Feeds

back to article Oracle revenues miss expectations – AGAIN

Oracle posted lower than expected revenue for the first quarter of its 2014 financial year, making it the third consecutive quarter that the database giant has missed analyst expectations. Oracle announced revenues of $8.38bn and earnings per share of 59 cents on Wednesday, compared to the $8.48bn and EPS of 56 cents that the …

COMMENTS

This topic is closed for new posts.
Anonymous Coward

We have a bottle of Champagne in the fridge waiting for the day some time soon where we will be out of Oracle's clutches, having gotten in to them by the Sun acquisition and our legacy hardware. Since then it has been down hill in every way that mattered to us as a customer, and we have been looking for Linux+anyone else solutions.

Sorry, for got: hostages!

7
1
LDS
Silver badge

Have you ever asked yourself why Sun ended to be bought by Oracle? It didn't look its "good products" attracted so many customers. It OS was wiped out by Linux running on x86 (which wiped out SPARCs) while Java was killed by a plethora of other languages much better at web development. MySQL didn't bring any revenue as long as most customers kept on using it for free (despite the double license), or migrated to Ppstgres.

It looks that many customer left the Sun boat far earlier than the Oracle acquisition. So blame yourself is Sun couldn't walk in its own legs.

0
1
Silver badge

So having murdered their business partner

So having murdered their business partner to get the whole profit on the deal, HW as well as the SW. They find that they can't keep the HW anyway.

0
0
Bronze badge
Joke

I the Oracle was all-knowing, all seeing...

I guess not, it seems:

http://en.wikipedia.org/wiki/File:Tos_city-on-the-edge.png

0
0

Screwing sun..

Oracle are systematically destroying everything they got from sun, including any good will sun ever built up over the years...

Many former sun customers, as well as former solaris users are now going out of their way to migrate to something else.

5
0
Silver badge

Re: Screwing sun..

>Oracle are systematically destroying everything they got from sun

Well if you buy it you can break it. Larry doesn't give a shit. He's too busy getting his ass handed to him by the kiwis right now.

4
0
Silver badge

Re: Screwing sun..

"Oracle are systematically destroying everything they got from sun, including any good will sun ever built up over the years..."

Yep, Oracle took all the Sun software/hardware in, announced, "Now it's MY way or the highway!", and everybody noticed that the highway went where they wanted to go.

4
0
Silver badge

Re: Screwing sun..

@Joe Montana ->"Many former sun customers, as well as former solaris users are now going out of their way to migrate to something else."

If I recall, Sun customers were fleeing in droves well before Oracle took over.

1
1
Anonymous Coward

Re: Screwing sun..

No, they ware gently ambling in the general direction of x86 (which Sun had finally got round to committing to) and Linux in the final days of Sun.

Then oracle took over, dropped the x86 line (that most customers wanted) as "not profitable enough" and decided to screw around with support costs and licensing terms to gouge them for more money. THEN the customers were fleeing in their droves!

3
0
DJV
FAIL

How sad, my heart bleeds...

All I want is for VirtualBox to get out of their clutches before they have their magical "screw it up" effect on that as well.

3
0
Silver badge

RIP SPARC

SPARC was pretty cool around 2000 or so but here in 2013 as an app developer SPARC's now not so slow death will not be mourned. The majority of SPARC production boxes out there today are much slower with less memory than the current sub $500 Dell desktop sitting on your desk at work.

3
1
Bronze badge

Re: RIP SPARC

I'd put the date earlier than that. SPARC started losing it's cool the moment the workstations started resembling PCs in all respects except the price - i.e. drop the SCSI, drop the SBUS and so on, but still charge five, ten or twenty grand. The last properly nice workstation from Sun was probably the 170MHz SS5 circa 1996 or so. I still have a couple here and they're still working albeit far too slow for production use. On the other hand, whenever my more modern, infinitely more powerful NetBSD PC gets bogged down under heavy I/O I find myself glancing at those machines thinking "It never happened on those".

1
0
Anonymous Coward

Re: RIP SPARC

much slower with less memory than the current sub $500 Dell desktop

Yeah, and that diesel Ford Sierra in the garage is much slower than your new Fiat Panda, so Ford cars are clearly not worth having.

