If HP and IBM can’t make money selling the same or lesser boxes for more, what reason is there to believe that Sun can? Scott McNealy used to describe x86 retailing as adding bruises to bananas, but now he’s doing it. Read the article at LinuxInsider.
An Analysis of Sun’s Linux and Opteron Strategy
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
McNealy is desperate, thats it. Hes doing what customers are telling him and he wants to keep his job. He actually dosent really like it–core sun technology–but he wants to keep his job.
Given Dell’s recent announcement that it would not be supporting Opteron in the near future (http://www.eweek.com/article2/0,1759,1542202,00.asp) this leaves three tier 1 players supporting Opteron: Sun, IBM, and HP.
HP’s Opteron strategy is fairly clouded at this point. They’re expected to release an Opteron server in the near future with anywhere from 1 – 4 processors. We’ll see how this plays out.
Sun and IBM are currently offering much the same thing: 2 way Opteron blade servers. The clear disadvantage of this is that they aren’t fully leveraging the Opteron architecture.
IBM doesn’t seem to have big Opteron plans, and as far as I’ve read they don’t have any plans at this point beyond the eServer 325.
Sun, on the other hand, has recently acquired an Opteron server manufacturer (http://news.com.com/2100-1010-5156828.html?tag=nl), and sent one of their top engineers (who is also a Sun co-founder) to the company to design the next generation of Sun Opteron servers. Here was Scott McNealy’s quote on the matter:
“He’s the most prolific and exciting and talented workstation and single-board computer designer on the planet,” McNealy said. “With this guy…designing Opteron servers, there ain’t going to be nobody who has the class and breadth of computers we have.”
Of the three tier 1 players backing Opteron, Sun certainly seems to be the most excited about it, and will most likely have the most diverse Opteron product line.
Also, Sun is the only player of the three offering a commercially supported Unix for the Opteron. Given HP’s deep investment into Itanium, I believe it’s unlikely they’re willing to support HP-UX on anything other than IA64 and PA-RISC.
So, Sun looks to remain the number one tier 1 Opteron player, at least in the near future.
My (completely ignorant) impression was that Linux’s inroads into the Unix market was mostly due to the cost of Intel/x86 hardware vs big iron unix vendors.
Sun was obliged to offer an x86 solution because many of its existing customers were unable to ignore the price differenial between a Sparc box and an x86 box and by offering a Sun branded x86 solution Sun can keep those customers Sun customers rather than lose them to IBM.
look at the amount of overhead involved. Of all modern hardware companies, the only one’s that have done even nominally well are Intel, Nvidia, and Cisco (somewhat). Sun dropped i386 architecture for a while, and look what happened to them (perhaps it was just a coincidence). They are doing what any company in their shoes would do, responding to market pressure. SPARC hardware is going the way of the dodo, and rapidly. Another case in point is apple…they have succeeded in capturing a niche market, but haven’t done anything revolutionary since ’84 (OSX is nice, but hardly revolutionary). Sun has more issues than just hardware though. I don’t have numbers in front of me, but I seriously doubt their solaris division is doing well (Sun is probably Sun’s number one consume of the OS), and Java needs a serious overhaul to remain competitive. IMHO opinion, all the execs at Sun are polishing brass on the titanic. I would think that perhaps they would be acquired, if there was something there to acquire.
Id have to agree with this article. I also wouldnt call the Sparc dead, just it will never have the popularity of x86. some people equate popularity with surivival, apple has cleary showed you dont need to run 90% of the boxes in the world in order to be raking in the dough. Thats my Opinion.
I’d have to say that it was one of the best articles on company strategy vs. competition I’ve ever read.
I’d agree with Jayson about hardware sales being tough, however his assertions about SPARC and Solaris fall a bit short. I don’t think SPARC is going to die and Solaris would be a hit with the AMD64 crowd. Java is a separate issue but currently being abused by Sun Marketing.
He doesn’t understand the strategy, simply because….IT’S NOT ABOUT MAKING MONEY ON LITTLE SERVERS.
It’s about winning mind share, market share, and offering the slickest path to the meaty high end. Oh, YEAH, and that is only A TINY part of the whole STRATEGY of SERVICES SERVICES SERVICES.
