Oracle, .... a busted flush with no future leading IP
And who thinks those tall tales, and any and/or all of any good goods news in those carefully prepared statements, are as akin to Hurd polishing a turd of a fluffed ego?
It is hard to watch the continuing decline in the Oracle systems business and not be concerned, but that is precisely what Oracle co-founder and CEO Larry Ellison wants you to do. Oracle is focusing on making the Sun businesses it has pruned (as distinct from the ones it has culled) profitable, and will worry about growth later …
SPARC T4 is doing very well, as previously announced.
SPARC T5 coming with new IP such as compression, database acceleration, more cores, and faster clock speeds.
There are still a lot of unreleased technology (but tested in limited quantities), such as transactional memory (may show up in T5?)
There are still a lot of very interesting patented silicon technology, such as Proximity Computing.
Let the games continue!
"There are still a lot of unreleased technology (but tested in limited quantities), such as transactional memory (may show up in T5?)"
Not going to happen. Not because Oracle doesn't have the cash, but because they understand that to put billions into CPU development would be to throw it away. x86 is where the vast majority of people will be in 5-10 years, rightly or wrongly. IBM Power will probably still chug along, much like IBM has done in previous generations (AS/400 is the last of the minis, z is the last of the mainframes, AIX will be the last of the Unixs). Oracle practically admits as much by making all of their Exa flagship systems x86 based. The opportunity cost and cost generally is too great to invest in CPU for 15% of the declining Unix market which is going to be relegated to the highest of the high end. Could they invest billions and come up with something competitive with IBM's Power in a few years? Probably... but Safra would rather spend that cash on high margin software or hardware which requires the purchase of high margin software, Exa. Oracle won't openly say it, but I think they are fine with ceding the Unix market to IBM. It is a blip on their total revenues and profit and doesn't do anything to promote their real business, software.
What utter nonsense. I've never seen Oracle invest in the future more. The Cloud is growing faster for Oracle than other cloud vendors, Fusion is starting to take hold, R&D has been in the 1-1.2 Billion range each quarter for the past two years (that's well over 4 billion a year in R&D). That's all investment in the future. That's in line with IBM, almost double SAP, and about 25% more than HP.
"I heard the cloud services team is being forced to use Exa-data and Exa-Logic to double count the hardware dollars and cloud dollars"
That is the Oracle "cloud." They had Oracle OnDemand customers. They are now selling Oracle software and hardware to those same customers that have always used Oracle OnDemand and wrapping the whole thing up as "cloud", like it some new offering that went from zero to over a billion in revenue practically overnight. "Cloud" is not a business unit that they need to report to in audited financials, so many of those Oracle DB dollars, Oracle Exa dollars, etc are the "cloud" (a double count). Then Oracle can say, "we are the largest enterprise cloud provider in the world"... really, because you were lambasting "cloud" as ridiculous like six months ago? Very much in the VMware spirit of just calling what they already sell and have been selling for a decade "the cloud." If analysts want "cloud" revenue, "cloud" revenue they shall have. This all just speaks to the absurdity of the "cloud" it is everything and nothing all at the same time. Marketing gimmick.
I liked Larry's original stance on the cloud, but someone must have told him that they are going to get heat from the analysts as an antiquated business model if they don't have "cloud."
Larry being Larry, totally agree with his statement:
If you pay attention to what Larry said he was not against cloud, he was against the term and in general everyone's discovery of something that had been around since time share systems were created. Very consistent in my opinion. The only thing that Larry was late with on Cloud was adopting the term, since their cloud service was around since the early to mid 2000's.
>I heard the cloud services team is being forced to use Exa-data and Exa-Logic to double count the hardware >dollars and cloud dollars. Rumor also has it they are not happy with Sun x86 hardware. Notice how they said >Sun hardware.
I only notice how you say these things. You like to say "I heard" allot and then act like it's fact.
I have it on great authority that the Exa-folks are very happy with the Sun Hardware, especially when compared with what they got from HP. I don't have to make up "rumors".
