Thread regarding Oracle Corp. layoffs

@OHQYJRs-3ozy

Agree with you. Also everyone knows AWS and Azure are much better than OPC.

Unfortunately Oracle can not beat them even though Oracle spends large amount money to invest in building Oracle's Cloud Data Center using Oracle's X86, storage and Oracle Linux. It is too late.

If I am a customer, I wouldn't bet my IT on SPARC and Solaris because I don't know how long it will stay on the market.

You guys shouldn't bet your career on it too because all these accelerator projects will become obsolete when you stop SPARC processor development. And AWS already has GPU accelerated cloud service now with just a few clicks to turn it on.

@OHQYJRs-2qqu OPC is competing with other public clouds? where exactly? ah! sorry, you were joking, yeah funny.... and you may have missed that O actually have a SPARC based OPC offering, that will be also improved soon.

@OHQYJRs-2fvk:

Found your post super funny despite the tough times.

I was counting on the 26 accelerators too!!

Every other real cloud provider develops it's own accelerator hardware. Wouldn't Oracle do the same. (Or I guess it could be argued O will never be a real cloud provider. )

@OHQYJRs-2rfw Cloud is already rolled out on x86 as it should be. OPC is competing with other public clouds. It just needs to use the hardware and software everyone is expecting, plain and simple. Some incestuous cloud offering running Solaris on SPARC would be a disaster and whoever stood his or her ground in those meetings to keep SPARC out of the cloud made the right call and the revenue/stock bump proves it. I say this as a literal part of the cancer weighing down Oracle's profits. Making a case for the continuation of SPARC is just impossible.

Is there any truth to the post here that the *ax products are canned? Man, I thought all 26 accelerator projects (AAX, BAX, CAX, DAX, EAX...) were going to keep me employed.

@OHQYJRs-2zgl

With the proper commitment, support and incentives, it should be possible to sell M7 and M8 based systems for 2 to 4 years, as long as the upper management has the desire and willingness to do so.

If we cant sell those outside (on premise), it should be possible to use them in our own cloud.

Its clear what the future holds. The M8 can be cutting edge but Moore's law dictates that won't last for more than a year. With no plans for M9 then a second law comes in to play. Murphy's law takes over.

Simply put you probably have enough time to scramble, get new certifications and train and interview. If you start now the severance package might just cary you through so you only miss a paycheck or two.

@OHQYJRs-1tcz Now that 10% won't survive either. Since there is no hope on M9.

@OHQYJRs-1tcz what is the implication of the rejection of the accelerator proposals ?

these proposals were not including SPARC to start with . I rather have a proposal that includes sparc instead of one small HW product that only requires 10% of the headcount if any.

This from MS? RL?

All accelerator options have been rejected. Now - not hear for fear mongering. Some of the HW folks are still under the impression this will survive.

Lists not made yet. Soon.

It is starting to feel like people are crying wolf

Another wave may coming next Thurs. Fasten your seat belt !!

Itanium family of systems were actually shipped in descent volume, but clearly it did not fly; if you don't know, there were some generations of Itanium microprocessors introduced. Not just said it loud.

HP / Intel just did something like this. They introduced a new generation of Itanium, and said out loud: this is it, no more after this.

If sparc is actually good, they might sell it off. But it does not matter how good sparc is if only Oracle is selling it.

And very fast CPUs are only impressive to people who do not know how to program massively scaled multiple processors without interlocks, which is something LE never trusted.

It's very interesting to see these kind of talks; some of what you hear from "internal sources" are actually supposed to be internal confidential info and probably not supposed to be out. For those who talk these info outside of the company would face some criminal charges potentially... Or these are because of the lack of company's information security awareness in the organization...

Yes, I agree with your analysis, O will still layoff people to reduce cost, so eventually it will be small group that run effectively!

Oracle HW sales is dead. The deck will be cleared soon. They might continue small dev and maintain the base through partners. Look at the trend after Sun acquisition. It has contracted 20% every year, writing on the wall. Many customers get it. The rest will soon find out. Some fools are still hanging around. It is only good for folks who want to retire in a year.

I probably don't understand. You are asking why Oracle is NOT doing layoffs?

If i might answer, SPARC is still generating nice revenue. Development cost's are irrelevant, since even if they cancel SPARC, they will have ho keep most of hardware engineers, because they are also developing lot of ASIC controllers. And it gives them competitive advantages.

Having own platform developed in-house have some nice advantages. They have features which no other processor can offer (software in silicon, real time encryption, memory protection). And they are not dependent on single/outside supplier. Intel can anytime dump them in favor of Microsoft cartel for example.

From technical side, SPARC M7 is still most powerful procesor in this world. After 3 years there is no other CPU which can match it in general performance. Plus they have accelerators for SQL and data processing, which makes it even better suited for database server.

Last thing i will mention is, that today is becoming generally accepted, that Intel already lost manufacturing process advantage with slowing development progress. On the other hand, foundries are still "on track" with 1-2 year cadence. By 2021 foudries might be 2 generations ahead of Intel. And this process is available for Oracle. For SPARC M9 that means that they will be able to use more cores or bigger SRAM cache. Or on the other hand, significantly reduce die size (which will reduce costs).

So M8/M9 will have superior architecture, will be manufactured on superior process and most probably will be significantly cheaper.

Seriously, i can't predict management decisions, but dumping SPARC today will be bad idea.