I wonder how Moon Macrosystems would compete against Makrotuf software?
It is hard to believe that three whole years have passed since software giant Oracle instantly became a systems player by snapping up beleaguered Sun Microsystems for a cool $7.4bn – about $5.6bn net of Sun's cash hoard. No one took El Reg up on our suggestion to start up a Sun clone we called Moon Macrosystems by grabbing all …
"The question is, will customers want to be locked so tightly into Larry's World?"
Answer: Of course not.
When in time and in what field did customers want to be all-in? Maybe if in the sense of being snow blinded into it, or otherwise left with no choice, then OK. I just can't think of a moment in business, where I could see this being a voluntarily decision.
And was it me, or was this article a little too long to say "Oracle might do something"?
I can see scenarios (purely fictional for now, but bear with me) where one would want to go all-in.
It definitely makes sense from a simplicity, standardisation and cost point of view. Which is why, for example, airlines of smaller size tend not to mix Airbus with Boeing craft. Similarly most air forces tend to standardise around a narrow range of platforms and engines.
On the other hand, the only way to avoid having your bank account plundered, or to be hostage to the supplier's whims is to actually be a shareholder of said supplier much in the way automotive Tier 1 suppliers feature most of their clients in their shareholder register.
Come to think of it, I'm surprised this hasn't happened already, actually.
Oracle was able to essentially kill MySQL, which was the biggest threat to its cash cow. DB2 and Oracle now rule the market.
Oracle owns JAVA and SAP is screwed. SAP retooled their who application stack on JAVA and is about freaked about asking Oracle for the re-license cost.
Oracle is milking that cow better than anyone else. They have no customer compassion so eliminating 9-5 service and forcing every dusty box to be under service.
Oracle knows how to take account control to the max and suck more wallet share from customers, personally I prefer the dual vendor strategy. We played IBM vs. Amdahl and now its Oracle and IBM.
Microsoft is the odd man out and now that M$ft is investing in Dell expect HP to sink even further into the abyss. Microsoft is going to buy Nokia and Dell to become competitive again vs. Apple. They will fail but that is their strategy.
Oracle owns JAVA ...
"Java" is not an acronym.
Microsoft is going to buy Nokia and Dell to become competitive again vs. Apple.
What's mildly interesting about that prediction is that it would answer Morgan's question about Oracle's position as the only IT vendor since IBM to control "so much of the stack" (ie, from hardware through systems software including database and other major components). Whether the market would be interested is a separate question; but I admit I'd be a tiny bit curious to see Oracle and Microsoft bare-knuckling it out in the integrated-systems town square while IBM leans against a post with hat brim pulled down and watches thoughtfully. If you know what I mean.
Amazing that HP didn't buy Sun instead. They have no problem blowing far more cash on much more expensive and risky acquisitions, which all get the goodwill written down in a few years anyway, but they ignore the one buy that was staring them in the face as as a perfect opportunity. Truly the most mismanaged of the Fortune 100 companies for the past decade.
"Truly the most mismanaged of the Fortune 100 companies for the past decade."
Thats what happens when a company becomes so management top heavy that the technical people who should be giving advice to the board about the worth of potential purchases of other tech companies are so far down the food chain that they never even get to look through the window, much less give their opinion. Its a shame because back in the day HP produced some good kit - the HP/9000s were a stonking line of machines and HP-UX was a worthy unix. But now , oh well, all good things come to an end....
Agreed. I didn't work with their servers, but their printers and science lab equipment were killer. I worked for a company that pretended to compete with them in some of the science lab market, and privately our execs would admit we couldn't go head to head with them. On the couple of occasions when we DID come out with something better, HP were quick to ink a deal with us to either put their Brand on our equipment, or work with us to develop the product and sell a varient to which they had the exclusive license.
I don't see the HP-Sun synergy you do. From my perspective both were essentially hardware vendors who haven't got a good handle on service, which is where the cash is these days. I do see the Oracle-Sun synergy since Oracle was a service and software company. I concur with the poster above that this is bad for customers because Ellison basically rapes his customers (and I think if I had been on the government committee reviewing the deal I would have killed it based on the db segment overlap), but I do see the business synergy from the Oracle and Sun standpoint.
But then I'm in the commodity end of the business, not the big iron.
I've been in a number of Fortune 100s in the past year. Sun hardware is going down hard, check IDC Unix tracker chart also. One interesting story.. got in a discussion with a fellow at a Fortune 50, re: hardware refreshes and application rollouts. The preferred platform is Linux on mainframe. If you want, you can do pseries and AIX. HP/UX *only* gets a pass if the application will only run on HP/UX and "oh by the way, don't expect another go-round." Sun and Solaris. If your application will only run on Solaris and you need a hardware refresh, you need to find a new application. I suspect there is a LOT of that going around. Certainly don't want to risk the company on an OS that is trending down dramatically, do we?