Try looking at a modern multi-core multi-thread SPARC box, there's not much on the planet than can touch it.

0
0
Silver badge

Re: RIP SPARC

>Try looking at a modern multi-core multi-thread SPARC box, there's not much on the planet than can touch it.

Including customers which is why you don't see them out in the wild hardly at all. Mostly what's out there is at a minimum 6 or 7 years old.

0
0
Bronze badge
Thumb Up

Re: RIP SPARC

"On the other hand, whenever my more modern, infinitely more powerful NetBSD PC gets bogged down under heavy I/O I find myself glancing at those machines thinking "It never happened on those"."

+1

Ah yes, I do remember back in the days some people proclaiming SPARC is crap, x86 is faster (usually based on them looking at clock speed and doing some simple single thread operations). Shove tons of users and/or database on it and the old SPARC (we're talking back in the sun4m/sun4u days here) would keep going not missing a beat whereas the x86 PC was slobbering heap and just couldn't do the I/O at all.

0
0
Anonymous Coward

Selective pickings

Right, so you selectively pick out the (few) worst highlights of Oracle's results, headline those and make it seem like their business is in the trash, then drop in as a sidenote at the end that overall revenues and income were actually up, and really it's business as usual.

Jesus Christ, Reg, c'mon.

2
4
Silver badge

Re: Selective pickings

They implied more that Oracles hardware division is a dead man walking than the company in general. Generally its a pyrrhic victory touting higher hardware unit sales when your hardware revenue is down double digits (ie you are basically giving away boxes to existing customers to maintain support revenue). Also touting your hardware support revenue increasing without many new customers is also a bit of a warning sign for anyone contemplating locking themselves into your ecosystem.

5
0

Hardware will soon only be support services

If you look at the financial statement you will see that hardware system products fell from 779 to 669 in 2013 but system support rose from 574 to 592.

Obviously people are trying not to buy Oracle hardware and the sales reps are now relying on audits and automatic support cost increases to increase revenue. When looking at hardware decline you really should only look at hardware which was down 14% NOT INCLUDE SUPPORT RANSOM which make it look like it is only down 7%

BTW we dislike Larry and will someday get rid of all their software even peoplesoft just waiting for the other offerings to get to the required function which is progressing quickly

cheers you fanboys

4
1
Bronze badge

Re: Hardware will soon only be support services

And I thought I was the only one who actually read Oracle's financials, though I stopped last quarter after a quick Google Finance look-see confirmed my long negative outlook. I still have pre-2000 10-Q and 10-K print-outs the back of which I use for scrap paper when programming. At least those trees will not have died in vain.

0
0
Bronze badge

Re: Hardware will soon only be support services

"If you look at the financial statement you will see that hardware system products fell from 779 to 669 in 2013 but system support rose from 574 to 592."

No doubt a chunk of that saying "No thanks, Larry" (and it is difficult to disagree with that).

Part of it could, however, also be people buying T4s and T5s and consolidating several EOL systems onto those (which is what Oracle's marketing seems to be pushing).

0
0
Anonymous Coward

SPARC hardware

Maybe Oracle can explain to their customers why a $1200 laptop made in Taiwan and running Linux is more than four times faster than some "Enterprise-Class" SPARC M-something T4 Server with 32 cores and 128GB RAM, ZFS and all the bells and whistles, and which I'm sure costs somewhere in the tens of thousands.

I'm not even talking about floating-point performance. The comparison above is based on compiling a certain piece of software. 40 minutes on the laptop. 2 1/2 hours on the SPARC Big Deal.

0
6
Anonymous Coward

Re: SPARC hardware

> 40 minutes on the laptop. 2 1/2 hours on the SPARC

Buy a better compiler.

0
0
Silver badge

Re: SPARC hardware

Citation?

0
0
Anonymous Coward

Re: SPARC hardware

> Citation?

Compile BOOST for example, with gmake -j4.

