteron!! This is versus a 1.1Ghz SPARC and a 1.8 GHz Opteron. Opteron is now 2Ghz, so that should pretty much tell you that SPARC is no where near competitive.
The story was very similar about a year ago with PowerPC, but they released a new chip and ramped up the clock speed. Now they are competitive again. SPARC is a great design, Itanium stole some design concepts from the SPARC. However, they have not kept on the edge of technology, and as such without a miracle SPARC is a distant memory.
Why would I want to go dump a lot of money into a really slow proprietary SPARC beast when I could buy a much faster x86 machine, with a full 3 year warranty, that will run pretty much anything instead?
Itanium is scaling way better on 2 – 8 way systems as well. Better than even Alpha did, amazing how throwing money at something (L2,L3 cache and increase bus bandwidth) can boost performance. Opteron also scales VERY well, 2 – 8 way. Much better than XEONs and SPARCs and pretty much on par with PowerPC, only Itanium stands out above Opteron.
Itanium is a proprietary solution. Who makes Itanium? Intel. Who controls Itanium? Intel. Who has the final say? Intel. Compare that to SPARC and x86-64 which are openstandard.
They have a clear vision now, Solaris x86 and SPARC on the server, Linux on the desktop and SUN One Server Stack software running on either Solaris x86 or SPARC. All this can be combined into an end to end solution for around $125 per user, per year, of it you have over 100,000 users, there is flat rate agreement.
I must nitpick here, because it makes a huge difference. It is NOT $125 per year PER USER, but $125 per year PER EMPLOYEE. This makes a huge difference to ISPs, who may have 10,000 users but have 20-30 employees. This is a license to run Solaris at, as Schwartz put it, infinite scale.
They never purchased Forte Software. Forte Software is SUN Workshop rebranded. The only thing they did purchase in terms of development was Netbeans which was later opensources under the SUN Community License.
I agree with all of your post except for this bit. Forte Software was a multi-million dollar international business built around a product for producing enterprise systems. Sun purchased it in 1997 or 98. You’re right Sun about the rebranding NetBeans. The idea was to take the Forte Application Environment, migrate it to java and continue selling the enterprise product to businesses with an integrated low-end/free development enviroment (netbeans) for those who didn’t need the enterprise features.
The interesting thing about the Forte Application Enviromment was that it achieved all that J2EE attempts to achieve, but in a much more elegant and developer friendly way. It was similar to java in that to use it a “Forte Run Time Environment” (FRE) was needed for each machine it was to be used on.
Programs were compiled from a language called TOOL (Transactional Object Oriented Language) to C++ object code which could be executed by the FRE. TOOL is very similar to python in many respects.
Each machine having a FRE knew about every other machine in the “environment” and a distributed application could be built simply by “Partitioning” (deploying) the relevant code (TOOL objects) to the client or server boxes in the “environment”. Nodes in an environment could be specified as being clients, servers or for failover or loadbalancing. All networking and object sharing, loading and instantiation was taken care of behind the scenes leaving developers free to develop. Code lived in a central repositary, against which developers did their checkouts or branches. Finally, Forte shipped with a complete object Framework (similar to NeXTSTEP), and provided wrappers for C code, ActiveX etc.
Sun have just recently stopped selling this product as Sun ONE Unified Server. A complete waste of opportunity since they owned the code and the intelluctual property prior to formulating the J2EE specs.
The once vibrant developer community has now mostly evaporated – many moved on to MTS/COM and to J2EE. Those that are left want the code to be openSourced. Sun don’t seem too keen.
Interesting, Merrill Lynch analyst didn’t get fired for this… Obviously SUN is not Microsoft.
IT analysts are like politicians. They are not here to inform,
but to misinform and misguide. There are always three components that have to be taken into account:
– what is said
– who is saying that
– the moment when it has been said
Then, maybe, one can guess what the truth is.
DG
Sun Microsystems will be in future! I love this company and will love in future.
Only there is need Itanium systems and support in Solaris for Itanium processor!