I've heard similar stories from a friend who works in a government lab that does require big iron. Their first choice use to be Sun for new workstations. When Oracle took over they doubled the price of the existing service contracts. So all the new *nix equipment they buy comes from an outfit I've never seen mentioned in a Reg article, and whose name I can't recall at the moment. Granted the sales numbers are maybe $20K a year on a new server at the location plus existing service contracts, but I imagine that's happening at more than one location.
I know the biggest Solaris shop I ever worked in (a certain very large semiconductor company that really really loves the state of Texas) was pushing to move all their in house software off SPARC to x86. Solaris might hang around awhile but the SPARC architecture is a dead man walking.
The T4 and newer line of SPARC-based hardware have improved significantly since the "blunder" years of the Niagara fiasco.
Like most others in related to this field, I was skeptical about Oracle's dedication to keep the Sun portfolio alive and nurse it to health. I must admit I have grudgingly conceded that they seem sincere in their efforts to revive the Sun hardware line.
Having worked with sun gear for a good part of two decades, I must say that some of the moves being made are interesting and have potential for success. Moreover, Oracle is so very (brutally) fiscal minded that I don't doubt they will be able to turn things around. Only unfortunate thing is that the timing of it all might be too late.
I have found the T4 line very impressive and when we start considering the cost of ownership of the stack it starts making a lot of sense financially (with virtualization and the management software etc thrown in for "Free").
The exadata platform is very impressive (albeit super expensive) -- we were able to run a monster dwh workload on the exadata platform and see 40x performance improvements (no joke).
So love them or loathe them, Oracle seems to have done something that the bumbling management at Sun wasn't able to do. Draw a line and stand by it.
That may very well be true as I haven't worked on a T4 system yet but one thing Oracle doesn't do well is drop prices. As previous mentioned with x86 now having most RAS features and being largely commoditized it does very much it seem like its too late for SPARC.
The article says:
"... is that Sun took a huge wonking write-off while the Oracle deal was pending, shedding nearly 6,000 workers and writing off the value of some acquisitions. This was done to gussy up Sun's books before the company was to be absorbed by Oracle, should the governments of the world approve the acquisition."
So what would have happened to Sun if the governments had refused the takeover? Would it have died (whether slowly or in a blaze of disgrace)? Or would it have emerged leaner and meaner, rebranded itself (as Moon Macrosystems) and thrown itself back into the fray?
I'm not a business analyst - just a sysadmin that built most of his career on Sun/Solaris - so I'm intrigued to know if a "new Sun" could have emerged from the ashes of the old.
PS: Where's the "inquiring minds want to know" icon?
Had the deal gone bad, Sun might've been able to recover if they had actually sell the stuff on their software stack. Their LDAP and IDM solutions are very good; in fact, most of the people from the Identity suite went on and forked the whole thing. Check out ForgeRock!
All good and interesting. One aspect to add is how Oracle would have performed without Sun, and/or if someone else had purchased Sun. Plus the ol' tax write-off of previous losses...
So a successful acquisition in terms of Oracle shareholder value. The world is changing yet again, will Oracle adapt again? If yes, that will be extremely impressive...
I'm curious, as you were providing so many charts in this article:
Can you also provide a simple chart mapping software cpu needs vs. cpu performances?
This chart will prove, that since a couple years, CPU enhancements outperform software needs in a tremendous way. Which makes it easy to re-write Moore's Law as:
"Every two years HW-revenues are reduced by a factor of two."
So, assuming, that growth in HW only is possible is neglecting that simple fact of Moore's Law. Cisco can only grow at the expenses of IBM, HP, DELL, and, yes, Sun/Oracle. In the end, all HW-only companies will be shrinking. There's a good reason, Oracle does not provide the numbers by line of business, others will follow. It will be a systems vs. systems fight, and we will still see companies going out of business or being acquired by other bigger ones.
maybe it's because I'm not a finance wizard but your charts seem to suggest that Sun's portfolio is still in decline even if the rate of decline has slowed.
Clearly, Oracle is benefiting from the ability to sell integrated systems such as Exadata but I'm not sure if they're really making headway elsewhere. They will certainly have pissed off shops with the way the handled OpenSolaris which has led to an apparently viable business for Illumos. Personally, I think they have handled MySQL right from a business perspective: keeping free users at arm's length; upping charges but devoting resources to paying customers. Quite how that will play out in terms of the forks (MariaDB and Percona) interests me less but it seems that the Postgres ecosystem continues to benefit from fears about lock-in. The EnterpriseDB presentation of 9.2 last week was very impressive in terms of performance, Oracle compatibility (the object type) and migration tools that are obviously coming from working with customers who are moving from Oracle to Postgres.