1
0

re: SPARC hardware

the T4 servers aren't M-something. the M-Somethings are the Fujitsu Sparc64 based servers.

people don't buy M series servers to run compilers, believe it or not. they buy them to run large enterprise relational databases with 100s of concurrent transactions

3
0
Silver badge

Re: re: SPARC hardware

>people don't buy M series servers.

Is all you had to say. SPARC is as dead as Itanium. Virtually no CS grad after 2016 or so will ever work on either.

2
2
Anonymous Coward

Re: re: SPARC hardware

>people don't buy M series servers to run compilers, believe it or not. >they buy them to run large enterprise relational databases with 100s of >concurrent transactions

Because running compilers as opposed to running relational databases is different exactly how?

Both compilers and relational databases have three very important common computational requirements:

- both use mainly integer-based calculations, and very little floating-point.

- both are I/O intensive and have large memory footprint requirements

- both make extensive use of IPC

So, please, kindly illuminate all of us here: exaclty why is a SPARC Big Cheese so appropriate for a relational database server, when it's a complete dog as a compile server?

0
0
Anonymous Coward

Re: re: SPARC hardware

> So, please, kindly illuminate all of us here: exaclty why is a SPARC Big Cheese so appropriate for a relational database server, when it's a complete dog as a compile server?

Multithreading. A compiler is doing one thing, sequentially, an RDBMS server is doing thousands of things in parallel. If you think they are comparable you have no clue about computing.

3
1
Roo
Bronze badge

Re: re: SPARC hardware

Often in a large source tree you can exploit quite a lot of parallelism make -j is your friend here when it comes to keeping all those cores busy. Also some compilers have been written so that the stages run in parallel, piping the results from one to stage to the next. As for "thousands of things in parallel" some source trees are huge, with 100,000s of files to be munged in a build, so in principle I think you can find compiling jobs that would be comparable to running a DB. YMMV of course. :)

1
1
Silver badge

Re: re: SPARC hardware

My experience of parallel make on huge source trees is that storage bandwidth is more of a bottleneck than CPU.

0
0
Anonymous Coward

Re: re: SPARC hardware

Depends entirely on the software you are building. I've seen C++ code taking 20 minutes wall time to compile just one file on SPARC, and it was all memory and CPU-bound, not I/O-bound. Writing the *.o to disk at the end does not take that much.

0
0
Anonymous Coward

Controlled descent into terrain

Services + Support was 51% last year, now up to 53% of total revenues.

Take notice of this trend.

Milking the existing customers at the expense of new businesses is typical of a company that is past its peak.

Existing customers are fed-up of increased costs and getting milked. Being held hostages -- as their critical systems are too "Oracle encumbered" -- they slowly abandon ship, replacing subsystem after subsystem to reduce their own costs wherever feasible. But realize that this process takes years.

It is a slow process of becoming irrelevant. Indicators to watch: top line (revenue) lack of growth and service/support percentage of revenue. Earnings can be more easily manipulated so they should be ignored.

For more perspective, look around you: all new, up-and-coming, at scale companies (Google, Facebook, Twitter, Amazon...) use distributed commodity hardware, cloud or not, with Free databases, relational (MariaDB, Percona, PostgresSQL) or not (mongo, cassandra, redis, hadoop+Impala, etc.). The number of good options here is large and the ROI of replacing Oracle makes the switch a no brainer. All those who start from scratch, without historical baggage, do their data without Oracle.

Depending on Oracle is proving to be a very costly business mistake. Lesson learned: when their salespeople and lawyers come calling, run away fast.

Their fall may not be as spectacuar as Sun's because their huge installed base (>95% of Fortune 500 companies are literally hostages) but make no mistake: they are a dinosaur in a world of much better alternatives. Survival of the fittest, plain and simple.

4
0
Anonymous Coward

Re: Controlled descent into terrain

Great insight, could not have said it better...

0
0
Silver badge

Now's the time for Larry to open the checkbook....

He's already got Mike Olsen in his speed dial ....

Time to buy Cloudera.

Just saying!

0
0

Arrogant fuckers!! They'll get whats coming to them , rule No.1 Listen to your customer!!!!!

0
0
This topic is closed for new posts.