Sun wants to give you the complete solution: low end, high end, middle, and desktop (Java Desktop System).
Sun’s strategy is that they can give you it all.
This guy can’t see the CONTINENT (let alone the forest) for the trees.
Sun should we wary of open source. Why? I used to like open source too. But then I realized — would you open source your wife, and let anybody have access to all her parts and use her? NO! Just like that, you shouldn’t open source valuable IP. Keep it to yourself, like society intended.
the GPL is known as “communist software”.. the BSD/Apache/MIT is known as “do whatever you want with us software”.
Sun wants to open source java because the open source community is huge and will do the work for them, it will also promote it further and adoption will be quicker.
As for SPARC/Solaris. They should have that as their main product like But I do agree with what IBM is doing except they are using big intel servers instead of pushing power/powerPC to the big ones all across. I dont ever see a power server commercial from IBM.
As for HP supporting PA-RISC, well explain why HP has announced that they will be discontinuing it.. HP is supporting an x86 and IA-64 line only. HP sucks that they are getting rid of the alpha!!! >:(
As for x86 being so much better and cheaper than SPARC, Maybe you should do a comparison yourself. Get a $900 SPARC machine from the online sun store and a $900 x86 machine from price grabber and tell me, which one is more solid and faster.
jizzles (IP: —.CS.UCLA.EDU)
He doesn’t understand the strategy, simply because….IT’S NOT ABOUT MAKING MONEY ON LITTLE SERVERS.
You might try reading the article a bit more closely. The point of the first section was that they will be losing money on their Opteron offerings, but the last two parts are an analysis of why backing Opteron is strategically advantageous to Sun despite the fact that they will be losing money. Namely, Sun offering Solaris/AMD64 is value added for Solaris/SPARC.
Anonymous (IP: —.cpe.net.cable.rogers.com)
Sun should stick to SPARC/Solaris
Sun should we wary of open source. […]you shouldn’t open source valuable IP. Keep it to yourself, like society intended.
Well, except for GNOME, Apache, samba, and a handful of other open source utilities which have been a part of Solaris for years, there isn’t really much open source involved here. The V20z is only the beginning of what Sun is planning to offer in its Opteron line, but when the next generation of Sun Opteron servers is released (designed and engineered by Sun Co-Founder Andy Bechtolsheim), which will coincide more or less with the release of Solaris 10 later this year, Sun will soon be ready to release a complete 64-bit build of Solaris 10 for AMD64, which will implement Solaris/SPARC-like security features such as a non-executable user stack, as well as include the 64-bit optimizations in the kernel (such as a 64-bit VMM).
So, for now, Sun will support Linux or Solaris x86 on the V20z, but keep in mind their real strategy is Solaris/Opteron, which, as the article describes, is ultimately value added for Solaris/SPARC.
Scott had nothing to say when Sun credit was downgraded to junk status. Nada. He has been MIA for some time as his company circles the toilet. He has sent Schwartz out to be his point man as he huddles under his desk. Maybe he should just pack it up and let someone else try. Things couldn’t get much worse for Sun.
Scott had nothing to say when Sun credit was downgraded to junk status. Nada. He has been MIA for some time as his company circles the toilet. He has sent Schwartz out to be his point man as he huddles under his desk. Maybe he should just pack it up and let someone else try. Things couldn’t get much worse for Sun.
I find that rather humorous considering the number of IBM linux customers who are now lining up purchasing Opteron servers loaded with Solaris 10 as part of their server fleet once it becomes available; second half of this year.
Linux had its time. Its time was based on being the cheap UNIX for the the “rest of us”, SUN has now started to provide the same thing WITH commercial support; Solaris x86. Why on earth would you want to use Linux when the real thing is available at a cheaper price?
too bad Sun didn’t go with Newisys. Newisys understood the potential of scaling the Opteron to many-cpu systems. Sun must have gotten scared since scaling the opteron to 8, 16, 32 cpu’s would destroy their high end market.
too bad they still push Solaris over Linux.
too bad they have no long term plan other than to maintain sales of their high end sparc and solaris systems. Sun treats opteron and linux as a loss-leader and a gateway to solaris/sparc sales.
too bad these decisions are going to bite them in the ass. bye bye sun.