Of course, driving towards integrated systems might mean that Oracle is less interested in selling the database as a standalone product: IBM is after all more than happy to have its applications run on Oracle as well as DB2, but I do think that Oracle is possibly more dependent upon the reputation of its database.
Here's the thing buy a locked in user base - double the support costs - force up employee attrition levels through aggressive management - reduce overheads at the same time release redundancy provisions straight onto the bottom line. Rebadge legacy SPARC and Intel as 'Engineered Systems' and you have the basis of a highly profitable asset strip - magic trick!
"...will customers want to be locked so tightly into Larry's World?" - Probably not - they didn't like it with IBM in the 1980s and they don't like it now.
Vendor independence is the aim, not least because it does not matter anymore where that x86 server comes from. As long as it can run VMWare or RHEV, and companies don't want to be held hostage by someone's licensing and maintenance plan anymore. Also, those Sandy Bridge chips are really performing - SPARC or POWER7 hardware may be 10-20 times more expensive, but by all means they are not 10-20 times more performing or resilient than x86. That said, Oracle 11g is not the last word in databases either. Neither is MySQL. There is Percona and PostgreSQL. While we are on it, Intel's 30 year old x86 architecture is not wisdom's last conclusion either: I can hardly wait until there are eventually enterprise scale ARM 64bit servers.
Intel/VMware/Redhat is not vendor independence its standardizing on substandard offerings and latching them together hoping it creates a "good enough" technology to run your business.
A lot of companies can lose more in one hour of downtime than the server costs. Not sure where 10-20x comes from last I checked RISC was only more expensive because it takes two x86 boxes to have similar reliability.
x86 is the garbage architecture of the world. We should be running RISC based hardware by now, but Intel cranked up the clock rates to get their processors to match RISC. Notice that ARM was born sometime around the late-80s and yet manages to have a fairly good performance without running hot or drawing too much power.
Hopefully, we're closer to an ARM takeover, maybe we'll finally get R&D for non-Intel architectures...
I thought that as well some time ago, but it does not matter if x86 is "garbage" or not. What matters is the commodification which results in RISC architectures losing their unique selling poiint.
I had to learn it the hard way myself, too, but the times of RISC and proprietary hardware archiectures is over. I've seen CIOs making strategic decisions to move away from RISC to x86 - not because they love x86 but because they had no continued business justification for SPARC and POWER. To give you a few examples, NYSE Euronext and the London Stock Exchange moved to Linux on x86, petaflop systems are being built using x86 (http://www.theregister.co.uk/2013/01/30/atipa_pnnl_hpcs_4a_supercomputer/)
Run some STREAM benchmarks side-by-side on x86 cores and ARM cores and see what the results are... Driving large chunks of external memory tends to burn a lot of power, this is old news. That said I think that x86s are overkill for a lot of apps, and ARMs will change the game over the long run. It's another iteration of the "Attack of the Killer Micros", except x86 is the big iron now. ;)
At the time of Oracle's purchase of Sun, the SPARC was one of a limited number of choices if you wanted RAS features - you could get an IBM or Unisys mainframe, or use an Itanium. So I figured right away that having the ability to run Oracle's databases on something appropriate (for some accounts) other than IBM hardware (they have their own databases) was the big reason behind the purchase - not Java as the wave of the future.
Then, shortly after the purchase, Intel announced Nehalem-EX, now the Xeon E7, and so RAS is available in the commodity x86 world. It seems like Oracle wasted its money.
Didn't Ellison insist he was going to make $2bn profit off the Sun bizz in the first year, but instead he's staggered to just over $1bn (by the most generous of estimates) after three years, and he only got there by gutting all the staff he actually needs to develop new systems? And you want to paint that as a success?
No. According Mr. Morgan's March 2010 article located here: http://www.theregister.co.uk/2010/03/26/oracle_q3_f2010_numbers/
"In the call, Catz said that even with Sun included in the mix, Oracle's operating margins were at 45 per cent, and she reiterated that Oracle was on track to get $1.5bn in non-GAAP operating income from the Sun division in fiscal 2011 and $2bn in fiscal 2012."
So Oracle's hardware business is now supposed to hit $2bn in fiscal 2012. They may get to it. I vaguely recall hardware business’ gross margins improved to 55% in at least one of their quarters. Regardless of units sold, the hardware business is now a profitable one for Oracle.
Biting the hand that feeds IT © 1998–2019