"I only notice how you say these things. You like to say "I heard" allot and then act like it's fact..... I have it on great authority that the Exa-folks are very happy with the Sun Hardware, especially when compared with what they got from HP."
You are apparently not familiar with the concept of irony. As you just criticized her for stating something she heard as fact... and then proceeded to state something you heard as fact.
Regardless, she isn't talking about Exa design teams who are using Sun x86 nodes saying they don't like Sun hw as much as HP, she is talking about Oracle "cloud", i.e. Oracle Services, not liking the Exadata, Logic as compared to whatever they were using before Exa came along.
"You are apparently not familiar with the concept of irony. As you just criticized her for stating something she heard as fact... and then proceeded to state something you heard as fact."
LMAO. Pointing out irony to someone that was obvious being ironic is very ironic. Irony is dead. Long live sarcasm!
"We're just about finishing with the downsizing phase," Ellison said cheerfully.
==> yeah after two years of firing, cancelling, delaying, cutting so they squeeze every penny out of the install base they are amost done. There is no future in hardware at Oracle, but then again neither is there at Teradata or what was Netezza. Just OEM the Intel boards.
Hardware systems products sales fell by 23 per cent, to $734m, and support revenues on the installed Oracle/Sun hardware base fell by 6 per cent to $587m. Add them up, and the combined hardware systems sales fell by 16 per cent to $1.32bn.
==> Mark Turd is doing the same thing to the Oracle business he did to HP, the only difference it is a cash cow to the software business.
===> Anyone that buys those Sun x86 servers embedded in "EXA-crap" better get ready for RAC cluster evacuations, and rebuilds.
In his own words - he doesn't care if Sun hardware goes to $0
(from October 2012)
Engineered systems on the other hand - are a different matter, those
are growing nicely.
Larry says when the commodity stuff goes to zero then the growth will
show because the stuff that is dying won't be a drag any more.
Well, in fairness, he said that he doesn't care if their general purpose x86 hardware for running MS, RHEL and what have you goes to zero and he wants out of the general purpose x86, non-Exa, business.
I do agree, and Oracle openly states, that they want to sell Exa first. They will sell SPARC, if they can't sell Exa, but it is not preferable for obvious financial reasons.
There are obviously less SPARC customers due to fears of what Oracle may or may not do to SPARC. However, it can’t be ignored that the current server line is offering better performance than the older Sun servers for a fraction of a cost. This is impacting the number of servers sold and profit. The only question is will this help drive long term growth. Based on the SPARC roadmap, including a new CPU (to be called M4?) and server line, and the SPARC Super Cluster (SSC) offering general purpose as well as running Exadata and Exalogic Oracle clearly believes that there is still a market. I expect them to move SSC as the preferred engineered system once the newer SSC is announced. Because of this and the performance improvements in SPARC, I believe that it’s very likely that SPARC will make a comeback. The only question is if Oracle will foster the much needed customer base. They’re getting better, but they are not even close to caring for their customers as Sun did.
I don't see SSC ever becoming the "preferred engineered system" at least in current general purpose form. If Sparc can surpass Xeon in OLTP performance, it is possible that Oracle will use SPARC instead of Xeon in Exadata, but the primary motivation for Oracle's hw business is to drag EE software licenses. I doubt Oracle has any interest in selling someone a server and then letting them put WebSphere or DB2 on it. The bulk of the profit is in the software..... Although people in the IT space are, apparently, passionate about SPARC. Look at SPARC vs. Xeon in say an Exadata from Oracle's financial perspective. We are talking in the $10,000 or so in additional margin capture (the differential between Xeon and SPARC profit) for Oracle by moving from Xeon to SPARC.... when you are talking about a multi-million system, that is peanuts. It isn't like there is a huge financial incentive for Oracle to use SPARC instead of Xeon. The money is in the software.
Biting the hand that feeds IT © 1998–2019