“Things couldn’t get much worse for Sun.”
Why in the heck are they posting profits then? Revenue is higher than its ever been. GROW UP
They do have long term plans and these decisions arent going to bite them in the ass, however, they were very late in offering x86… and that caused them alot of pain. But right now they are set to grow and they are doing better than all of you are saying so If you dont know much about sun then just be quiet
Sun, on the other hand, has recently acquired an Opteron server manufacturer, and sent one of their top engineers (who is also a Sun co-founder) to the company to design the next generation of Sun Opteron servers.
This statement is factually incorrect. Andy Bechtolsheim was the CEO of Kealia, the company Sun acquired. In effect, the have re-acquired Bechtolsheim. They have not dispatched him as some agent to handle a special project. This was his own company that Sun has subsequently purchased from him. (And whomever else happen to be stakeholders…)
I don’t have numbers in front of me, but I seriously doubt their solaris division is doing well
I assume you meant to say you had no factual basis for making such an outlandish statement?
Microsoft is the only company that looks at an operating system as an actual product that you sell to generate revenue. Every other systems company uses their operating systems to leverage other software built on their platforms. Every other major UNIX vendor has put their OS on life support, but Sun continues to invest and innovate in ways that save customers money and allow companies leverage their software on many different platforms. Solaris 10 has features that no other OS does. Sun obviously isn’t trying to make money on Solaris, as they’re using it to make money on everything else.
Scott had nothing to say when Sun credit was downgraded to junk status. Nada. He has been MIA for some time as his company circles the toilet. He has sent Schwartz out to be his point man as he huddles under his desk. Maybe he should just pack it up and let someone else try. Things couldn’t get much worse for Sun.
Exactly how does this relate to Sun’s Opteron strategy? This comment is inapposite to the article and the discussion, and is frankly incendiary.
too bad Sun didn’t go with Newisys. Newisys understood the potential of scaling the Opteron to many-cpu systems. Sun must have gotten scared since scaling the opteron to 8, 16, 32 cpu’s would destroy their high end market.
Too bad Kealia was already looking into doing things like this, and Sun has already aquired the expertise. Seriously, though, how do you know that Sun isn’t looking into replacing their high-end line with big-iron Opterons? Kealia was looking at building some major high-end systems before they were acquried, and what’s to stop Sun from proceeding with this expertise? In fact, it would make a lot of sense for Sun to eventually replace their high-end systems with ones that are grounded in “commodity” technologies.
In fact, I would worry more about HP and their middle to high-end markets. They’ve killed off Alpha, and PA-RISC and are in the process of trying to sell both Itanium and Opteron to their customers. Talk about a confused strategy? In fact, they were the biggest proponents of Itanium and have been trying to force their customers off of their older platforms onto newer ones so they can cash in. However, Sun is providing upgrade paths for owners of SPARC hardware, who don’t want to part with it, along with new platform options for those interested in x86 and x86-64. I would contend that HP’s strategy is worse for their customers.
“Why on earth would you want to use Linux when the real thing is available at a cheaper price?”
Because Solaris is not cheaper. Not by a long shot.
Because Solaris is not cheaper. Not by a long shot.
Your statement is incorrect.
In a for-profit business setting Solaris is available for $100 per employee per year, whereas RedHat Enterprise Linux AS is available for $2,400 – $18,000 per machine per year. So if you’re a small business with 5 employees and 20 machines, you pay $500, not $48,000 and this price includes all of Sun’s software stack from end to end. You’ll have to pay even more to get solutions from other vendors.
If you’re talking about purely personal use, or for use as a developer, then you likely qualify for a free license from Sun. I don’t see how it is possible for you to claim that Solaris is more expensive than Linux.
If you don’t believe me see:
http://wwws.sun.com/software/solaris/binaries/index.html“>Free
<a href=”http://wwws.sun.com/software/javaenterprisesystem/index.html“>JE…
http://www.redhat.com/software/rhel/purchase/“>RedHat
It’s all spelled out there. All you have to do is know how to read and understand english, and maybe do a little arithmetic.